EGU General Assembly 2023  

Traditional post-disaster building damage assessments were performed manually by the response team, which was risky and time-consuming. With advanced remote sensing technology, such as an unmanned aerial vehicle (UAV), it would be possible to acquire high-quality aerial videos and operate at a variety of altitudes and angles.  The collected data would be sent into a neural network for training and validating. In this study, the Object Detection model (YOLO) was utilized, which is capable of predicting both bounding boxes and damage levels. The network was trained using the ISBDA dataset, which was created from aerial videos of the aftermath of Hurricane Harvey in 2017, Hurricane Michael and Hurricane Florence in 2018, and three tornadoes in 2017, 2018, and 2019 in the United States. The Joint Damage Scale was used to classify the buildings in this dataset into four categories: no damage, minor damage, major damage, and destroyed. However, the number of major damage and destroyed classes are significantly lower than the number of no damage and minor damage classes in the dataset. Also, the damage characteristics of minor and major damage classes are similar under such type of disaster. These caused the YOLO model prone to misclassify the intermediate damage levels, i.e., minor and major damage in our earlier experiments. This study aimed to improve the YOLO model using a stacking ensemble deep learning approach with a image classification model called Mobilenet. First, the ISBDA dataset was used and refined to train the YOLO network and the Mobilenet network separately, and the latter provides two classes predictions (0 for no damage or minor damage, 1 for major damage or destroyed) rather than the four classes by the former. In the inference phase, the initial predictions from the trained YOLO network, including bounding box coordinates, confidence scores for four damage classes, and the predicted class, were then extracted and passed to the trained Mobilenet to generate the secondary predictions for each building. Based on the secondary predictions, two hyperparameters were utilized to refine the initial predictions by modifying the confidence scores of each class, and the hyperparameters were trained during this phase. Lastly, the trained hyperparameters were applied to the testing dataset to evaluate the performance of the proposed method. The results show that our stacking ensemble method could obtain more reliable predictions of intermediate classes.

How to cite: Sim, L., Tsai, F.-J., and Lin, S.-Y.: A Stacking Ensemble Deep Learning Approach for Post Disaster Building Assessment using UAV Imagery, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-338, https://doi.org/10.5194/egusphere-egu23-338, 2023.

        Since Taiwan is located at the Pacific Ring of Fire, seismic activity of varying magnitudes occurs almost every day. Among them, some of these seismic activities have in turn caused severe disasters, resulting in loss of personal property, casualties and damage to important public facilities. Therefore, investigating the long-term spatiotemporal pattern of seismic activities is a crucial task for understanding the causes of seismic activity and to predict future seismic activity, in order to carry out disaster prevention measures in advance. Previous studies mostly focused on the causes of single seismic events on the small spatiotemporal scale. In this study, the data from 1987 to 2020 are used, including seismic events from the United States Geological Survey (USGS), the ambient environmental factors such as daily air temperature from Taiwan Central Weather Bureau (CWB) and daily sea surface temperature data from National Oceanic and Atmospheric Administration (NOAA). Then the temperature difference between the land air temperature and the sea surface temperature (SST) to the correlation between the occurrence of seismic activities and the abnormal occurrence of temperature difference are compared. The results show that lots of seismic activities often have positive and negative anomalies of temperature difference from 21 days before to 7 days after the seismic event. Moreover, there is a specific trend of temperature difference anomalies under different magnitude intervals. In the magnitude range of 2.5 to 4 and greater than 6, almost all of the seismic events have significant anomalous differences in the temperature difference between land air temperature and SST compared with no seismic events. This study uncovers anomalous frequency signatures of seismic activities and temperature differences between land air temperature and SST. The significant difference in temperature difference between seismic events and non-seismic events was compared by using statistical analysis. Additionally, the deep neural network (DNN) of deep learning model, logistic regression and random forest of machine learning model was used to identify whether there will be a seismic event under different magnitude intervals. It is hoped that it can provide relevant information for the prediction of future seismic activity, to more accurately prevent disasters that may be caused by seismic activity.

How to cite: Chen, Y.-H. and Lin, Y.-C.: Investigating the Correlation between the Characteristics of Seismic Activity and Environmental Variables in Taiwan, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2564, https://doi.org/10.5194/egusphere-egu23-2564, 2023.

The 2010-2011 Canterbury Earthquake sequence (CES) led to unprecedented building damage in the Canterbury region, New Zealand. Commercial and residential buildings were significantly affected. Due to New Zealand’s unique insurance setting, around 80% of the losses were covered by insurance (Bevere & Balz, 2012; King et al., 2014). The Insurance Council of New Zealand (ICNZ) estimated the total economic losses to be more than NZ$40 billion, with the Earthquake Commission (EQC) and private insurers covering NZ$10 billion and NZ$21 billion of the losses, respectively (ICNZ, 2021). As a result of the CES and the 2016 Kaikoura earthquake, EQC’s Natural Disaster Fund was depleted (EQC, 2022). This highlighted the need for improved tools enabling damage and loss analysis for natural hazards.
This research project used residential building claims collected by EQC following the CES to develop a rapid seismic loss prediction model for residential buildings in Christchurch. Geographic information systems (GIS) tools, data science techniques, and machine learning (ML) were used for the model development. Before the training of the ML model, the claims data was enriched with additional information from external data sources. The seismic demand, building characteristics, soil conditions, and information about the liquefaction occurrence were added to the claims data. Once merged and pre-processed, the aggregated data was used to train ML models based on the main events in the CES. Emphasis was put on the interpretability and explainability of the model. The ML model delivered valuable insights related to the most important features contributing to losses. Those insights are aligned with engineering knowledge and observations from previous studies, confirming the potential of using ML for disaster loss prediction and management. Care was also put into the retrainability of the model to ensure that any new data from future earthquake events can rapidly be added to the model. 

How to cite: Roeslin, S.: Development of a Rapid Seismic Loss Prediction Model for Residential Buildings using Machine Learning - Christchurch, New Zealand, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2996, https://doi.org/10.5194/egusphere-egu23-2996, 2023.

We investigate the use of CCTV cameras and deep learning to automatically monitor trash screen blockage. 

Trash screens are installed to prevent debris from entering critical parts of river networks (pipes, tunnels, locks,...). When the debris piles up at the trash screens,  it  may block the waterway and can cause flooding. It is thus crucial to clean blocked trash screens and avoid flooding and consequent damage. Currently, the maintenance crews must manually check a camera or river level data or go on site to check the state of the screen to know if it needs cleaning. This wastes valuable time in emergency situations where blocked screens must be urgently cleaned (e.g., in case of forecast  heavy rainfall). Some initial attempts at trying to predict trash screen blockage exist. However, these have not been widely adopted in practice.  CCTV cameras can be easily installed at any location and can thus be used to monitor the state of trash screens, but the images need to be processed by an automated algorithm to inform whether the screen is blocked.

With the help of UK-based practitioners (Environment Agency and local councils), we have created a dataset of 40000 CCTV trash screen images coming from 36 cameras, each labelled with blockage information. Using this database, we have compared 3 deep learning approaches to automate the detection of trash screen blockage: 

• A binary image classifier, which takes as input a single image, and outputs a binary label that estimates whether the trash screen is blocked.
• An approach based on anomaly detection which tries to reconstruct the input image with an auto-encoder trained on clean trash screen images.  In consequence, blocked trash screens are detected as anomalies by the auto-encoder.
• An image similarity estimation approach based on the use of a siamese network, which takes as input two images and outputs a similarity index related, in our case, to whether both images contain trash. 

Using performance criteria chosen in discussion  with practitioners (overall accuracy, false alarm rate, resilience to luminosity / moving fields of view, computing capabilities), we show that deep learning can be used in practice to automate the identification of blocked trash screens. We also analyse the strengths and weaknesses of each of these approaches and provide guidelines for their application.

How to cite: Vandaele, R., Dance, S. L., and Ojha, V.: Comparison of deep learning approaches to monitor trash screen blockage from CCTV cameras, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3928, https://doi.org/10.5194/egusphere-egu23-3928, 2023.

Roadways are critical infrastructure in our society, providing services for people through and between cities. However, they are prone to closures and disruptions, especially after extreme weather events like hurricanes.

At the same time, traffic flow data are a fundamental type of information for any transportation system.

We tackle the problem of traffic sensor placement on roadways to address two tasks at the same time. The first task is traffic data estimation in ordinary situations, which is vital for traffic monitoring and city planning. We design a graph-based method to estimate traffic flow on roads where sensors are not present. The second one is enhanced observability of roadways in case of extreme weather events. We propose a satellite-based multi-domain risk assessment to locate roads at high risk of closures. Vegetation and flood hazards are taken into account. We formalize the problem as a search method over the network to suggest the minimum number and location of traffic sensors to place while maximizing the traffic estimation capabilities and observability of the risky areas of a city.

How to cite: Gazzea, M., Arghandeh, R., and Miraki, A.: Traffic Monitoring System Design considering Multi-Hazard Disaster Risks, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4455, https://doi.org/10.5194/egusphere-egu23-4455, 2023.

Earthquake-induced land deformation and structure failure are more severe over soft soils than over firm soils and rocks owing to the seismic site effect and liquefaction. The site-specific seismic site effect related to the amplification of ground motion, liquefaction, and landslide has spatial uncertainty depending on the local subsurface, surface geological, and topographic conditions. When the 2017 Pohang earthquake (M 5.4), South Korea’s second strongest earthquake in decades, occurred, the severe damages influenced by variable site response and vulnerability indicators were observed focusing on the basin or basin-edge region deposited unconsolidated Quaternary sediments. Thus, nationwide site characterization is essential considering empirical correlations with geotechnical site response and hazard parameters and surface proxies. Furthermore, in case of so many variables and tenuously related correlations, machine learning classification models can prove to be very precise than the parametric methods. This study established a multivariate seismic site classification system using the machine learning technique based on the geospatial big data platform.

The supervised machine learning classification techniques and more specifically, random forest, support vector machine (SVM), and artificial neural network (ANN) algorithms have been adopted. Supervised machine learning algorithms analyze a set of labeled training data consisting of a group of input data and desired output values. They produce an inferred function that can be used for predictions from given input data. To optimize the classification criteria by considering the geotechnical uncertainty and local site effects, the training datasets applying principal component analysis (PCA) were verified with k-fold cross-validation. Moreover, the optimized training algorithm, proved by loss estimators (receiver operating characteristic curve (ROC), the area under the ROC curve (AUC)) based on confusion matrix, was selected.

For the southeastern region in South Korea, the boring log information (strata, standard penetration test, etc.), geological map (1:50k scale), digital terrain model (having 5 m × 5 m), soil map (1:250k scale) were collected and constructed as geospatial big data. Preliminarily, to build spatially coincided datasets with geotechnical response parameters and surface proxies, the mesh-type geospatial information was built by advanced geostatistical interpolation and simulation methods.

Site classification systems use seismic hazard parameters related to the geotechnical characteristics of the study area as the classification criteria. The current site classification systems in South Korea and the United States recommend Vs30, which is the average shear wave velocity (Vs) up to 30 m underground. This criterion uses only the dynamic characteristics of the site without considering its geometric distribution characteristics. Thus, the geospatial information included the geo-layer thickness, surface proxies (elevation, slope, geological category, soil category), and Vs30. For the liquefaction and landslide hazard estimation, the liquefaction vulnerability indexes (i.e., liquefaction potential or severity index) and landslide vulnerability indexes (i.e., a factor of safety or displacement) were also trained as input features into the classifier modeling. Finally, the composite status against seismic site effect, liquefaction, and landslide was predicted as hazard class (I.e., safe, slight-, moderate-, extreme-failure) based on the best-fitting classifier.  

How to cite: Kim, H.: Machine Learning-based Site Classification System for Earthquake-Induced Multi-Hazard in South Korea, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4757, https://doi.org/10.5194/egusphere-egu23-4757, 2023.

The automatic anticipation and detection of extreme events constitute a major challenge in the current context of climate change, which has changed their likelihood and intensity. One of the main objectives within the EXtreme Events: Artificial Intelligence for Detection and Attribution (XAIDA) project (https://xaida.eu/) is related to developing novel approaches for the detection and localization of extreme events, such as tropical cyclones and severe convective storms, heat waves and droughts, as well as persistent winter extremes, among others. Here we introduce the XAIDA4Detection toolbox that allows for tackling generic problems of detection and characterization. The open-source toolbox integrates a set of advanced ML models, ranging in complexity, assumptions, and sophistication, and yields spatio-temporal explicit detection maps with probabilistic heatmap estimates. We included supervised and unsupervised methods, deterministic and probabilistic, neural networks based on convolutional and recurrent nets, and density-based methods. The toolbox is intended for scientists, engineers, and students with basic knowledge of extreme events, outlier detection techniques, and Deep Learning (DL), as well as Python programming with basic packages (Numpy, Scikit-learn, Matplotlib) and DL packages (PyTorch, PyTorch Lightning). This presentation will summarize the available features and their potential to be adapted to multiple extreme event problems and use cases.

How to cite: Cortés-Andrés, J., Gonzalez-Calabuig, M., Zhang, M., Williams, T., Fernández-Torres, M.-Á., Pellicer-Valero, O. J., and Camps-Valls, G.: XAIDA4Detection: A Toolbox for the Detection and Characterization of Spatio-Temporal Extreme Events, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4816, https://doi.org/10.5194/egusphere-egu23-4816, 2023.

Damage assessment is a critical step in crisis management. It must be fast and accurate in order to organize and scale the emergency response in a manner adapted to the real needs on the ground. The speed requirements motivate an automation of the analysis, at least in support of the photo-interpretation. Deep Learning (DL) seems to be the most suitable methodology for this problem: on one hand for the speed in obtaining the answer, and on the other hand by the high performance of the results obtained by these methods in the extraction of information from images. Following previous studies to evaluate the potential contribution of DL methods for building damage assessment after a disaster, several conventional Deep Neural Network (DNN) and Transformers (TF) architectures were compared.

Made available at the end of 2019, the xView2 database appears to be the most interesting database for this study. It gathers images of disasters between 2011 and 2018 with 6 types of disasters: earthquakes, tsunamis, floods, volcanic eruptions, fires and hurricanes. For each of these disasters, pre- and post-disaster images are available with a ground truth containing the building footprint as well as the evaluation of the type of damage divided into 4 classes (no damage, minor damage, major damage, destroyed) similar to those considered in the study.

This study compares a wide range DNN architectures all based on an encoder-decoder structure. Two encoder families were implemented: EfficientNet (B0 to B7 configurations) and Swin TF (Tiny, Small, and Base configurations). Three adaptable decoders were implemented: UNet, DeepLabV3+, FPN. Finally, to benefit from both pre- and post-disaster images, the trained models were designed to proceed images with a Siamese approach: both images are processed independently by the encoder, and the extracted features are then concatenated by the decoder.

Taking benefit of global information (such as the type of disaster for example) present in the image, the Swin TF, associated with FPN decoder, reaches the better performances than all other encoder-decoder architectures. The Shifted WINdows process enables the pipe to process large images in a reasonable time, comparable to the processing time of EfficientNet-based architectures. An interesting additional result is that the models trained during this study do not seem to benefit so much from extra-large configurations, and both small and tiny configurations reach the highest scores.

How to cite: Lagrange, A., Dublé, N., De Vieilleville, F., Dupuis, A., May, S., and Walker-Deemin, A.: Vision Transformers for building damage assessment after natural disasters, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5581, https://doi.org/10.5194/egusphere-egu23-5581, 2023.

Natural and man-made disasters pose a threat to human life, flora-fauna, and infrastructure. It is critical to detect the damage quickly and accurately for infrastructures right after the occurrence of any disaster. The detection and assessment of infrastructure damage help manage financial strategy as well. Recently, many researchers and agencies have made efforts to create high-resolution satellite imageries database related to pre and post-disaster events. The advanced remote sensing satellite imageries can reflect the surface of the earth accurately up to 30 cm spatial resolution on a daily basis. These high spatial resolutions (HSR) imageries can help access any natural hazard's damage by comparing the pre- and post-disaster data. These remote sensing imageries have limitations, such as cloud occlusions. Building under a thick cloud cannot be recognised in optical images. The manual assessment of the severity of damage to buildings/infrastructure by comparing bi-temporal HSR imageries or airborne will be a tedious and subjective job. On the other hand, the emerging use of unmanned aired vehicles (UAV) can be used to assess the situation precisely. The high-resolution UAV imageries and the HSR satellite imageries can complement each other for critical infrastructure damage assessment. In this study, a novel approach is used to integrate UAV data into HSR satellite imageries for the building damage assessment using a convolution neural network (CNN) based deep learning model. The research work is divided into two fundamental sub-tasks: first is the building localisation in the pre-event images, and second is the damage classification by assigning a unique damage level label reflecting the degree of damage to each building instance on the post-disaster images. For the study, the HSR satellite imageries of 36 pairs of pre- and post natural hazard events is acquired for the year 2021-22, similarly available UAV based data for these events is also collected from the open data source. The data is then pre-processed, and the building damage is assessed using a deep object-based semantic change detection framework (ChangeOS). The mentioned model was trained on the xview2 building damage assessment datasets comprised of ~20,000 images with ~730,000 building polygons of pre and post disaster events over the globe from 2011-2018. The experimental setup in this study includes training on the global dataset and testing on the regional-scale building damage assessment using HSR satellite imageries and local-scale using UAV imageries. The result obtained from the bi-temporal assessment of HSR images for the Indonesia Earthquake 2022 has shown an F1 score of ~67%, while the Uttarakhand flooding event 2021 has reported an F1 score of ~64%. The HSR imageries from the UAV Haiti earthquake event in 2011 have also shown less but promising F1 scores of ~54%. It is inferred that merging HSR imageries from satellite and UAV for building damage assessment using the ChangeOS framework represents a robust tool to further promote future research in infrastructure maintenance strategy and policy management in disaster response.

How to cite: Gupta, S. and Nair, S.: A novel approach for infrastructural disaster damage assessment using high spatial resolution satellite and UAV imageries using deep learning algorithms., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5778, https://doi.org/10.5194/egusphere-egu23-5778, 2023.

Artificial intelligence (AI) methods have emerged as a powerful tool to study and in some cases forecast natural disasters [1,2]. Recent works have successfully combined deep learning modeling with scientific knowledge stemming from the SAR Interferometry domain propelling research on tasks like volcanic activity monitoring [3], associated with ground deformation. A milestone in this interdisciplinary field has been the release of the Hephaestus [4] InSAR dataset, facilitating automatic InSAR interpretation, volcanic activity localization as well as the detection and categorization of atmospheric contributions in wrapped interferograms. Hephaestus contains annotations for approximately 20,000 InSAR frames, covering the 44 most active volcanoes in the world. The annotation was performed  by a team of InSAR experts that manually examined each InSAR frame individually. However, even with such a large dataset, class imbalance remains a challenge, i.e. the InSAR samples containing volcano deformation fringes are orders of magnitude less than those that do not. This is anticipated since natural hazards are in principle rare in nature. To counter that, the authors of Hephaestus provide more than 100,000 unlabeled InSAR frames to be used for global large-scale self-supervised learning, which is more robust to class imbalance when compared to supervised learning [5]. 

Motivated by the Hephaestus dataset and the insights provided by [2], we train global, task-agnostic models in a self-supervised learning fashion that can handle distribution shifts caused by spatio-temporal variability as well as major class imbalances. By finetuning such a model to the labeled part of Hephaestus we obtain the backbone for a global volcanic activity alerting system, namely Pluto. Pluto is a novel end-to-end AI based system that provides early warnings of volcanic unrest on a global scale.

Pluto automatically synchronizes its database with the Comet-LiCS [6] portal to receive newly generated Sentinel-1 InSAR data acquired over volcanic areas. The new samples are fed to our volcanic activity detection model. If volcanic activity is detected, an automatic email is sent to the service users, which contains information about the intensity, the exact location and the type (Mogi, Sill, Dyk) of the event. To ensure a robust and ever-improving service we augment Pluto with an iterative pipeline that collects samples that were misclassified in production, and uses them to further improve the existing model. 

[1] Kondylatos et al. "Wildfire danger prediction and understanding with Deep Learning." Geophysical Research Letters 49.17 (2022): e2022GL099368.

[2] Bountos et al. "Self-supervised contrastive learning for volcanic unrest detection." IEEE Geoscience and Remote Sensing Letters 19 (2021): 1-5.

[3] Bountos et al. "Learning from Synthetic InSAR with Vision Transformers: The case of volcanic unrest detection." IEEE Transactions on Geoscience and Remote Sensing (2022).

[4] Bountos et al. "Hephaestus: A large scale multitask dataset towards InSAR understanding." Proceedings of the IEEE/CVF CVPR. 2022.

[5] Liu et al. "Self-supervised learning is more robust to dataset imbalance." arXiv preprint arXiv:2110.05025 (2021).

[6] Lazecký et al. "LiCSAR: An automatic InSAR tool for measuring and monitoring tectonic and volcanic activity." Remote Sensing 12.15 (2020): 2430.

How to cite: Bountos, N. I., Michail, D., Herekakis, T., Thanasou, A., and Papoutsis, I.: Pluto: A global volcanic activity early warning system powered by large scale self-supervised deep learning on InSAR data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5913, https://doi.org/10.5194/egusphere-egu23-5913, 2023.

It has become increasingly apparent over the past few decades that environmental degradation is something of a common concern for humanity and it is difficult to deny that the present environmental problems are caused primarily by anthropogenic activities rather than natural causes.

To minimize disaster’s risk, the role of geospatial science and technology may be a terribly helpful and necessary technique for hazard zone mapping throughout emergency conditions. 

This approach can definitively help predict harmful events, but also to mitigate damage to the environment from events that cannot be efficiently predicted.

With detailed information obtained through various dataset, decision making has become simpler. This fact is crucial for a quick and effective response to any disaster. Remote sensing, in particular RADAR/SAR data, help in managing a disaster at various stages. 

Prevention for example refers to the outright avoidance of adverse impacts of hazards and related disasters; preparedness refers to the knowledge and capacities to effectively anticipate, respond to, and recover from, the impacts of likely, imminent or current hazard events or conditions.

Finally relief is the provision of emergency services after a disaster in order to reduce damage to environment and people.

Thanks to the opportunity proposed by ASI (Italian Space Agency) to use COSMO-SkyMed data, in NeMeA Sistemi srl we developed two projects: “Ventimiglia Legalità”, “Edilizia Spontanea” and 3xA.

Their main objective is to detect illegal buildings not present in the land Legal registry.

We developed new and innovative technologies using integrated data for the monitoring and protection of environmental and anthropogenic health, in coastal and nearby areas. 

3xA project addresses the highly challenging problem of automatically detecting changes from a time series of high-resolution synthetic aperture radar (SAR) images. In this context, to fully leverage the potential of such data, an innovative machine learning based approach has been developed. 

The project is characterized by an end-to-end training and inference system which takes as input two raw images and produces a vectorized change map without any human supervision.

More into the details, it takes as input two SAR acquisitions at time t1 and t2, the acquisitions are firstly pre-processed, homogenised and finally undergo a completely self-supervised algorithm which takes advantage of DNNs to classify changed/unchanged areas. This method shows promising results in automatically producing a change map from two input SAR images (Stripmap or Spotlight COSMO-SkyMed data), with 98% accuracy.

Being the process automated, results are produced faster than similar products generated by human operators.

A similar approach has been followed to create an algorithm which performs semantic segmentation from the same kind of data.

This time, only one of the two SAR acquisitions is taken as input for pre-processing steps and then for a supervised neural network. The result is a single image where each pixel is labelled with the class predicted by the algorithm. 

Also in this case, results are promising, reaching around 90% of accuracy. 

How to cite: Pennino, I.: A new approach for hazard and disaster prevention: deep learning algorithms for change detection and classification RADAR/SAR, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6522, https://doi.org/10.5194/egusphere-egu23-6522, 2023.

 With the gradual warming of the global climate, natural catastrophes have caused billions of dollars in damage to ecosystems, economies and properties. Along with the damage, the loss of life is a very serious possibility. With the unprecedented growth of the human population, large-scale development activities and changes to the natural environment, the frequency, and intensity of extreme natural events and consequent impacts are expected to increase in the future. 

 To be able to mitigate and to reduce the potential damage of the natural catastrophe, continuous monitoring is required. The collection of data using earth observation (EO) systems has been valuable for tracking the effects of natural hazards, especially with their near real-time capabilities for tracking extreme natural events. Remote sensing systems from different platforms also serve as an important decision support tool for devising response strategies, coordinating rescue operations, and making damage and loss estimations.

 Synthetic aperture radar (SAR) imagery provides highly valuable information about our planet that no other technology is capable of. SAR sensors emit their own energy to illuminate objects or areas on Earth and record what’s reflected back from the surface to the sensor. This allows data acquisition day and night since no sunlight is needed. SAR also uses longer wavelengths than optical systems, which gives it the unsurpassed advantage of being able to penetrate clouds, rain, fog and smoke. All of this makes SAR imagery unprecedentedly valuable in sudden events and crisis situations requiring a rapid response.

 In this talk we will be focusing on flood monitoring using our ICEYE SAR images, taking into account multi-satellites, multi-angles and multi-resolutions that are inherent from our constellation and capabilities. We will present the different steps necessary that have allowed us to improve the consistency of our generated flood maps.

How to cite: Dupeyrat, A., Almaksour, A., Vinholi, J., and friberg, T.: Deep learning for automatic flood mapping from high resolution SAR images, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6790, https://doi.org/10.5194/egusphere-egu23-6790, 2023.

Increasing climatic extremes resulted in frequency and severity of urban flood events during the last several decades. Significant economic losses were point out the urgency of flood response. In recent years, the government gradually increased the layout of CCTV water level monitoring facilities for the purpose of decision-making in flood event. However, it is difficult for decision makers to recognize multiple images in the same time. Therefore, the aim of this study attempts to establish an automatic water level recognition method for given closed-circuit television (CCTV) system.

In the last years, many advances have been made in the area of automatic image recognition with methods of artificial intelligence. Little literature has been published on real-time water level recognition of closed-circuit television system for disaster management. The purpose of this study is to examine the possibilities in practice of artificial intelligence for real-time water level recognition with deep convolutional neural network. Proposed methodology will demonstrate with several case studies in Taichung. For the potential issue that AI models may lacks of learning target, the generative adversarial network (GAN) may be adopted for this study. The result of this study could be useful to decision makers responsible for organizing response assignments during flood event.

How to cite: Chen, B. and Li, C.-Y.: A study on the establishment of computer vision for disaster identification based on existing closed-circuit television system, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7435, https://doi.org/10.5194/egusphere-egu23-7435, 2023.

Extra-Tropical Cyclones are major systems ruling and influencing the atmospheric structure at mid-latitudes. They are characterised by strong winds and heavy precipitation, and can cause considerable storm surges potentially devastating for coastal regions. The availability of historical observations of the extreme events caused by intense ETCs are rather limited, hampering risk evaluation. Increasing the amount of significant data available would substantially help several fields of analysis influenced by these events, such as coastal management, agricultural production, energy distribution, air and maritime transportation, and risk assessment and management.

Here, we address the possibility of generating synthetic ETC atmospheric fields of mean sea level pressure, wind speed, and precipitation in the North Atlantic by training a Generative Adversarial Network (GAN). The purpose of GANs is to learn the distribution of a training set based on a game theoretic scenario where two networks compete against each other, the generator and the discriminator. The former is trained to generate synthetic examples that are plausible and resemble the real ones. The input of the generator is a vector of random Gaussian values, whose domain is known as the “latent space”. The discriminator learns to distinguish whether an example comes from the dataset distribution. The competition set by the game-theoretic approach improves the network until the counterfeits are indistinguishable from the originals.

To train the GAN, we use atmospheric fields extracted from the ERA5 reanalysis dataset in the geographic domain with boundaries 0°- 90°N, 70°W - 20°E and for the period 1^st January 1979 - 1^st January 2020. We analyse the generated samples’ histograms, the samples’ average fields, the Wasserstein distance and the Kullback-Leibler divergence between the generated samples and the test set distributions. Results show that the generative model has learned the distribution of the values of the atmospheric fields and the general spatial trends of the atmosphere in the domain. To evaluate better the atmospheric structure learned by the network, we perform linear and spherical interpolations in the latent space. Specifically, we consider four cyclones and compare the frames of their tracks to those of the synthetic tracks generated by interpolation. The interpolated tracks show interesting features consistent with the original tracks. These findings suggest that GANs can learn meaningful representations of the ETCs’ fields, encouraging further investigations to model the tracks’ temporal evolution.

How to cite: Dainelli, F., Taormina, R., Ascenso, G., Scoccimarro, E., Giuliani, M., and Castelletti, A.: Synthetic Generation of Extra-Tropical Cyclones’ fields with Generative Adversarial Networks, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8419, https://doi.org/10.5194/egusphere-egu23-8419, 2023.

Understanding the relationships between different drought drivers and observed drought impact can provide important information for early warning systems and drought management planning. Moreover, this relationship can help inform the definition and delineation of drought events. However, currently, drought hazards are often characterized based on their frequency of occurring, rather than based on the impacts they cause. A more data-driven depiction of “impactful drought events”- whereby droughts are defined by the hydrometeorological conditions that, in the past, have led to observable impacts-, has the potential to be more meaningful for drought risk assessments.

In our research, we apply a data-mining method based on association rules, namely fast and frugal decision trees, to link different drought hazard indices to agricultural impacts. This machine learning technique is able to select the most relevant drought hazard drivers (among both hydrological and meteorological indices) and their thresholds associated with “impactful drought events”. The technique can be used to assess the likelihood of occurrence of several impact severities, hence it supports the creation of a loss exceedance curve and estimates of average annual loss. An additional advantage is that such data-driven relations in essence reflect varying local drought vulnerabilities which are difficult to quantify in data-scarce regions.

This contribution exemplifies the use of fast and frugal decision trees to estimate (agricultural) drought risk in the Volta basin and its riparian countries. We find that some agriculture-dependent regions in Ghana, Togo and Côte d’Ivoire face annual average drought-induced maize production losses up to 3M USD, while per hectare, losses can mount to on average 50 USD/ha per year in Burkina Faso. In general, there is a clear north-south gradient in the drought risk, which we find augmented under projected climate conditions. Climate change is estimated to worsen the drought impacts in the Volta Basin, with 11 regions facing increases in annual average losses of more than 50%.

We show that the proposed multi-variate, impact-based, non-parametric, machine learning approach can improve the evaluation of droughts, as this approach directly leverages observed drought impact information to demarcate impactful drought events. We evidence that the proposed technique can support quantitative drought risk assessments which can be used for geographic comparison of disaster losses at a sub-national scale.

How to cite: Wens, M., Hamed, R., de Moel, H., Massabo, M., and Mapelli, A.: Towards probabilistic impact-based drought risk analysis – a case study on the Volta Basin, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8944, https://doi.org/10.5194/egusphere-egu23-8944, 2023.

The Canada Centre for Mapping and Earth Observation (CCMEO) uses Radarsat Constellation Mission (RCM) data for near-real time flood mapping. One of the many advantages of using SAR sensors, is that they are less affected by the cloud coverage and atmospheric conditions, compared to optical sensors. RCM has been used operationally since 2020 and employs 3 satellites, enabling lower revisit times and increased imagery coverage. The team responsible for the production of flood maps in the context of emergency response are able to produce maps within four hours from the data acquisition. Although the results from their automated system are good, there are some limitations to it, requiring manual intervention to correct the data before publication. Main limitations are located in urban and vegetated areas. Work started in 2021 to make use of deep learning algorithms, namely convolutional neural networks (CNN), to improve the performances of the automated production of flood inundation maps. The training dataset make use of the former maps created by the emergency response team and is comprised of over 80 SAR images and corresponding digital elevation model (DEM) in multiple locations in Canada. The training and test images were split in smaller tiles of 256 x 256 pixels, for a total of 22,469 training tiles and 6,821 test tiles. Current implementation uses a U-Net architecture from NRCan geo-deep-learning pipeline (https://github.com/NRCan/geo-deep-learning). To measure performance of the model, intersection over union (IoU) metric is used. The model can achieve 83% IoU for extracting water and flood from background areas over the test tiles. Next steps include increasing the number of different geographical contexts in the training set, towards the integration of the model into production.

How to cite: Turgeon-Pelchat, M., McGrath, H., Esfahani, F., Tolszczuk-Leclerc, S., Rainville, T., Svacina, N., Zhou, L., Langari, Z., and Houngbo, H.: Improving near real-time flood extraction pipeline from SAR data using deep learning, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9091, https://doi.org/10.5194/egusphere-egu23-9091, 2023.

Throughout the past couple years, changes in global climate have been turning wildfires into an increasingly unpredictable phenomenon. Many environmental parameters that have been linked to wildfires, such as the number of consecutive hot days, are becoming increasingly unstable. This leads to a twofold problem: adequate fire risk assessment is at the same time more important and more difficult than ever. 

In the past, physical models were the prevalent approach to most questions in the domain of wildfire science. While they tend to provide accurate and transparent results, they require domain expertise and often tedious manual data collection.

In recent years, increased computation capabilities and the improved availability of remote sensing data associated with the new space movement have made deep learning a beneficial approach. Data-driven approaches often yield state of the art performance without requiring expert knowledge at a fraction of the complexity of physical models. The downside, however, is that they are often intransparent and offer no insights into their inner algorithmic workings. 

We want to shed some light on this interpretability/performance tradeoff and compare different approaches for predicting wildfire hazard. We evaluate their strengths and weaknesses with a special focus on explainability. We built a wildfire hazard model for South America based on a spatiotemporal CNN architecture that infers fire susceptibility from environmental conditions that led to fire in the past. The training data used contains selected ECMWF ERA5 Land variables and ESA world cover information. This means that our model is able to learn from actual fire conditions instead of relying on theoretical frameworks. Unlike many other models, we do not make simplifying assumptions such as a standard fuel type, but calculate hazard ratings based on actual environmental conditions. Compared to classical fire hazard models, this approach allows us to account for regional and atypical fire behavior and makes our model readily adaptable and trainable for other ecosystems, too.

The ground truth labels are derived from fusing active fire remote sensing data from 20 different satellites into one active wildfire cluster data set. The problem itself is highly imbalanced with non-fire pixels making up 99.78% of the training data. Therefore we evaluate the ability of our model to correctly predict wildfire hazard using metrics for imbalanced data such as PR-AUC and F1 score. We also compare the results against selected standard fire hazard models such as the Canadian Fire Weather Index (FWI). 

In addition, we assess the computational complexity and speed of calculating the respective models and consider the accuracy/complexity/speed tradeoff of the different approaches. Furthermore, we aim to provide insights why and how our model makes its predictions by leveraging common explainability methods. This allows for insights into which factors tend to influence wildfire hazard the most and to optimize for relatively lightweight, yet performant and transparent architectures.

How to cite: Strebl, J., Gottfriedsen, J., Laux, D., Helleis, M., and Tresp, V.: Fire hazard modelling with remote sensing data for South America, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9426, https://doi.org/10.5194/egusphere-egu23-9426, 2023.

For recent years, Machine Learning (ML) models have been proven to be useful in solving problems of a wide variety of fields such as medical, economic, manufacturing, transportation, energy, education, etc. With increased interest in ML models and advances in sensor technologies, ML models are being widely applied even in civil engineering domain. ML model enables analysis of large amounts of data, automation, improved decision making and provides more accurate prediction. While several state-of-the-art reviews have been conducted in each sub-domain (e.g., geotechnical engineering, structural engineering) of civil engineering or its specific application problems (e.g., structural damage detection, water quality evaluation), little effort has been devoted to comprehensive review on ML models applied in civil engineering and compare them across sub-domains. A systematic, but domain-specific literature review framework should be employed to effectively classify and compare the models. To that end, this study proposes a novel review approach based on the hierarchical classification tree “D-A-M-I-E (Domain-Application problem-ML models-Input data-Example case)”. “D-A-M-I-E” classification tree classifies the ML studies in civil engineering based on the (1) domain of the civil engineering, (2) application problem, (3) applied ML models and (4) data used in the problem. Moreover, data used for the ML models in each application examples are examined based on the specific characteristic of the domain and the application problem. For comprehensive review, five different domains (structural engineering, geotechnical engineering, water engineering, transportation engineering and energy engineering) are considered and the ML application problem is divided into five different problems (prediction, classification, detection, generation, optimization). Based on the “D-A-M-I-E” classification tree, about 300 ML studies in civil engineering are reviewed. For each domain, analysis and comparison on following questions has been conducted: (1) which problems are mainly solved based on ML models, (2) which ML models are mainly applied in each domain and problem, (3) how advanced the ML models are and (4) what kind of data are used and what processing of data is performed for application of ML models. This paper assessed the expansion and applicability of the proposed methodology to other areas (e.g., Earth system modeling, climate science). Furthermore, based on the identification of research gaps of ML models in each domain, this paper provides future direction of ML in civil engineering based on the approaches of dealing data (e.g., collection, handling, storage, and transmission) and hopes to help application of ML models in other fields.

How to cite: Kim, J. and Jung, D.: State-of-the-Art Review of Machine Learning Models in Civil Engineering: Based on DAMIE Classification Tree, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11636, https://doi.org/10.5194/egusphere-egu23-11636, 2023.

Digital twin is now recognized as digital copies of physical world's objects stored in digital space and utilized to simulate the sequences and consequences of target phenomena. By incorporating physical world’s data into the digital twin, developers and users have a full view of the target through real-time feedback. Recent advances in high-performance computing and large-scale data fusion of sensing and observations of both natural and social phenomena are enhancing applicability of digital twin paradigm to natural disaster research. Artificial intelligence (AI) and machine learning are also being applied more and more widely across the world and contributing as essential elements of digital twin. Those have significant implications for disaster response and recovery to hold out the promise of dramatically improving our understanding of disaster-affected areas and responses in real-time.

A project is underway to enhance resilience of disaster response systems by constructing "Disaster Digital Twin" to support disaster response team in the anticipated tsunami disaster. “Disaster Digital Twin” platform consists of a fusion of real-time hazard simulation, e.g. tsunami inundation forecast, social sensing to identify dynamic exposed population, and multi-agent simulation of disaster response activities to find optimal allocation or strategy of response efforts, and achieve the enhancement of disaster resilience.

To achieve the goal of innovating digital twin computing for enhancing disaster resilience, four preliminary results are shown;

(1) Developing nation-wide real-time tsunami inundation and damage forecast system. The priority target for forecasting is the Pacific coast of Japan, a region where Nankai trough earthquake is likely to occur.

(2) Establishing a real-time estimation of the number of exposed population in the inundation zone and clarifying the relationship between the exposed population and medical demand.

(3) Developing a reinforcement learning-based multi-agent simulation of medical activities in the affected areas with use of damage information, medical demands, and resources in the medical facilities to fid optimal allocation of medical response.

(4) Developing a digital twin computing platform to support disaster medical response activities and find optimal allocation of disaster medical services through what-if analysis of multi-agent simulation.

How to cite: Koshimura, S. and Mas, E.: Digital twin computing for enhancing resilience of disaster response system, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11756, https://doi.org/10.5194/egusphere-egu23-11756, 2023.

Daman Landslide had blocked one of the three cross-island roads in Taiwan, and a road section has been under control since last October. During the period, more than thousands of small-scale post-failures occurred whose irregular patterns affected the safety of engineering workers for slope protection construction and road users. Therefore, we installed one time-lapse camera and two geophones at the crown and closed to the toe of the Daman landslide, respectively to train a classification model to offer in-situ alarm. According to time-lapse photos, those post failures can be categorized into two types. One is rock/debris moving and stopping above the upper slope or road, named type I, and the other is the rock/debris going through the road to download slope, named type II. Type I was almost recorded by the crown station, and type II was shown by both stations with different arrival times and the toe station’ high-frequency signals gradually rising (up to 100 Hz). Those distinct features were exhibited by spectrograms. To keep characteristics simultaneously, we merge two stations’ spectrograms as one to indicate different types of post-failures. However, frequent earthquakes affect the performance of the landslide’s discrimination, which should be involved in the classification model. A total of three labels, type I, type II, and earthquake, contained more than 15,000 images of spectrogram, have been used for deep neural network (DNN) to be a two-station-based automatic classifier. Further, user-defined parameters for the specific frequency band within fixed time span windows, including a sum of power spectrogram density, the arrival time of peak amplitude, cross-correlation coefficient, and signal-to-noise ratio, have been utilized for the decision tree algorithm. Both model results benefit the automatic classifier for post-failure alarms and can readily extend to monitor other landslides with frequent post-failures by transfer learning.

How to cite: Chang, J.-M., Chao, W.-A., and Huang, W.-K.: Classification Seismic Spectrograms from Deep Neural Network: Application to Alarm System of Post-failure Landslides, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12240, https://doi.org/10.5194/egusphere-egu23-12240, 2023.

Humans significantly control the natural environment and natural processes. Global fire ignitions are a prime example of how human actions change the frequency of occurrence of otherwise rare events like wildfires. However, human controls on fire ignition are insufficiently characterised by global fire models because impacts are often indirect, complex, and collinear. Hence, modelling fire activity while considering the complex relationships amongst the input variables and their effect on global ignitions is crucial to developing fire models reflecting the real world. 

This presentation leverages causal inference and machine learning frameworks applied to global datasets of fire ignitions from Earth observations and potential drivers to uncover anthropogenic pathways on fire ignition. Potential fire controls include human predictors from Earth observations and statistical data combined with variables traditionally associated with fire activity, like weather, and vegetation abundance and state, derived from earth observations and models.

Our research models causal relationships between fire control variables and global ignitions using Directed Acyclic Graphs(DAGs). Here, every edge between variables symbolises a relation between them; the edge weight indicates the strength of the relationship, and the orientation of the edge between the variables signifies the cause-and-effect relationship between the variables. However, defining a fire ignition distribution using DAGs is challenging owing to the large combinatorial sample space and acyclicity constraint. We use Bayesian structure learning to make these approximations and infer the extent of human intervention when combined with climate variables and vegetation properties. Our research demonstrates the need for causal modelling and the inclusion of anthropogenic factors in global fire modelling.

How to cite: Pande, N. and Dorigo, W.: Investigating causal effects of anthropogenic factors on global fire modeling, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12716, https://doi.org/10.5194/egusphere-egu23-12716, 2023.

Flood early warning systems are vital for preventing flood damages and for reducing disaster risks. Such systems are particularly important for forecasting compound events where multiple, often dependent flood drivers co-occur and interact. In this research an early warning system for prediction of coastal-fluvial floods is developed to provide a robust, cost-effective and time-efficient framework for management of flood risks and impacts. This three-step method combines a cascade of three linked models: (1) statistical model that determines probabilities of multiple-driver flood events, (2) hydrodynamic model forced by outputs from the statistical model, and finally (3) machine learning (ML) model that uses hydrodynamic outputs from various probability flood events to train the ML algorithm in order to predict the spatially and temporarily variable inundation patterns resulting from a combination of coastal and fluvial flood drivers occurring simultaneously.

The method has been utilized for the case of Cork City, located in the south-west of Ireland, which has a long history of fluvial-coastal flooding. The Lee  River channelling through the city centre may generate a substantial flood when the downstream river flow draining to the estuary coincides with the sea water propagating upstream on a flood tide. For this hydrological domain the statistical model employs the univariate extreme values analysis and copula functions to calculate joint probabilities of river discharges and sea water levels (astronomical tides and surge residuals) occurring simultaneously. The return levels for these two components along a return level curve produced by the copula function are used to generate synthetic timeseries, which serve as water level boundary conditions for a hydrodynamic flood model. The multi-scale nested flood model (MSN_Flood) was configured for Cork City at 2m resolution to simulate an unsteady, non-uniform flow in the Lee  River and a flood wave propagation over urban floodplains. The ensemble hydrodynamic model outputs are ultimately used to train and test a range machine learning models for prediction of flood extents and water depths. In total, 23 machine learning algorithms including: Artificial Neural Network, Decision Tree, Gaussian Process Regression, Linear Regression, Radial Basis Function, Support Vector Machine, and Support Vector Regression were employed to confirm that the ML algorithm can be used successfully to predict the flood inundation depths over urban floodplains for a given set of compound flood drivers. Here, the limited flood conditioning factors taken into account to analyse floods are the upstream flood hydrographs and downstream sea water level timeseries. To evaluate model performance, different statistical skill scores were computed. Results indicated that in most pixels, the Gaussian Process Regression model performs better than the other models.

The main contribution of this research is to demonstrate the ML models can be used in early warning systems for flood prediction and to give insight into the most suitable models in terms of robustness, accuracy, effectiveness, and speed. The findings demonstrate that ML models do help in flood water propagation mapping and assessment of flood risk under various compound flood scenarios.

How to cite: Olbert, A. I., Moradian, S., and Uddin, G.: Machine learning modelling of compound flood events, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13083, https://doi.org/10.5194/egusphere-egu23-13083, 2023.

Climate change is one of the most pressing challenges to humankind today. The number and severity of wildfires are increasing in many parts of the world, with record-breaking temperatures, prolonged heat waves, and droughts. We can minimize the risks and consequences of these natural disasters by providing accurate and timely wildfire progression predictions through fire spread modeling. Knowing the direction and rate of spread of wildfires over the next hours can help deploy firefighting resources more efficiently and warn nearby populations hours in advance to allow safe evacuation.
Physics-based spread models have proven their applicability on a regional scale but often require detailed spatial input data. Additionally, rendering them in real-time scenarios can be slow and therefore inhibit fast output generation. Deep learning-based models have shown success in specific fire spread scenarios in recent years. But they are limited by their transferability to other regions, explainability, and longer training time. Accurate active fire data products and a fast data pipeline are additional essential requirements of a wildfire spread early-warning system.
In this study, physical models are compared to a deep learning-based CNN approach in terms of computational speed, area accuracy, and spread direction. We use a dataset of the 30 largest wildfires in the US in the year 2021 to evaluate the performance of the model’s predictions.
This work focuses in particular on the optimization of a cloud-based fire spread modeling data pipeline for near-real-time fire progression over the next  2 to 24 hours. We describe our data pipeline, including the collection and pre-processing of ignition points derived from remote sensing-based active fire detections. Furthermore, we use data from SRTM-1 as topography, ESA Land Cover and Corine Land Cover for fuel composition, and ERA-5 Reanalysis products for weather data inputs. The application of the physics-based models is derived from the open-source library ForeFire, to create and execute physical wildfire spread models from single fire ignition points as well as fire fronts. The predictions of the ForeFire model serve as a benchmark for the evaluation of the performance of our Convolutional Neural Network. The CNN forecasts the fire outline based on a spatiotemporal U-Net architecture. 
The scaling of the algorithms to a global setting is enabled by the Leibniz Supercomputing Centre. It enables large-scale cloud-based machine learning to provide a time-sensitive solution for operational fire spread modeling in emergency management based on real-time remote sensing information. 

How to cite: Bauer, T., Miller, J., Gottfriedsen, J., Mollière, C., Durillo Barrionuevo, J., and Hammer, N.: ML-based fire spread model and data pipeline optimization, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14126, https://doi.org/10.5194/egusphere-egu23-14126, 2023.

Multiple landslide events occur often across the world which have the potential to cause significant harm to both human life and property. Although a substantial amount of research has been conducted to address the mapping of landslides using Earth Observation (EO) data, several gaps and uncertainties remain when developing models to be operational at the global scale. To address this issue, we present the HR-GLDD, a high-resolution (HR) dataset for landslide mapping composed of landslide instances from ten different physiographical regions globally: South and South-East Asia, East Asia, South America, and Central America. The dataset contains five rainfall triggered and five earthquake-triggered multiple landslide events that occurred in varying geomorphological and topographical regions. HR-GLDD is one of the first datasets for landslide detection generated by high-resolution satellite imagery which can be useful for applications in artificial intelligence for landslide segmentation and detection studies. Five state-of-the-art deep learning models were used to test the transferability and robustness of the HR-GLDD. Moreover, two recent landslide events were used for testing the performance and usability of the dataset to comment on the detection of newly occurring significant landslide events. The deep learning models showed similar results for testing the HR-GLDD in individual test sites thereby indicating the robustness of the dataset for such purposes. The HR-GLDD can be accessed open access and it has the potential to calibrate and develop models to produce reliable inventories using high-resolution satellite imagery after the occurrence of new significant landslide events. The HR-GLDD will be updated regularly by integrating data from new landslide events.

How to cite: Catani, F., Meena, S. R., Nava, L., Bhuyan, K., Puliero, S., Pedrosa Soares, L., Dias, H. C., and Floris, M.: A globally distributed dataset using generalized DL for rapid landslide mapping on HR satellite imagery, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15711, https://doi.org/10.5194/egusphere-egu23-15711, 2023.

Fluvial and flash floods can have devastating effects if they occur without warning. In Denmark, management of flood risk and performing preventative emergency service actions has been the sole responsibility of local municipalities. However, motivated by the disastrous 2021 floods in Central Europe, the Danish government has recently appointed the Danish Meteorological Institute (DMI) as the national authority for flood warnings in Denmark, and DMI is in the process of building capacity to fulfill this role.

One of the most cost-effective ways to mitigate flood damages is a well-functioning early warning system. Flood warning systems can rely on various methods ranging from human interpretation of meteorological and hydrological data to advanced hydrological modelling. The aim of this study is to generate short-range streamflow predictions in Danish river systems with lead times of 4-12 hours. To do so, we train and test models with hourly data on 172 catchments.

Machine learning (ML) models have in many cases been shown to outperform traditional hydrological models and offer efficient ways to learn patterns in historical data. Here, we investigate streamflow predictions with LightGBM, which is a gradient boosting framework that employs tree-based ML algorithms and is developed and maintained by Microsoft (Ke et al., 2017). The main argument for choosing a tree-based algorithm is its inherent ability to represent rapid dynamics often observed during flash floods. The main advantages of LightGBM over other tree-based algorithms are efficiency in training and lower memory consumption. We benchmark LightGBM’s performance against persistence, linear regression and various LSTM setups from the Neural Hydrology library (Kratzert et al., 2022).

We evaluate the algorithm trained using different input features. This analysis include model explainability, such as SHAP, and the results indicate that simply using lagged real-time observations of streamflow together with precipitation leads to the best performing and most parsimonious models. The results show that the LightGBM setup outperforms the benchmarks and is able to generate predictions with high Klinge-Gupta Efficiency scores > 0.9 in most catchments. Compared to the persistence benchmark it especially shows strong improvements on peak timing errors.

How to cite: Martinsen, G., Sweeney, Y., Pedersen, J. W., Alexandru, R., Capape, S., Harris, C., Butts, M., and Diaz, M.: Danish national early warning system for flash floods based on a gradient boosting machine learning framework, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16626, https://doi.org/10.5194/egusphere-egu23-16626, 2023.

Increases in extreme rainfall intensities as a result of climate change pose a great risk due to the possible increases in pluvial flooding. But evidence is emerging that the observed increases in extreme rainfall are not resulting in universal increases in flooding. Here, we begin by presenting historical evidence for changes in extreme rainfalls and floods discussing the underlying mechanisms for the changes, before examining the implications of climate change projections on engineering design.

Extreme rainfall is intensifying universally across the globe with more extreme events experiencing larger degrees of intensification. Simultaneously, and somewhat paradoxically, the magnitude of frequent floods (those expected to occur on average once per year) are in general decreasing, particularly in the tropical and arid regions of the world. We suggest this is likely due to the dominance of drying antecedent soil moisture conditions and shorter storm durations at higher temperatures offsetting any increases in rainfall intensity. However, for rare magnitude floods (those expected, on average, to occur less than once every twenty years) the increase in rainfall appears to outweigh any decrease in soil moisture or change in the temporal pattern of the storm.

Climate model projections, downscaled through a continental scale water balance model and locally calibrated rainfall-runoff models, show that future projections of flood responses follow historical trends – with the rarer the flood, the more likely it is to be increasing. To deepen our understanding, we focus our analysis on event runoff coefficients as an indicator of future runoff changes. Across Australia we find runoff coefficients are projected to decrease, that is, reduced runoff resulting from the same amount of rainfall. These results indicate drier conditions and a compounding of the reduced average rainfall and drier conditions already being experiences in many arid parts of the world.

With these historical changes and projections in mind we conclude with some insights and implications on how best to incorporate the additional uncertainty due to climate change when estimating floods for planning and design purposes. As floods constitute a large portion of the inflows into reservoirs, we suggest that future water resources planning will need to account for reduced runoff yields. To assess the potential impacts of future climate change for planning and design purposes we need to consider how changes to rainfall intensity vary with both storm duration and storm rarity, as well as how antecedent conditions influence the proportion of rainfall that appears as runoff. There remains significant work in adapting our current flood guidance to reflect these historical and projected changes.

How to cite: Wasko, C., Ho, M., Nathan, R., Sharma, A., Dykman, C., and Vogel, E.: Implications for engineering design of shorter more extreme rainfalls and increased flood variability, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1224, https://doi.org/10.5194/egusphere-egu23-1224, 2023.

Disaster risk determines the potential loss of life, injury, or destroyed or damaged assets which could occur to a system, society or a community in a specific period of time, determined probabilistically as a function of hazard, exposure, vulnerability and capacity. This paper focuses on the exposure elements, that expresses people, infrastructure, housing, production capacities and other tangible human assets located in hazard-prone areas (UNDRR, 2017).  In performing risk analyses, an accurate exposure model should be constructed and specified according to the purpose and spatial scale of the assessment.

The scope of the present work is the flood displacement risk assessment for two small island developing states in the Pacific Ocean, Fiji and Vanuatu, where a new methodology is proposed, that considers different but intrinsically linked components in assessing the contribution of disasters to displacement. In this assessment, three main elements are supposed to trigger (or at least contribute to cause) flood displacements: the loss of housing, the loss of livelihoods or the loss of access to basic services. This implies that, besides the classical vulnerability characterization of a asset based on occupancy (residential, commercial, industrial, etc.) and structural elements (number of stories, basement, etc.), the exposure model must also consider a spatial representation of the population relying on the specific function of that asset: residential population in case of residential building; population working in that building in case of commercial, industrial, or service buildings; population working in crop or grazing areas in case of agricultural field; number of students in case of school.

In this context, a procedure for avoiding potential double counting was also implemented. It means that, to evaluate the ratio of population that could suffer impacts due to floods on both livelihoods and housing, each worker must be associated to his/her home with his/her workplace.

Regarding the spatial scale, the small size of the countries allows for the definition of a high-resolution exposure model, that entails a characterization at building Level.

The construction of the exposure model is articulated in three main steps: 1) analysis and integration of different sources of employment and residential data (from global to local information); 2) physical characterization of assets at building scale, using building footprints from the Open Street Map layer and attributes from existing exposure models, such as Pacific Catastrophe Risk Assessment and Financing Initiative (PCRAFI) project that lasted from 2012 to 2017 and Global Earthquake Model (GEM) within the project Global Exposure Map (v2018.1); 3) the procedure to avoid double counting, which associates each worker to his/her home with his/her workplace, following the criterion of minimum geometric distance between workplace and residence.

The exposure model is then used in a probabilistic risk assessment, where different flood scenarios and related damage scenarios are computed at building scale. Physical damage above a certain threshold is considered to cause the unavailability of asset function (residence, workplace), thus triggering the displacement of people relying on that function.

How to cite: Ottonelli, D., Ponserre, S., Rossi, L., Rudari, R., and Trasforini, E.: High resolution exposure model for a flood displacement risk assessment, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5418, https://doi.org/10.5194/egusphere-egu23-5418, 2023.

Floods have triggered about 166 million displacements globally since 2008, according to the Internal Displacement Monitoring Center (IDMC). Since 2008, most of the displacements triggered by floods have been localized in Asia and the Pacific and with an overall estimate of 129 million displacements. Small Island Developing States (SIDS) states bear the greatest displacement risk relative to their population size. Climate change combined with vulnerability of exposed infrastructure, and housing poses an existential threat for some Pacific islands that could see their populations move not only internally but also across borders. These magnitudes of forced movement highlight the importance of the phenomenon. In this context, we present a first attempt to estimate present and future riverine flood displacement risk at the national and sub-national level for two countries in the Pacific Ocean: Fiji and Vanuatu.

This work proposes a new methodology that provides a more comprehensive assessment of vulnerability in the context of disaster displacement risk and recognizes that people’s vulnerability depends on several physical and social factors. Such elements, however, are not yet included in standard risk models because difficultly quantifiable. While quantitative approaches to disaster displacement risk assessment generally consider the likelihood of housing rendered unhabitable as a proxy for displacement, this new methodology expands this concept by taking into account different elements that may trigger displacements or may increase the susceptibility to forced movement: 1) impact on houses; 2) impact on livelihoods; 3) impact on critical facilities and services.

A probabilistic risk assessment was performed at building scale in present and future climate conditions: under current climate conditions (1979-2016); under medium-term projected climate conditions (2016 - 2060); under long-term projected climate conditions (2061 – 2100). As results, displacement risk information - expressed in annual average displacement (AAD) and probable maximum displacement (PMD) - were calculated at national and subnational (NUTS2) scales, allowing for a geographic and quantitative comparison both within and between countries. The computation performed at building scale also allowed for result aggregation by sectors.

The outputs of the probabilistic model  show an important role of climate change in determining future likelihood to displacement due to riverine floods in the area. Flood displacement risk is likely to double by 2060 in both countries, and under the pessimistic long-term scenarios AAD is expected to triple in Fiji and quadruple in Vanuatu. These analyses are an important step in risk awareness processes and key to pushing for risk reduction, adaptation, and management mechanisms to be put in place.

How to cite: Trasforini, E., Rossi, L., Ponserre, S., Campo, L., Libertino, A., Ottonelli, D., and Rudari, R.: From vulnerability to vulnerabilities for a probabilistic flood displacement risk model: the case study of Fiji and Vanuatu., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5563, https://doi.org/10.5194/egusphere-egu23-5563, 2023.

Natural disasters are one of the main causes of worry for the majority of nations because they severely harm the global economy. One of the natural disasters that occur on a global scale that seriously damages infrastructure and claims thousands of lives is flooding. Due to its geographical location, India is one of the high-risk nations that is negatively impacted by floods every year. It ranks in the top 20% of countries worldwide for the number of flood-related fatalities. A natural tragedy cannot be prevented. However, if some preventative actions were taken in advance, a sizable portion of the potential damage may be prevented. Professionals and authorities need accurate figures regarding flood depth, amount of flow, scale, and distinct datasets in order to reduce and manage the effects of such catastrophes. The management of flood risk is heavily dependent on flood modelling. One of the many software tools that assists in computing the discharge, depth, magnitude, and statistics of rivers located in high-risk flood zones is HEC-RAS (Hydrologic Engineering Centre-River Analysis System).This study employed the Purna River's 1D hydrodynamic floods modelling (50 and 100 years out) on HEC-RAS and it has been found that the great portion of populated area will be affected in future. The goal of this study is to assess the prediction power and carry out a sensitivity analysis to identify the sensitive zones. This research project would enable different flood modelling and risk zone delineation for diverse flood-affected areas in India and around the world. In places affected by flooding, these technologies can also be used to create emergency response protocols.

How to cite: Khan, M. U. S., Abdullah, M., and Khan, A. A.: Flood Modelling and Simulation using HEC-RAS, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5564, https://doi.org/10.5194/egusphere-egu23-5564, 2023.

Biodiversity loss, land degradation, and climate change are acknowledged environmental challenges faced by humanity. Human activities including land-use changes are key stressors for biodiversity, thus, future projections of biodiversity impacts need to include both climate change and land-use change. While a lot of studies focused on mapping and projecting the vulnerability of multiple species based on different climate mitigation scenarios or warming levels, land-use trajectories are often not included in these projections. Recent work made first steps to address these deficiencies. For example, Hof et al. (2018) evaluated potential future impacts of climate and land-use changes on global species richness of terrestrial vertebrates under a low and high emission scenario. However, they used the same land-use change assumptions for both emission scenarios. In this study, we aim to fill the described research gap by combining future climate scenarios and a matrix of land-use projections derived from integrated assessment modeling (IAM) to estimate the fractional land-use patterns, underlying land-use transitions, and key agricultural management information, to assess the impact of climate change on biodiversity and quantify the additional impact of land-use change.

To this end, we use the global simulations with a species distribution model from the Hof et al. (2018) study forced by four GCMs and both RCP2.6 and RCP6.0 climate scenarios following the ISIMIP2b simulation protocol and apply a land-use filter on the species occurrence probabilities to determine the implications for the world’s amphibians, mammals and reptiles at a 0.5° resolution. The land use data used to include future projections of land-use change is the Land Use Harmonization dataset v2 (LUH2). LUH2 reconstructs and projects changes in land use among 12 categories. To match the species’ habitat preferences, data from IUCN Habitat and Classification Scheme for each species is mapped onto the 12 land-use types represented in the LUH2 dataset according to the conversion table from Carlson et al. (2022). The land-use data is then used to refine the climatic envelope and filter out regions where species cannot persist.

This approach allows to quantify the change of the proportion of affected species distributions between different climate and land-use scenarios and combinations of both. In addition, it provides quantitative information on the impact of future climate change on biodiversity accounting for the combination of land-use change projections and climate-driven species distribution models.

Key Reference:

Hof, C., Voskamp, A., Biber, M. F., Böhning-Gaese, K., Engelhardt, E. K., Niamir, A., Willis, S. G., & Hickler, T. 2018: Bioenergy cropland expansion may offset positive effects of climate change mitigation for global vertebrate diversity. Proceedings of the National Academy of Sciences of the United States of America, 115(52), 13294–13299.

How to cite: Hari, C., Vanderkelen, I., Fischer, M., and Davin, É.: Combining future projections of land-use and climate change to assess their impact on biodiversity, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7110, https://doi.org/10.5194/egusphere-egu23-7110, 2023.

Climate change is driving increased urgency for better quantification of climate hazards and their impacts for stakeholders across multiple economic sectors. Flooding has been highlighted as one of the most significant climates risk to UK economic infrastructure, with costs expected to increase with climate-driven changes to rainfall, such as increased intensity of summer storms. To accelerate climate change adaptation and enable economic resilience to climate change impacts, close collaboration is needed between climate scientists, impact modellers, and stakeholders, and technology advances can support this by enabling and streamlining the process of developing and deploying climate impact modelling workflows to translate complex datasets and scientific models into actionable information.

In this presentation, we describe the application of such a technology for the case of pluvial flooding, undertaken as part of the IBM Research and Science and Technology Facilities Research Council partnership, the Hartree National Centre for Digital Innovation (HNCDI), a 5-year programme established to develop and apply new technology to key economic challenges in the UK. Here, we model pluvial flood hazard for a case study region in northeastern England, using a 2-d physical simulation model of flood inundation, driven by open-access geospatial and climate datasets. Flood hazard maps are translated to impact using open asset location data and damage functions.

We consider the sensitivity and scalability (in terms of computational cost) of the hazard and impact predictions to multiple factors, including (1) DEM/DSM representation of land surface (2) soil and land use parameterisation, and (3) model spatial resolution. We also contrast the use of drivers in the form of extreme rainfall scenarios created using a traditional design storm approach, and ensembles of synthetic storms from a stochastic weather generator, both derived from hourly 1km gridded rainfall observations. Finally, we reflect on key gaps to be addressed in the models, data and technology to meaningfully inform climate adaptation across industry sectors.

How to cite: Jones, A., Taylor, A., Butt, J., Edwards, B., Diaz, J. L. G., and Avegliano, P. B.: End-to-end modelling of flood risk and impact for climate change resilience, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7346, https://doi.org/10.5194/egusphere-egu23-7346, 2023.

Climate change is causing increased risks linked to extreme weather events. In order to develop effective adaptation strategies and policies, there is an urgent need for methodologies able to assess how the socio-economic risks associated with extreme climate-related events will change in the coming decades especially at local scale. The development of these methodologies require the expertise from many different scientific disciplines, including: modelling of global and local climatic phenomena, assessment of the intensity and probability of extreme events, representation of their impacts on the society and quantification of the associated risk.

In this work we propose a methodological chain linking the risk of extreme events in a changing climate with both direct and indirect impacts on the socio-economic sector from regional to local scale. The proposed chain integrates the knowledge of three scientific fields: climatologists, engineers and macro-economist. Here, we present agreements and differences between communities (e.g., aim, terminology, methodology, etc.), and evaluate advantages and constrains of the combined used of high-resolution regional climate models, engineering risk assessment models and economic input-output models compared to the state of the art in this field.

To illustrate the advantages of the proposed methodology and its practical feasibility, we present preliminary results from an applied pilot study in the Italian context.

How to cite: Arosio, M., Caiani, A., Fosser, G., and Di Noia, J.: A transdisciplinary chain to assess the risk of direct and indirect impacts linked to extreme climate events from regional to local scale, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7522, https://doi.org/10.5194/egusphere-egu23-7522, 2023.

Aviation’s contribution to anthropogenic global warming is estimated to be between 3 – 5% [1]. This assessment comprises two contributions: the well understood atmospheric impact of carbon dioxide (CO₂) and the more uncertain non-CO₂ effects. The latter pertain to persistent contrails and pollutants like nitrogen oxides (NO_x), water vapor (H₂O), sulfur oxides (SO_x) and soot particles. NO_x emissions are involved in non-linear processes that result in the short-term production of ozone (O₃) and longer-term destruction of methane (CH₄), stratospheric water vapor (SWV), and primary mode ozone (PMO). The aviation-attributable impacts arising from this short-term increase in O₃ can vary by more than a factor of 1.5 depending on the selected modelling approach. This O₃ increase is associated with the second largest warming effect across aviation’s main climate forcers [1]. We therefore quantify this figure using three modelling approaches (an Eulerian and a Lagrangian tagging scheme as well as a perturbation approach) at three potential aircraft cruise altitudes (200, 250 and 300 hPa) at which NO_x pulse emissions are introduced in the Americas, Africa, Eurasia and Australasia. In general, the tagging method computes the contribution by an emission source to the concentration of a chemical species while a perturbation approach consists in calculating the total impact of an emission to the concentration of a species by means of subtracting two simulations: one with all emissions and a second without the specific source’s emissions. We compare results from Eulerian and Lagrangian simulations using the same climate-chemistry code: the ECHAM5/MESSy Atmospheric Chemistry (EMAC) model. With the Eulerian setup, we are able to capture non-linear processes and feedback effects, but not track the transport of emitted species in detail. The Lagrangian setup [2], on the other hand, allows for the accompaniment of thousands of air parcel trajectories, but at the cost of assuming a simplified linear chemistry mechanism. We find that the Lagrangian tagging approach provides the largest estimates for O₃ production and radiative forcing (RF), followed by the Eulerian tagging scheme and lastly by the perturbation method. We therefore investigate the appropriateness of each of these in quantifying aviation’s total and marginal climate effects by addressing the following research questions: 1) By how much are the estimates for the short-term NO_x-induced O₃ perturbation and consequent RF varying across the three modelling approaches and why? 2) How does this RF vary with emission altitude within the upper Troposphere/lower Stratosphere (UTLS)?

[1] Lee, D.S., Fahey, D.W., Skowron, A., Allen, M.R., Burkhardt, U., Chen, Q., Doherty, S.J., Freeman, S., Forster, P.M., Fuglestvedt, J., Gettelman, A., De León, R.R., Lim, L.L., Lund, M.T., Millar, R.J., Owen, B., Penner, J.E., Pitari, G., Prather, M.J., Sausen, R., and Wilcox, L.J.: The contribution of global aviation to anthropogenic climate forcing for 2000 to 2018, Atmos. Environ., 244, 117834, https://doi.org/10.1016/j.atmosenv.2020.117834, 2021.

[2] Maruhashi, J., Grewe, V., Frömming, C., Jöckel, P., and Dedoussi, I. C.: Transport patterns of global aviation NOx and their short-term O3 radiative forcing – a machine learning approach, Atmos. Chem. Phys., 22, 14253–14282, https://doi.org/10.5194/acp-22-14253-2022, 2022.

How to cite: Maruhashi, J., Mertens, M., Grewe, V., and Dedoussi, I.: Assessing the appropriateness of different climate modelling approaches for the estimation of aviation NOx climate effects, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9246, https://doi.org/10.5194/egusphere-egu23-9246, 2023.

Sub-daily extreme precipitation can generate fast hydro-geomorphic hazards such as flash floods and debris flows, which cause fatalities and damages especially in mountainous regions. Reliable projections of extreme future precipitation is fundamental for risk management and adaptation strategies. Convection-permitting climate models (CPMs) esplicitely resolve large convective systems and represent local processes, especially sub-daily extreme precipitation, more realistically than coarser resolution models, thus leading to higher confidence in their projections. Given their high computation cost, however, the available CPM simulations cover relatively short time periods (10–20 years), too short for deriving precipitation frequency analyses with conventional extreme value methods based on annual maxima or threshold exceedances.

In this work, we evaluate the potential of a non-asymptotic approach based on “ordinary” events, the so-called Simplified Metastatistical Extreme Value (SMEV), to provide information on the future change of short-duration precipitation extremes. We focus on a complex-orography region in the Eastern Italian Alps, where significant changes in sub-daily annual maxima have been already observed. The study is based on COSMO-crCLIM model simulations at 2.2 km resolution under the RCP8.5 scenario and uses three 10-year time periods: historical 1996-2005 (the control period), near-future 2041-2050 and far future 2090-2099. We estimate extreme precipitation for durations ranging from 1 h to 24 h and assess the projected changes with respect to the control period. Specifically, we analyze annual maxima, return levels up to 50 years, and the parameters of the statistical model. A bootstrap procedure is used to evaluate the uncertainty of the estimates, and a permutation test is applied to assess the statistical significance of the projected changes. We compare our results with a modified Generalized Extreme Value (GEV) approach, recently applied for the study of extremes in CPM future time periods.

We found that annual maxima and higher return levels exhibit a general increase in the future especially for the far future and the shorter event durations. On average, the magnitude of the far future change decreases with the precipitation temporal scale. The changes show an interesting spatial organization that can be associated with the orography of the region: significant future increases are mostly located at high elevations, while lowlands and coastal zones show no clear pattern.

This work shows that SMEV reduces the uncertainty in the estimates of higher return levels compared to GEV and can thus provide improved estimates of their future changes from short CPM runs. These findings advance our knowledge about the projected changes in extreme precipitation and their spatial distribution at the different time scales. They can thus help improving risk management and adaptation strategies.

How to cite: Dallan, E., Fosser, G., Schaer, C., Roghani, B., Canale, A., Marani, M., Borga, M., and Marra, F.: Future changes in sub-daily extreme precipitation over a complex-orography area from a convection-permitting climate model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9338, https://doi.org/10.5194/egusphere-egu23-9338, 2023.

We present the results of an NHESS (Natural Hazards and Earth System Sciences) 20th anniversary survey, in which 350 natural hazard community members responded to two questions: (Q1) “what are the top three scientific challenges you believe are currently facing our understanding of natural hazards” and (Q2) “what three broad step changes should or could be done by the natural hazard community to address natural hazards in achieving the Sustainable Development Goals”? We have analysed the data quantitatively and qualitatively. According to the 350 respondents, the most significant challenges (Q1) are the following (within brackets % of 350 respondents who identified a given theme): (i) shortcomings in the knowledge of risk and risk components (64 %), (ii) deficiencies of hazard and risk reduction approaches (37 %), (iii) influence of global change, especially climate change (35 %), (iv) integration of social factors (18%), (v) inadequate translation of science to policy and practice (17 %), and (vi) lack of interdisciplinary approaches (6 %). In order for the natural hazard community to support the implementation of the Sustainable Development Goals (Q2), respondents called for (i) enhanced stakeholder engagement, communication and knowledge transfer (39 %), (ii) increased management and reduction of disaster risks (34 %), (iii) enhanced interdisciplinary research and its translation to policy and practice (29 %), (iv) a better understanding of natural hazards (23 %), (v) better data, enhanced access to data and data sharing (9 %), and (vi) increased attention to developing countries (6 %). We note that while the most common knowledge gaps are felt to be around components of knowledge about risk drivers, the step changes that the community felt were necessary related more to issues of wider stakeholder engagement, increased risk management and interdisciplinary working.

How to cite: Malamud, B. D., Šakić Trogrlić, R., and Donovan, A.: Three Hundred and Fifty Views on what the Natural Hazard Community should do to Support the Implementation of the SDGs, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9379, https://doi.org/10.5194/egusphere-egu23-9379, 2023.

Can total annual streamflow in any given year be largely characterised by a relatively small number of high flow events? A comprehensive assessment of this is of high value as there is evidence to suggest that as flood events increase in rarity a more consistent response between streamflow extremes and temperature increases can be established — providing greater reliability in projections of rare events. We propose here a novel methodology to characterise streamflow regimes in the context of total annual streamflow for water supply. Using the Australian Bureau of Meteorology’s Hydrologic Reference Station database, we developed annual event flow distributions that standardise the relationship between total annual streamflow and event flows. It was found that the annual event flow distributions are primarily a function of local climate and catchment size and were largely insensitive to interannual variability represented by the El Nino Southern Oscillation Index, mean annual temperature, or total annual rainfall volume. Statistically significant trends were found in the timeseries of annual event flow distribution values, signalling a move to a less even distribution in the southern latitudes and a more even distribution in the northern latitudes. Our results show that total annual streamflows can be characterised by a small number of high flow events. This suggests that for Australia’s most critical surface drinking water supply catchments the streamflow yields can be represented by changes in a few, high flow events, independent of interannual variability. As these relationships are non-stationary, they may provide a basis for understanding changes in water supply into the future.

How to cite: Dykman, C., Sharma, A., Wasko, C., and Nathan, R.: Can annual streamflow volumes be characterised by flood events alone?, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10210, https://doi.org/10.5194/egusphere-egu23-10210, 2023.

In the past 5 years large efforts have been made to improve our understanding of scale-interactions in the Earth system, and to better resolve atmospheric and oceanic meso-scale processes and their response to greenhouse warming. Here, we provide an overview of the technical and scientific achievements of a new collaboration between the IBS Center for Climate Physics (South Korea) and the Alfred Wegener Institute for Polar and Marine Research (Germany) to simulate the climate system at km-scale resolution using the AWI Climate Model, version 3 (AWI-CM3). AWI-CM3 is based on the OpenIFS-FESOM2 coupled model and we conducted several control simulations and transient greenhouse warming runs in a medium-resolution (MR) configuration (31 km in the OpenIFS and 5~27 km in the FESOM2, ‘Tco319-DART’). These simulations will be used in future as initial conditions for shorter coupled storm-resolving (SR) simulations with target resolutions of 9 km and 4 km (Tco1279 and Tco2559). Our presentation focuses on the performance of the MR configuration (with a throughput of about 7 simulation years per day on 350 nodes) and its representation of the mean climate, climate variability such as the El Niño-Southern Oscillation, tropical cyclone statistics. We will also present preliminary estimates of the expected scaling behavior of the AWI-CM3 SR configuration on different multi-Petaflop supercomputing systems.

How to cite: Lee, S.-S., Timmermann, A., Jung, T., Semmler, T., Chu, J.-E., Streffing, J., and Raavi, P. H.: Towards global km-scale greenhouse warming simulations with the AWI-CM3, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10715, https://doi.org/10.5194/egusphere-egu23-10715, 2023.

Flood-related damage has increased dramatically in recent decades with direct and indirect economic impacts accounting for a large share of gross national products. Therefore, there is an urgent need to acquire more quantitative knowledge about flood damage to mitigate economic losses and reduce exposure to flood risk.

Firms are especially affected in case of flood. Still, flood damage assessment to businesses is hindered by the paucity of available data to characterize the enterprises, the lack of high-quality damage data to derive new models or validate existing ones, and the high variability of activity types which hampers generalization. This study contributes at improving knowledge about types and extent of damage of flood events on economic activities through the analysis of empirical data, focusing on direct damage and with specific reference to the Italian context.

In detail, the investigated dataset is composed by around a thousand of observed damage records collected after four flood events in Italy, along with additional information on the dimension (i.e., surface and number of employees) and the typology of the affected firms (i.e., NACE category) as well as on local water depth levels. Damage data are further classified in damage to the building structure, the stock, and the equipment.

Several econometric models have been implemented to better understand the links among the damage, the characteristics of the economic activities and the water depth. Since the heterogeneity of the affected firms is very high, in terms of surface, water depth levels, and number of employees and this might have had influence on the firm’s damage reporting, data has been analyzed with Heckman's selection bias model.

Obtained results show the absence of a constant return scale relationship, therefore, the total damage increases less than proportionally to the firm’s surface; the water depth plays an important role to explain the damage to the stock that results the more vulnerable asset.  Information on the NACE category made it possible to quantify the differences in damage by economic sector. The results reveal as the most vulnerable sectors for building structure, stock and equipment, respectively, human health, commercial, and manufacture. The accuracy of the prediction models represented by adjusted R² varies between 0.25, 0.36 depending on the damage component.

Despite characterized by significant uncertainty, obtained results supply a first model for the prediction of flood damage to firms for the Italian context, in the support of more effective risk mitigation actions. In fact, the model identifies the more vulnerable elements within the business sectors orienting modelers and decision-makers choices.

Acknowledgements:

Authors acknowledge with gratitude: Francesca Carisi, Alessio Domeneghetti and Armando Brath (from University of Bologna), Giovanni Menduni, Giulia Pesaro and Guido Minucci (from Politecnico di Milano), Simone Sterlacchini and Marco Zazzeri (from the Italian National Research Council) for their collaboration in collecting the observed damage records analysed in the research. A special thanks to Marta Galliani (from Politecnico di Milano) for providing the refined dataset used in this study.

How to cite: Ballocci, M., Molinari, D., Ballio, F., and Marin, G.: Econometric modelling for the  estimation of direct flood damage to enterprises: a local-scale approach from post-event records in Italy, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11320, https://doi.org/10.5194/egusphere-egu23-11320, 2023.

In this paper we combine traditional econometric time series techniques and machine learning algorithms to construct skillful monthly maize price prediction models for four southern African countries – namely, Malawi, Mozambique, Zambia, and Zimbabwe. Theoretical models of price transmission commonly assume that shocks are transmitted from an external market (typically modeled as the world market) to the largest domestic city or port within a country and then, depending on the degree of market integration within the country, these shocks are transmitted to local markets. However recent evidence suggests that internal shocks have a larger impact on prices than external shocks. In an analysis of 554 local commodity markets across 51 countries during the period between 2008-2012, Brown and Kshirsagar (2015) find that 20% of local market prices were affected by domestic weather disturbances in the short-run in comparison to 9% by international price changes. This finding has prompted more recent literature to relax assumptions about international price transmission to investigate how shocks are transmitted through local and regional markets. 

Here we investigate the effects of domestic weather disturbances on regional maize price transmission. We then use these results of to build skillful price prediction models that use limited price data, weather disturbances, and other readily accessible free secondary data to predict monthly grain prices three, six, and nine months ahead in four Southern African countries. The collection of subnational price data in developing countries is costly and often difficult to obtain. We limit the amount of price data used by first determining if monthly price series in each country co-move and how these co-movements are influenced by domestic climate disturbances. We then use bivariate error correction models to both assess whether price movements in each country follow well-defined paths and identify influencing and influenced markets.

From this analysis we classify markets that act as price anchors in each country. Because local climate conditions have been found to affect and accurately predict agricultural prices, price dispersion, and yields in developing countries we use climate conditions at both the market location and anchor market locations as predictors. We show that during periods classified by drought, price prediction models using anchor market prices and high-resolution climate data have high degrees of predictive accuracy. We hope the results presented in this paper will assist policymakers, government stakeholders, and researchers in systematically constructing subnational price forecasts with minimal price data to be used in early warning and food security monitoring models.

How to cite: Anderson, P., Davenport, F., Baylis, K., and Shukla, S.: Using domestic weather disturbances and price transmission for maize price predictions in Southern Africa, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11432, https://doi.org/10.5194/egusphere-egu23-11432, 2023.

The understanding of climate change impacts on tropical storms (TS) activity is crucial for better planning and risk assessment. Despite the theory and modeling suggest an increase in the TS activity with warming, the change in TS characteristics remain uncertain due to the limitations in the global climate models and tracking algorithms (tracker). Here, we performed tracker-inter-comparison and model-evaluation to find out the reliability of trackers and models at simulating the TS characteristics. We found that both trackers produce qualitatively similar results but quantitative different results due to different specifications of the algorithms and model bias. Our results show a decline in the frequency but rise in the strength of TS in the future for the Ganges and the Mekong basin.

How to cite: Ali, H., Fowler, H., Roberts, M., and Vanniere, B.: Change in the Tropical Storms activity in the future over the Ganges basin, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13826, https://doi.org/10.5194/egusphere-egu23-13826, 2023.

Measuring carbon emissions at high temporal and spatial resolution covering all parts of the globe is key to understanding the sources and sinks of carbon. These measurements are critical for informing both climate modeling and policy decisions to mitigate climate change. Fragmented data sources and the requirement of significant programming knowledge to retrieve, clean, and analyze data from existing data sources pose a significant barrier for climate researchers. As understanding of climate science becomes crucial for fields beyond geophysical sciences, it is especially urgent to build tools that can enable researchers from diverse academic backgrounds to analyze carbon emission data from satellites. 

In this presentation, we will present a novel, user-friendly platform which has pre-built functions and analysis pipelines allowing scientists to perform common data analysis tasks without the need to write code. The underlying data lake combines NASA’s Orbiting Carbon Observatory (OCO-2 and OCO-3) data with other data sources (e.g. MODIS-based fire data) that facilitate a more accurate and complete understanding of the dynamics of the carbon cycle and the factors that influence it. 

We highlight how our approach integrating data discovery, access, and analysis of climate data can help democratize climate research and inform policymaking.

Potential research questions that can be addressed using this approach include: 

(i) studying the impacts of fires on the global carbon cycle with MODIS fire products providing information on the location, intensity, and types of fires, 

(ii) studying the photosynthetic activity of plants and the carbon cycle assimilating OCO-2 SIF data. OCO2-SIF data measures the fluorescence emitted by plants as a result of photosynthesis, which can be used as an indicator of plant health and productivity and 

(iii) AI-assisted audit of industrial emissions incorporating publicly available data on critical CO2 emitting sectors e.g. power plants, steel mills, cement plants, atmospheric “spillover” from agricultural and forest fires, traffic emissions, demographic and economic variables, etc

How to cite: Odhiambo, M., Jain, R., Gorana, Y., Kaushal, N., and Mishra, A.: Measuring Carbon: A Tool for Analysing Gridded, Continuous, Carbon Measurements at High Temporal and Spatial Resolution, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15643, https://doi.org/10.5194/egusphere-egu23-15643, 2023.

Cervest, a climate intelligence startup, addresses the need for effective adaptation strategies by bridging the gap between disciplines. We use cutting-edge science techniques such as high-resolution convection-permitting models, remote sensing, hydrology, bayesian statistical modeling, machine learning, and data science to provide accurate, localized physical risk assessments for assets. Our climate intelligence product also accounts for assets vulnerability and multi-hazard, multi-risks. It can be used to assess not only the direct impacts of extreme events but also their indirect effects on supply chains and economic production networks. In this session, we will present our vision for the future of climate intelligence and share our novel probabilistic approach to assessing the impacts of climate change.

How to cite: Morris Boyne, E. P., Prodhomme, C., Pain, A. J., and Laken, B.: Bridging the Gap: Cervest's Climate Intelligence Approach for Effective Adaptation Strategies, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17497, https://doi.org/10.5194/egusphere-egu23-17497, 2023.

The aviation-induced non-CO₂ climate effects, being responsible for two-thirds of aviation radiative forcing [1], have a direct dependency on atmospheric location and time of emissions. This implies that their associated impacts can be mitigated by planning climate-aware trajectories to avoid areas of airspace with large climate effects [2]. However, for the efficiency of such a mitigation strategy, one needs to consider various sources of uncertainty. In fact, if not accounted for within flight planning a priori, the rather immature scientific understanding of aviation-induced climate effects and uncertainty associated with emissions calculation and meteorological conditions can lead to inefficient aircraft trajectories. In addition, the mitigation potential achieved by the climate-optimal routing option increases the operating costs as the aircraft flies longer by re-routing climate-sensitive areas. In this respect, there is a need to plan robust climate-optimal aircraft trajectories having a minimum cost increase compared to the Business-as-usual (BAU) scenario.

In the current study, we present robust climate optimal aircraft trajectory planning, considering meteorological uncertainties. The airspace is assumed to be fully free routing. The information on the spatio-temporal dependency of aviation-induced climate effects is based on the latest version of the prototype algorithmic climate change functions (aCCF V1.1) [3]. An ensemble prediction weather forecast is used to characterize meteorological uncertainty. The flight planning objective is to find an efficient balance between the increased operating costs and the mitigated climate effects with acceptable ranges of uncertainty. The general approach for decision-making between conflicting objectives relies on building a Pareto-frontier by running the optimization many times, each corresponding to a weighting parameter in the objective function (see e.g., [4]). In this study, by proposing a more efficient modeling scheme in the definition of the aircraft trajectory optimization within the context of optimal control theory, we provide an ability to determine the highest possible mitigation potential with a user-specified limit on the increased operating cost and vice versa only in two iterations. In this approach, we define the “Lagrangian” term of the performance index (used to represents climate effects in the objective function) as an additional state variable, enabling to impose path and boundary constraints on the climate effect and its dispersion.

The effectiveness of the proposed approach is illustrated by considering the optimization of 10 flights on June 13, 2018, 0000UTC. Due to the strong variability among different members of relative humidity within the EPS weather forecast, the climate impact of contrails is highly uncertain. This in turn leads to high uncertainty in quantifying the net climate effects due to the dominant impact of contrails compared to the remaining species. For the considered case studies, it is shown that by employing the proposed trajectory optimizer, it is possible to minimize the climate effects while respecting the specified available extra operating cost in US dollars. In addition, the uncertainty on the quantified climate effects lies within the user-defined range, implying that the sensitivity of climate impact to the uncertainty in the forecasted weather conditions can be controlled.

How to cite: Simorgh, A. and Soler, M.: Cost-Effective Climate-Friendly Aircraft Flight Planning, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17504, https://doi.org/10.5194/egusphere-egu23-17504, 2023.

How to cite: Baneshi, F. and Soler, M.: Network Assessment of the Aviation Climate Impact Considering the European Structured Airspace, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17517, https://doi.org/10.5194/egusphere-egu23-17517, 2023.

The simultaneous occurrence of increased river discharge and high coastal water levels may cause compound flooding. Compound flood events can potentially cause greater damage than the separate occurrence of the underlying extreme events, making them essential for risk assessment. Even though a general increase in the frequency and/or severity of compound flood events is assumed due to climate change, there have been very few studies conducted for larger regions of Europe. Our work, therefore, focuses on the high-resolution analysis of changes in extreme events of coastal water levels, river discharge, and their concurrent appearance at the end of this century in Northern and Central Europe (2070-2100). For this, we analyse downscaled data sets from two global climate models for the two emissions scenarios RCP2.6 and RCP8.5.

First, we compare the historical runs of the downscaled GCMs to historical reconstruction data to investigate if they deliver comparable results for Northern and Central Europe. Then we study changes in the intensity of extreme events, their number, and the duration of extreme event seasons under climate change. Our analysis shows increases in compound flood events over the whole European domain, mostly due to the rising sea level. This increase is concomitant with an increase in the annual compound flood event season duration.

Furthermore, the sea level rise associated with a global warming of 1.5K will result in a 50% increase in compound flood events for nearly every European river considered.

How to cite: Heinrich, P., Hagemann, S., and Weisse, R.: Changes in Compound Flood Event frequency in Northern and Central Europe under climate change, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1988, https://doi.org/10.5194/egusphere-egu23-1988, 2023.

Compound hot and dry events (such as recent summers of 2015, 2018 and 2022 in Europe) have an impact on a wide range of sectors, including health, transport, energy production, ecology, agriculture and forestry. The co-occurrence of extreme heat and drought poses a risk to water security in particular, since heat exacerbates moisture shortages during dry periods through increased evapotranspiration while at the same time water demand increases (e.g., for drinking water, cooling, irrigation). Current research suggests that climate change will increase the intensity, frequency, and duration of joint hot and dry extreme events in Europe. However, most studies focus on the drivers applying coarse-resolution global climate models.

This study exploits a 50-member single-model initial condition large ensemble (SMILE) of the Canadian Regional Climate Model, version 5, at 12 km resolution (CRCM5-LE, RCP 8.5 from 2006 onwards, driven by the Canadian Earth System Model Version 2 large ensemble, CanESM2-LE). The application of a regional SMILE provides an extensive database of compound events and, subsequently, robust estimations of their occurrence changes across Europe, from current to future states and in high geographical detail.

We define compound hot and dry summers based on joint exceedances of temperature and (negative) precipitation thresholds (2001-2020 JJA 95^th percentiles). By considering low soil moisture (below regional 2001-2020 JJA 10^th percentile) as an impact indicator, we further show the spatially varying connection between compound hot and dry summers and low water availability in Europe. Compound event occurrences are investigated in a current climate (2001-2020) and future 20-year slices at global warming levels (GWL, derived from the CanESM2-LE) of +2 °C and +3 °C, with each period represented by 1000 model years. Last, we investigate the underlying processes (e.g., heat budget terms) of changing event occurrences and their spatial distribution, and discuss the land use-specific (e.g., urban, agricultural, natural) exposure to impacts on water availability during compound hot and dry summers.

We identify areas in the Mediterranean and northern France as hotspots with a fivefold occurrence frequency of compound hot and dry summers for +2 °C GWL. With +3 °C GWL, the Mediterranean, France, Belgium, southern Germany, Switzerland, and the south of UK and Ireland are affected by a tenfold occurrence frequency with respect to current climate.

This study is an important boundary condition to the development of adaptation strategies for the affected regions.  At the same time, it quantifies the reduction of event occurrence in a +2°C world compared to the higher GWL of +3°C, highlighting the importance of climate mitigation strategies and policies.

How to cite: Böhnisch, A., Felsche, E., Mittermeier, M., Poschlod, B., and Ludwig, R.: Hotspots and impacts of present and future compound hot and dry summers in Europe, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2079, https://doi.org/10.5194/egusphere-egu23-2079, 2023.

Inter-variable correlations (e.g., between daily temperature and precipitation) are key statistical properties to characterize probabilities of simultaneous climate events and compound events. Their correct simulations from climate models, both in values and in changes over time, is then a prerequisite to investigate their future changes and associated impacts. Therefore, this study first evaluates the capabilities of one 11-single run multi-model ensemble (CMIP6) and one 40-member single model initial-condition large ensemble (CESM) over Europe to reproduce the characteristics of a reanalysis dataset (ERA5) in terms of temperature-precipitation correlations and their historical changes.

Next, the ensembles’ correlations for the end of the 21st century are compared. Over the historical period, both CMIP6 and CESM ensembles have season-dependent and spatially structured biases. Moreover, the inter-variable correlations from both ensembles mostly appear stationary. Thus, although reanalyses display significant correlation changes, none of the ensembles can reproduce them, with internal variability representing only 30% on the inter-model variability. However, future correlations show significant changes over large spatial patterns. Yet, those patterns are rather different for CMIP6 and CESM, reflecting a large uncertainty in changes. In addition, for historical and future projections, an analysis conditional on atmospheric circulation regimes is performed. The conditional correlations given the regimes are found to be the main contributor to the biases in correlation over the historical period, and to the past and future changes of correlation.

These results highlight the importance of the large-scale circulation regimes and the need to understand their physical relationships with local-scale phenomena associated to specific inter-variable correlations.

How to cite: Vrac, M., Thao, S., and Yiou, P.: Changes in temperature-precipitation correlations over Europe: Are climate models reliable?, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3133, https://doi.org/10.5194/egusphere-egu23-3133, 2023.

Concurrent extreme events exacerbate adverse impacts on humans, economy, and environment relative to those from independent extreme events. However, while the effects of climate change on the frequency of individual extreme events have been highly researched, the impacts of climate change on the interaction, dependence and joint occurrence of these extremes have not been extensively investigated, particularly in the East African region. Here, we investigate the joint occurrence of six categories of extreme events in East Africa, namely: river floods, droughts, heatwaves, crop failures, wildfires and tropical cyclones using bias-adjusted impact simulations under past and future climate conditions. We show that the change in the probability of joint occurrence of these extreme events in the region can be explained by the effects of climate change on the frequency, spatial distribution, and dependence of these extreme events. The analysis demonstrates that there is a higher positive correlation between most co-occurring pairs of extremes in the region under end-of-century global warming conditions leading to more frequent concurrence in comparison to the early-industrial period. Our results further highlight the most affected locations in the region by these concurrent events and consequently the main driver(s) in the various co-occurring pairs of extremes. Our results overall highlight that concurrent extremes will become the norm rather than the exception in East Africa under low-end warming scenarios.

How to cite: Muheki, D., Deijns, A., Bevacqua, E., Messori, G., Zscheischler, J., and Thiery, W.: The perfect storm? Concurrent climate extremes in East Africa, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3172, https://doi.org/10.5194/egusphere-egu23-3172, 2023.

Both droughts and heatwaves cause negative impact on human health, agriculture, economy and other areas while occurring separately. However, in recent years the impact of these phenomena acting together has been increasingly analysed as it was found that such events, called compound drought and heatwave events (CHDE), may induce even more damage. The aim of this research is to identify droughts, heatwaves and CDHE in the eastern part of the Baltic Sea region during the summer months (June-August) from 1950 to 2022 and to assess their frequency and intensity. For the purpose to identify droughts the 1-month Standard Precipitation Index (SPI) values calculated for each day were used. Droughts were distinguished if the SPI values were lower than -1 for at least five or more days in a row and this condition was met in at least one third of the study area.  Heatwaves were defined as a period of five or more consecutive days when daily maximum air temperature (T_max) was higher than 90th percentile of T_max of the study period (1951–2022) for each summer day (on a 5-day moving average) and for one or more days covered at least one third of the study area. Daily T_max data as well as precipitation data that was needed to calculate SPI were obtained from European Centre of Medium-range Weather Forecast ERA-5 reanalysis dataset with a spatial resolution of 0.25° x 0.25°. CDHE events were defined as time periods when heatwave occurs during the drought period. Study showed that the number of heatwaves in the study area since 1950 increased significantly (by 1.25 per decade). The number of droughts during investigation period slightly decreased. The majority of droughts were identified in 1990’s when dry periods were recorded during six summers in a row (from 1992 to 1997). In total, 19 CDHE during the summer months were distinguished, while a lot of them occurred during 1990‘s (5 events). As a consequence, statistically significant increase of such events during the study period was not observed. CHDE of the highest intensity was found in 1994 while the longest CDHE occurred in 2022 and lasted for 19 days (from August 11^th to August 29^th).

How to cite: Klimavičius, L. and Rimkus, E.: Compound drought and heatwave events in the eastern part of the Baltic Sea region, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3273, https://doi.org/10.5194/egusphere-egu23-3273, 2023.

Countless climate-related impacts are caused by compound events, i.e. by the combination of multiple climate processes at different spatial and temporal scales. For example, when precipitation and wind extremes coincide, the resulting impacts on infrastructure and humans can be very destructive. It is established that climate modes of variability, which are known to oscillate from seasonal to decadal timescales, such as the El Niño-Southern Oscillation (ENSO), North Atlantic Oscillation (NAO), Pacific-North American Pattern (PNA) and Atlantic Multidecadal Variability (AMV) favour the occurrence of extreme weather events such as heavy precipitation in several areas worldwide. However, little is known about the effect that these climate modes of variability have on compound events. In this context, understanding the physical modulators of compound events can contribute to an improved comprehension of their dynamics, and ultimately to a better prediction of their impacts. Here, focussing on compound wind and precipitation extremes, we contribute to closing this research gap by using large ensemble climate model simulations (CESM) and reanalysis data (ERA5). We identify hotspot regions in the northern hemisphere where winter (DJF) compound event occurrences are influenced by modes of variability. We also inspect whether particular combinations of modes of variability, e.g., superposition of extreme states of both ENSO and NAO indices, enhance compound event occurrences. Finally, the identified patterns in the observational data are compared to the model simulations. The findings allow us to understand whether climate modes of variability favour the simultaneous occurrence of compound events over different regions worldwide, and how well the current generation of climate model simulations represents these dynamics.  An improved understanding of these oscillating modes of variability could be used to enhance the development of sub-seasonal to seasonal forecasts of compound events, which therefore may reduce their impacts. 

How to cite: Teber, K., Francois, B., Gimeno-Sotelo, L., Küpfer, K., Brett, L., Leeding, R., Yavuzdogan, A., Domeisen, D., Suarez, L., and Bevacqua, E.: The influence of modes of variability and their interplay on compound extreme wind and precipitation events in the northern hemisphere., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5187, https://doi.org/10.5194/egusphere-egu23-5187, 2023.

The role of climate change in exacerbating the impacts of natural hazards has been the focus of extensive interest. However, while the emphasis is generally on a single hazard (e.g., heat stress, extreme precipitation, floods, droughts), their compounding effects under climate change have been the subject of a growing number of studies. Among compound events, heat stress was recently found to be a precursor of summer flooding across the central United States. We show for the first time that heat stress can trigger floods across large areas of North and South America, southern Africa, Asia and eastern Australia. Moreover, using global climate models from the sixth phase of the Coupled Model Intercomparison Project (CMIP6), we show that the compounding of heat stress and floods is projected to worsen under climate change with effects magnified as we move from the Shared Socioeconomic Pathways (SSPs) 1-2.6 to 5-8.5. Under future conditions, the compounding between heat stress and floods is projected to extend to Europe and Russia due to the increased warming. These results highlight the need towards improved preparation and mitigation measures that account for the compound nature of these two hazards, and how the compounding is expected to be exacerbated because of climate change.

How to cite: Noto, L. V., Treppiedi, D., and Villarini, G.: Climate Change to Exacerbate the Compounding of Heat Stress and Flooding, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5942, https://doi.org/10.5194/egusphere-egu23-5942, 2023.

Seasonal droughts are a common feature of the Iberian climate. They can have severe socioeconomic and ecological impacts – especially, when recurring in consecutive years. We investigate the recurrence of extreme drought events in the Iberian Peninsula (IP) for the past decades and in regional climate change projections. With this aim, we introduce and apply a new set of indices: the Recurrent Dry Year Index (RDYI) and the Consecutive Drought Year (CDY) Index. For the present climate, different gridded observational and reanalysis datasets at several spatial resolutions (10 to 25 km) are used. To analyse the potential impacts of climate change, we apply the indices to a large EURO-CORDEX multi-model ensemble with 12 km horizontal resolution consisting of 25 different global-to-regional model (GCM-RCM) chains for the historical period and future periods along the RCP8.5 scenario.

Results show that the IP is regularly affected by extreme droughts under present climate conditions, with roughly three individual events per decade. Especially the southern and central parts of IP are exposed to recurrent events, which last between two and six years. Under different global warming levels (GWLs), results reveal a general tendency towards more severe drought conditions. Moreover, recurrent drought events are projected to occur more frequent and last longer (up to 14 years). While the ensemble mean responses are only moderate for a GWL of +2°C (compared to the pre-industrial average), recurrent drought conditions are strongly enhanced for the +3°C GWL. The climate change signals are robust for most of IP and the considered recurrent drought indices, with a larger model agreement for the +3°C GWL. For both present and future climate conditions, results show some sensitivity on the choice of index and dataset.

We conclude that the new indices are suitable for the detection and evaluation of recurrent drought events under present and future climate conditions. With ongoing climate change, the Iberian Peninsula faces an increased risk of recurrent drought events. If global warming should exceed the +3°C threshold, the majority of models projects an almost permanent state of drought – with severe consequences for the Iberian population and ecosystems.

How to cite: Moemken, J., Koerner, B., Ehmele, F., Feldmann, H., and Pinto, J. G.: Recurrence of drought events over Iberia under present and future climate conditions, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6226, https://doi.org/10.5194/egusphere-egu23-6226, 2023.

Severe winter wind storms are related with strong impacts. We could show in recent studies that the co-occurrence of extreme wind and precipitation is leading to higher damages of residential buildings in comparison to non-compound events. This was done using ERA5 reanalysis data for the European winter season and daily insurance records of damages for residential buildings over Germany provided by the German Insurance Association (GDV).

This study investigates the representation of co-occurrence of extreme wind and precipitation for climate simulations of the Coordinated Regional Climate Downscaling Experiment (CORDEX) for Europe (EURO-CORDEX). We use multi-model ensemble simulations with horizontal resolution of 0.44° and 0.11°. Individual simulations are analysed with respect to the characteristic of compound events for historical projections. Climate change signals for future scenarios are discussed.

How to cite: Grieger, J. and Ulbrich, U.: Compound precipitation and wind extremes under recent and future climate conditions, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6283, https://doi.org/10.5194/egusphere-egu23-6283, 2023.

Extreme events such as heat waves and droughts can have major impacts on agriculture, human health, and the energy sector, especially during the co-occurrence of such events. Although there is evidence that heat waves and drought have increased in intensity and frequency in the last decades, the analysis, characterization, and impact assessment of the compound occurrence of drought and heat waves are not well documented yet in a common framework. There are still some open questions related to how changes in midlatitude circulation may transcend in the thermodynamical characteristics of these compound events in the future. Furthermore, the role of some local feedbacks and the relationship with other extremes are still a debating subject.

The purpose of this research is to shed some light and add evidence about the key drivers related to these extreme events. The main atmospheric characteristics of compound heat waves and drought events in Europe and North America are identified through the analysis of the ERA5 dataset during the historical period (1959-2022). Additionally, we evaluate the ability of CMIP6 models with respect ERA5 to reproduce the statistics of these compound events. Specifically, we aim at understanding what are the climatological characteristics of these events in the historical climate and what are the dynamical mechanisms leading to compound occurrence of heat waves and droughts.

How to cite: Castillo, N., Gaetani, M., and Martina, M.: Characterization of compound occurrence of heat waves and drought in Europe and North America, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6479, https://doi.org/10.5194/egusphere-egu23-6479, 2023.

In wintertime, infrastructure and property in NW Europe are threatened by multiple meteorological hazards, and it is increasingly apparent that these exacerbate risk by tending to co-occurring in events that last days to weeks. Impacted by Atlantic storms, Great Britain (GB) is a sentinel location for weather that later tracks into NW Europe.   A recent, dramatic storm sequence (Dudley, Eunice, Franklin) demonstrated the need for a multi-hazard view by bringing a mixture of damaging and disruptive extremes including extreme winds and flooding over 7-10 days in Feb 2022.

This work uses a stakeholder inspired, event-based approach to jointly consider these two hazards.  A wind event set (n = 3,426) is created from the 12km regional UK Climate projections (1981-1999, 2061-2079) to match previously created high-flow events (Griffin et al, 2023). Then, the two hazards’ time-series are merged using windows up to a maximum size (Δt = 1-180 days) positioned to maximize the size of the largest events’ impact. The benefits and limitations of this methodology are discussed, anatomy of storm sequences (Δt = 21 days) discussed, and potential drivers of co-occurrence in the multi-hazard sequences (e.g. jet stream position/strength) examined.

How to cite: Hillier, J., Bloomfield, H., Garry, F., Bates, P., and Shaffrey, L.: Co-occurring British flood-wind events (1980-2080): Their anatomy & drivers, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7203, https://doi.org/10.5194/egusphere-egu23-7203, 2023.

Heatwave events have been increasing in frequency, duration, and intensity along the past decades, leading to severe impacts on ecosystems, human health and basic resources. These events are projected to continue increasing associated to anthropogenic activity. Moreover, droughts have also been more recurrent and intense, which can significantly impact agriculture and reservoirs’ water level and quality.

Events of high temperature can occur both in the atmosphere and the seas. These warmer conditions, together with extremely dry episodes, have been affecting southern Europe and the Mediterranean region, which appear to be very sensitive to climate change. Additionally, the co-occurrence of droughts and heatwaves increases meteorological fire danger, rising the probability of wildfire occurrence and severity and resulting in economic, ecological, and even human losses.

In this sense, it becomes fundamental to pay special attention to the role of compound events and synergies in fueling extreme fire outbreaks. Therefore, we propose to address this problem by analyzing the occurrence of both marine and atmospheric heatwaves and drought conditions over Southern Europe, East Atlantic and Mediterranean Sea (relying on ERA5 reanalysis), as well as the recorded wildfires (through MODIS burned area product).

This work aims to address the occurrence of heatwaves (marine and atmospheric) and previous and contemporaneous drought episodes on a compound or cascading approach, estimating their contribution to the occurrence of extreme wildfires in the region in the last decades were analyzed on a seasonal scale.

Acknowledgments: This study is partially supported by the European Union’s Horizon 2020 research project FirEUrisk (Grant Agreement no. 101003890) and by the Portuguese Fundação para a Ciência e a Tecnologia (FCT) I.P./MCTES through national funds (PIDDAC) – UIDB/50019/2020- IDL,  DHEFEUS - 2022.09185.PTDC

How to cite: Santos, R., Russo, A., and Gouveia, C. M.: Assessing the impact of marine and atmospheric heatwaves on droughts and fire activity in the Mediterranean region, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7779, https://doi.org/10.5194/egusphere-egu23-7779, 2023.

Investigating the global connectivities of extreme events is vital for accurate risk reduction and adaptation planning. While human and natural systems have a certain resilience level against single extremes, they may be unable to cope with multiple extreme events whose impacts tend to be amplified in a non-linear relationship. Concurrent droughts and heatwaves are frequently linked to severe damage in socioeconomic sectors such as agriculture, energy, health, and water resources. They can also have detrimental effects on natural ecosystems. Here, we detect global scale dependencies of large-scale droughts and heatwaves using an AI-enhanced point process-based approach, where large-scale events are defined to occur when a certain amount of grid points (e.g., 20%) of a given region of interest experiences heatwave or drought conditions. The classic inhomogeneous and non-stationary J-function can determine whether the occurrence of the events shows clustering, inhibition or independence. However, the analysis and interpretation of this function are usually affected by a high degree of subjectiveness, and its application for large datasets and/or ensembles is challenging. The proposed AI-based automated interpretation tool replaces a subjective and user-dependent approach. Monte Carlo simulations based on standard point process models, reflecting the aforementioned dependence structures, are utilized, allowing the dependence structure to be labeled and the classification problem to be trained using Deep Learning algorithms. To identify the global connectivities of large-scale droughts and heatwaves, we first detect extreme events at the grid scale based on appropriately selected indices. A cluster analysis pinpoints areas with similar drought and heatwave patterns, thus identifying the regions of interest for the large-scale events. For these events we compute the J-functions, and the dependence structure of the large-scale events is then classified by the AI-tool. Links to teleconnections (such as the El Niño-Southern Oscillation and the North Atlantic Oscillation) can be further identified by analyzing the dependencies conditioning on the teleconnection phase under consideration. The proposed tool can be used in diverse research questions where a point process approach is appropriate, and thus has applications beyond climate science.

How to cite: Luther, N., Toreti, A., Pérez-Aracil, J., Salcedo-Sanz, S., Vlachopoulos, O., Ceglar, A., Hrast Essenfelder, A., and Xoplaki, E.: Detecting dependencies of large-scale heatwaves and droughts with AI-enhanced point process approaches, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8588, https://doi.org/10.5194/egusphere-egu23-8588, 2023.

Synchronous crop failure among multiple breadbaskets worldwide, a typical spatially compound event, may amplify threats to the global food system and food security and has been a growing concern among the scientific community in recent years. While the risk of simultaneous crop loss across multiple breadbasket regions has been analyzed, to date, little is known about how interdependence among regional crop production affects aggregated crop failure at the global scale. Quantifying the impact of dependencies among breadbasket regions on global food production and assessing how the dynamic of spatially compounding crop failures is simulated by climate and crop models is essential for informing the modeling of global food security risk. In this study, focusing on different crop types, we quantify the influence of dependence between crop production of individual regions on global aggregated crop yield based on the Inter-Sectoral Impact Model Intercomparison Project (ISIMIP) dataset. We find that spatial dependence between regional crop yields may aggravate global crop deficits and identify a characteristic spatial scale beyond which the dependence between crop production in different regions vanishes. Our results provide valuable information for designing risk strategies for food security at the suited scale.

How to cite: Feng, S., Zscheischler, J., Hao, Z., and Bevacqua, E.: Interdependence among subregional crop production affects global crop failure risk, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8705, https://doi.org/10.5194/egusphere-egu23-8705, 2023.

Polders can be found in coastal and alluvial lowlands all over the world. These polders need an internal drainage system consisting of drainage canals, weirs and/or pumps to discharge the water out of the polder. Next to these drainage canals, dikes can protect the low-lying polder areas that are situated several meters lower than the controlled water levels in these canals. This study investigates the joint impact of extreme rainfall events on water and dike systems within Dutch polders. Previous research has shown that the combined effect of heavy rainfall and storm surge can increase flood risk in coastal polders in the Netherlands. However, the impact of extreme rainfall on multiple water-and-dike systems within a single polder, resulting in multiple hazards, has received little attention. Our analysis uses physical models that are calibrated on measurements and forced by synthetic rainfall and evaporation time series to examine the response time and interdependencies between regional drainage systems and pore-water pressures in canal dikes. Water levels and pore-water pressures and their interrelationships were analyzed as indicators of flood hazards. Our findings demonstrate the importance of considering the joint impact of multiple hazards on flood risk in polders, as the functioning of regional drainage systems and canal dikes can be affected by similar weather events.

How to cite: Strijker, B. and Kok, M.: The joint impact of rainfall events on water- and dike systems in Dutch polders, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9048, https://doi.org/10.5194/egusphere-egu23-9048, 2023.

Although studies based on physiological models have repeatedly shown that high humidity levels lead to stronger heat stress in humans, findings from epidemiological studies have remained inconclusive on the matter till date. We aim to employ a ‘bottom-up’ strategy of identifying key drivers of compound events to explore the role played by humidity in high heat-related mortality events, spanning across multiple cities in multiple countries. We used daily data on all-cause mortality, mean temperature and mean relative humidity from 11 cities across the world and applied state-of-the-art epidemiological models to compute the daily observed total mortality counts attributable to heat (i.e., limited to days with average temperature exceeding the ‘temperature of minimum mortality’ (MMT) in each city). Each of these days with mean temperature exceeding MMT is considered as an ’event’ and events with highest mortality counts attributable to heat from multiple cities are analysed in a 2D scatter plot of the corresponding percentile rank of temperature and humidity observed during those events. The frequency of such high impact events in the temperature-humidity percentile space across multiple cities, categorised into sub-groups based on the temperature and humidity climatology of the cities, was then studied. It was observed that close to 90% of the high impact events occurred during high temperature (> 90^th percentile) and non-high humid (<50^th percentile) conditions. The events of high severity, where humidity conditions were comparatively high (> 50^th percentile), were mostly representative of cities with prevailing high humidity conditions on average during the warmest months, when compared across all the cities. Based on our preliminary findings, there is no conclusive evidence that high humidity conditions were prevalent during high heat-mortality impact events, but further analysis incorporating more cities and other climatological variables of interests such as absolute humidity, wet-bulb globe temperature etc. are planned. This novel framework provides valuable insights into the role of humidity in heat stress mortality and can be generalised to address other similar complex research questions in environmental epidemiology.

How to cite: Sivaraj, S., Lüthi, S., Lo, E., and Vicedo-Cabrera, A. M.: A bottom-up approach for exploring the role of humidity in high heat-related mortality events: A Multi-City, Multi-Country study, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10996, https://doi.org/10.5194/egusphere-egu23-10996, 2023.

Heavy rainfall events and urban flash floods pose a high risk potential for humans and the environment, as a concrete prediction of the regional impacts is difficult. The effects of heavy rainfall events and urban flash floods depend, among other things, on the characteristics of the respective affected area - such as land use, soil type or topographical factors - but also on prior conditions, especially the pre-rainfall index. River floods also pose a similarly high risk, even if they can be predicted more precisely than heavy rainfall events - especially in larger river systems - and thus a more focused flood risk management can be carried out. If these events overlap in the form of compound flooding from river floods and heavy rainfall, the hazards and the risk to people and the environment increase significantly. This was shown in particular by the flood disaster in July 2021 in Rhineland-Palatinate and North Rhine-Westphalia in Germany.

Investigations are carried out into the joint occurrence of river floods and heavy rainfall. Discharge data from various stream gauges in North Rhine-Westphalia (Germany) and precipitation data from radar data of the German Weather Service as well as ERA5-Land reanalysis data of the ECMWF are used for this purpose. First, the respective single events are identified and analysed with regard to various statistical parameters. Then the analysis of the compound events is carried out, considering only events that are identical in time and space. To take this into account, simultaneous series are formed from the time series available. Since not all catchments are equally at risk from compound river flood and heavy rainfall events, one focus is on determining vulnerable areas. Here, various characteristic attributes of the catchments but also weather conditions, such as the pre-rainfall index, are considered. It turns out that special attention must be paid to small to medium-sized catchments and to areas with steep and narrow valleys.

Furthermore, the joint occurrence probability of river floods and heavy rainfall is determined. This is done with archimedean copula functions. A statement on the joint probability of occurrence of river floods and heavy rainfall has not yet been included in practice or in standards but should be adopted for the correct determination of hazards and risks. Furthermore, based on the analyses carried out, a proposal for the preparation of flood hazard maps by compound river floods and heavy rainfall is presented.

How to cite: Simon, F. and Mudersbach, C.: Analysis of compound river flood and heavy rainfall events for a development of combined flood maps, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11000, https://doi.org/10.5194/egusphere-egu23-11000, 2023.

Abstract

The co-occurrence of temperature and precipitation extremes can have profound consequences than either individual extremes. The role of increasing warm spells in increasing CHDEs has been studied in various studies, but the role of active and break spells on CHDEs during monsoon has not been studied. As a result, in this study, we investigated the fraction of CHDEs in both active and break spells in India. We used copula and threshold-based methodology to characterize CHDE to investigate the uncertainty in the frequency of CHDEs during active and break spells. We also looked at how CHDEs in two different spells will impact society differently. We further investigated the changes in CHDEs to future projections of active-break spells of the Indian Summer Monsoon. The findings of the study may help to mitigate the severe impacts of compound hot and dry extremes in the future.

Keywords: Climate change, Compound extremes, active spells, dry spells

How to cite: Malik, I. and Mishra, V.: Unfolding the role of active/break spells in compound hot and dry extremes (CHDE) in India, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11029, https://doi.org/10.5194/egusphere-egu23-11029, 2023.

Floods are among the most impactful disasters especially in terms of economy in affecting humans’ activities and damaging infrastructures. This is particularly the case along the coast where coastal floods happen. Such floods can be due to three different factors: meteorological (precipitation), hydrological (river runoff) and oceanographic (storm surge). A single factor but also a combination of two or more of such factors happening at the same time can lead to coastal floods also called compound floods. Flood hazards can then be underestimating when compound effects are not considered. 

This study focuses on coastal compound floods from oceanographic and hydrological phenomena at the coastal city of Halmstad (Sweden). It aims to quantify the risk of such flood events at Halmstad and to analyse the sensitivity of data sources and copula’ approaches.

Here, the copula method is used to analyse compound floods based on annual maxima of river discharge and corresponding sea level and vice-versa. A comparison is carried out with the commonly used Extreme Value theory on a single factor and the compound approach. Effects from different data time-series available from observations and models for both river discharge and sea level are studied.

This paper concludes the presence of a higher risk of flooding when compound effects are not considered and that the choices made on input datasets and copulas can have a significant impact.

How to cite: Dubois, K., Andreas Dahl Larsen, M., Drews, M., Nilsson, E., and Rutgersson, A.: Compound storm surge and river flood events in the coastal zone: Exploring the influence of data sources and compound approach on extreme recurrence levels, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11220, https://doi.org/10.5194/egusphere-egu23-11220, 2023.

Climate change increases the frequency and severity of heat waves, which can negatively impact human health. Extreme heat can lead to heat stroke, dehydration, and other heat-related illnesses. Heatwaves are more severe for vulnerable populations such as older adults, young children, and people with pre-existing medical conditions. In this study, we analyze the occurrence of compound extreme heat-related mortality in five Swiss cities using neural networks.

To define the excess mortality due to compound heat extremes (Hot day, Tmax>30^oC, followed by a tropical night, Tmin>20^oC) we compared mortality during the four hot summers of 2003, 2015, 2018, and 2019 with long-term average mortality rates (1981-2020). We trained long short-term memory (LSTM) neural networks on 40-year time series of maximum and minimum temperatures, hot day / tropical night compound events, and mortality in Basel, Bern, Geneva, Lugano, and Zürich.  LSTM neural networks learn the important parts of the sequence seen so far and forget the less important ones. This makes these models predict with greater accuracy than traditional time series analysis methods.

In general, we found that over the past 40 years, more than six percent of deaths were caused by compound extreme heat waves in the five Swiss cities. Geneva and Lugano are the most affected cities by compound heat, but the risk of heat-related mortality has decreased in these two regions over time, which could be a result of the action plans that exist in the Latin regions of Switzerland.

We further used Switzerland's future climate model scenarios (CH2018), to predict mortality rates in Swiss cities in the near-future (2020–2050) and far-future (2070–2100). We projected that the number of people affected by mortality risks associated with heat could increase by three folds by the end of the century in most cities if no further adaptation is taken place.

Our results show how important it is for governments, public health agencies, and individuals to be aware of the potential impacts of climate change on heat-related mortality and to take steps to mitigate and adapt to these impacts.

How to cite: Ashraf Vaghefi, S., Muccione, V., Vicedo-Cabrera, A., Neukom, R., Huggel, C., and Salzmann, N.: Projecting the occurrence of extreme heat-related mortality using long short-term memory networks in cities of Switzerland, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12245, https://doi.org/10.5194/egusphere-egu23-12245, 2023.

Extreme meteorological events such as frost, heat, and drought can induce significant damage to vegetation and ecosystems. In particular, heat and drought events are projected to become more frequent under a changing climate. It is therefore crucial to predict the frequency (on climate timescales) and the occurrence (on timescales of weeks to months) of such extremes.

The subseasonal-to-seasonal (S2S) forecasting timescale refers to forecasting timescales from two weeks to a season. Skillful S2S forecasts of hydro-meteorological hazards can be of crucial importance to prevent large-scale vegetation damage. The utility of S2S forecasts for vegetation is very broad (agriculture, biodiversity and flora protection, wildfire risk management, forest management, etc.).

We focus here on forest damage, defined as negative anomalies of the normalized difference vegetation index (NDVI). We use the AVHRR dataset, providing NDVI data over Europe. Compound droughts and heat waves are known to trigger low NDVI events in summer. A dry summer combined with moist conditions during the previous autumn can also have a negative impact. The idea is to find, among all the hydrometeorological variables available as S2S forecast in the ECMWF model, the most relevant ones to predict forest damage. For that, we establish an automated procedure to identify the compound hydro-meteorological conditions leading to low NDVI events, up to several seasons before the impact. We train a model using ERA5 and ERA5-Land reanalysis datasets for the explicative variables. These variables include temperature, precipitation, dew point temperature, surface latent heat flux, soil moisture, snow water equivalent, soil temperature, etc. Several space and time aggregations are considered in order to find the optimal scales and most relevant combinations of variables to predict low NDVI events. The overall goal of this research project is to bridge the research gap between the S2S forecast of hydrometeorological variables and vegetation damage in general. For that, we assess the forecast skill of variables identified as responsible for compound low NDVI events and vegetation biodiversity loss.

How to cite: Rivoire, P., Domeisen, D., Guisan, A., and Vittoz, P.: Compounding hydro-meteorological drivers of forest damage over Europe, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12427, https://doi.org/10.5194/egusphere-egu23-12427, 2023.

Compound extreme events describe the simultaneous occurrence of two or more individual extreme weather or climate events that often have a significant impact on environment, society or economy. Many studies have investigated such events, often using different spatiotemporal scales for the same event, depending on e.g., the country or region of interest. Although appropriate from an impact point of view, this practice might lead to conflicting or inconsistent results. It is therefore necessary to find objective definitions of extreme events for attribution studies or to investigate how likelihoods of certain extreme events change over time.

Building on previous work for single extreme events, we propose a roadmap for obtaining objective compound event definitions, especially with regards to their spatiotemporal characteristics, by estimating multivariate probability distributions via copulas and then maximizing the rarity of the event across several scales. We present applications to past compound extreme events with considerable impact on e.g., human health and agriculture, such as the European heat wave/high ozone event in summer 2003, and also investigate how probabilities of these events change under different emission scenarios.

How to cite: Schuhen, N., Sillmann, J., Cattiaux, J., and Iles, C.: Defining compound extreme events on objective spatiotemporal scales, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13562, https://doi.org/10.5194/egusphere-egu23-13562, 2023.

From a risk management perspective, compound events are very relevant because they can significantly increase the intensity and/or the spatial and temporal extension of the impact. Thus, depending on their magnitude, they may overwhelm the capability of emergency-response services to cope with “unusual” situations of major damage and respond to a large number of emergency situations throughout the region at the same time, and/or have to maintain the level of response during a relatively long period of time. When an extreme compound event occurs, its characteristics depart from the idealized conditions that are usually analyzed and, from the risk management perspective, the problem becomes highly multidimensional. This will be illustrated with the impact of the Gloria storm on the Spanish Mediterranean coast in January 2020. During five days extreme conditions (with some record breakings) of multiple hazards (wind, waves, rainfall, river discharge and surge) were recorded. In places such as the mouth of the Tordera River, they occurred simultaneously, but the most common situation was that different extreme conditions of univariate hazard occurred in remote areas of the territory, although they had to be managed simultaneously. In addition, the storm caused massive damage of various kinds, affecting transportation infrastructure, railway services, breakwaters, docks, urban services, housing, agricultural land and four fatalities in Catalonia. As a result of this, although the storm lasted about five days, the management of its impacts was much more extended, so that several months later some repairs were still being carried out. Looking to the event, the analysis of its probability of occurrence will be significantly affected by the adopted perspective. Thus, from the “physical” point of view, the analysis would range from the simplest joint probability of some hazards occurring in a given location (classical 2-drivers multivariate events) to multiple hazards over the whole territory (spatially compound with up to four concurrent hazards). From a "management" point of view, the analysis would focus on the probability of different types of damage (and their corresponding services) occurring at the same time, and on the probability of providing services in remote parts of the territory (and, consequently, dividing the available services) within a short period of time. To illustrate this possible multidimensional study plane, we will map the compoundness of the Gloria storm encompassing its induced hazards, impacts, damage and response. 

This work was supported by the Spanish Agency of Research in the framework of the C3RiskMed project (PID2020-113638RB-C21/ AEI / 10.13039/501100011033).

How to cite: Aguilera Vidal, M., Jimenez, J. A., Llasat, M., Castan, S., and Llasat, C.: How compound can a compound event be? Mapping the compoundness of the Gloria storm, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14152, https://doi.org/10.5194/egusphere-egu23-14152, 2023.

Global climate change is altering the frequency, intensity, and timing of drought and late-spring frost (LSF). European beech, an ecological and economical cornerstone of European forestry, has been shown to be susceptible to both extremes. Since recovery from both drought and frost damage requires access to stored carbohydrate reserves, the joint occurrence of drought and late-frost exacerbates the deleterious effects on forest health. Both extremes are projected to increase in frequency with increasing temperatures, yet, a statistical model for compound drought and late-spring frost events over time is still lacking. Thus, in order to facilitate forest risk assessment, we quantify the joint probability of drought and spring late-frost risk in the historic domain and identify shifts in this dependency across multiple, future climate change scenarios. Analogously, we determine the individual probability of both drought and LSF to determine the contribution of each extreme to the joint probability. 

We determine frost risk based on the minimum temperature during the period of leaf flushing as predicted by a phenological model. Drought risk is quantified using the Standardized Precipitation Evapotranspiration Index (SPEI). To quantify the joint risk of these two extremes while accounting for climatic and topographical covariates, we use vine copula based models. Specifically,  we apply a novel, regular vine copula based regression model, Y-vine copula regression, designed for a two-response regression setting.

We establish a historical baseline for the joint probability of drought and LSF and identify critical climatic and topographic covariates. Subsequently, we repeat the analysis with climate projections for three different scenarios (RCP 2.6, RCP 4.5, RCP 8.5). We identify differences in the joint probability of drought and LSF across the three climate change trajectories, yet note, that the critical covariates remain constant across scenarios. To further disentangle the coupling between drought and LSF, we use a single response, D-vine copula to determine probability and critical covariates for each extreme separately. Consequently, we are able to determine whether the risk of frost and drought change in concert, how this differs between climate change scenarios, and which covariates drive each extreme. 

How to cite: Meyer, B. F., Tepegjozova, M., Rammig, A., Czado, C., and Zang, C. S.: Quantifying climate change induced shifts in the risk of jointly and individually occurring drought and late-spring frost, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15290, https://doi.org/10.5194/egusphere-egu23-15290, 2023.

Compound climate-related events are impactful extreme events in which the interactions between multiple variables amplify the final impact. They may be classified depending on the types of interaction and the scales involved. For example, temporal compounding events are characterized by the occurrence of subsequent events in time, as in case of a temporal clustering of precipitation. This last trigger is of great importance when the antecedent soil saturation shapes the intensity or occurrence of a given natural hazard, like for floods or deep landslides. Here, we focus on the characteristics of temporal clustering of precipitation over the Italian territory and its link with landslides occurrence. First, we investigate the spatial and temporal distribution of temporal clustering and the synoptic conditions more prone to it, using Era5-Land dataset. Second, we link the identified clusters with the occurrence of different movements’ types (complex, debris flow, fall, flow, and sliding), using a shuffling procedure to assess the significance. Regarding the first point, clear differences emerged between the Italian regions and the four seasons. Clusters were more widespread in autumn and spring and more localized in winter and summer. During winter, we observed a negative link between the number of clusters and the Mediterranean oscillation index in south-central Italy. Regarding the second point, differences were found between the five landslide typologies: fall events were mostly preceded by an intense precipitation event, debris flow by a temporal clustering over small windows and complex, flow, and sliding with a temporal clustering over long windows.

How to cite: Banfi, F. and De Michele, C.: Temporal compounding of precipitation and its occurrence before landslide events over the Italian territory, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15427, https://doi.org/10.5194/egusphere-egu23-15427, 2023.

Increases in the frequency and intensity of hydroclimatic extremes (floods and droughts) and their temporal swings have led to severe consequences in many regions around the world. Traditionally, these contrasting extremes have been assessed in isolation without considering their spatial and temporal interactions, implications for infrastructure design and management and the overall compounding risks. Nonetheless, understanding the changing characteristics of such lagged compound events is critical to developing effective mitigation and adaptation strategies. In this study, we propose a novel framework to identify and characterize the hydroclimatic whiplash events and investigate their spatiotemporal projections under climate change. Multiple hydroclimate variables such as precipitation, evapotranspiration, soil moisture, runoff, and streamflow are used to identify dry and wet extremes and their transitions. Different scenarios for nonstationary hydrological swings between flood and drought are investigated based on streamflow data. Meteorological wet and dry conditions are investigated using standardized drought indices calculated based on the downscaled and statistically bias-adjusted simulations of CMIP5 for 1.5°C-4 °C global warming levels over three major river basins in northwest North America. Further, three dry-wet spell indices estimated by precipitation, soil moisture, and runoff simulations are merged into an integrated indicator to provide a thorough perspective on the changing risks of such transitions across North America using the Canadian Regional Climate Model version 4 Large Ensemble. We apply an ensemble pooling approach to enhance the sample size for index estimation, which enables projecting the characteristics more robustly. Frequency, intensity, transition time, spatial fraction, aggregation index, and seasonality are quantified for each warming period and compared with those of the baseline period to investigate their projected changes. In addition, we assess the contribution of external forcing and internal variability to the historical and projected changes of the lagged compound events. The results of this study suggest that hydroclimatic whiplash across North America is expected to become more frequent and intensified in a warmer climate. Projections show overall increases in the frequency of hydroclimatic whiplash and a decrease in the corresponding transition times as the climate gets warmer. In addition, the magnitude, intensity, and duration of wet and dry components of such lagged compound events are projected to increase based on the analyses with streamflow. Increasing trends of spatial fraction and spatial aggregation during both transitions between dry and wet spells also imply higher risks and future challenges for water resources management. The findings of this study support the necessity of developing appropriate mitigation measures targeting lagged compound floods and droughts that can lead to severe environmental and socio-economic disasters in North America.

How to cite: Najafi, M. R., Na, W., Rezvani, R., and Rahimi Movaghar, M.: Characterization of Lagged Compound Floods and Droughts Under Climate Change, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16867, https://doi.org/10.5194/egusphere-egu23-16867, 2023.

The Bay of Bengal and the Arabian Sea adjoining the coastline of India are breeding grounds for depressions and tropical cyclones, with 2 to 3 cyclones making landfall every year on average. The frequency and intensity of compound coastal flooding events are expected to increase as the world continues to warm. The impact of these events will also be more due to the growing exposure and vulnerability of human settlements in the coastal areas of India. The compound coastal flooding events are primarily driven by extreme sea levels and heavy rainfall during tropical storms and depressions making landfall near the coast. However, there is no comprehensive study on the trends in compound flooding scenarios with reference to Indian coastline. This study presents results from an analysis of compound extreme flood events in the Indian coastal region and assesses the change in frequency and intensity of these events based on in-situ data for the period 1980-2020. The hourly sea-level data was obtained from 9 Tide Gauge stations (TGs) operated by the Survey of India. The daily rainfall data at these stations are extracted from 0.25° resolution gridded rainfall product of the India Meteorological Department (IMD). Harmonic analysis is carried out on the detrended sea-level data to separate the astronomical tide component and obtain skew surge time series at predicted high tide timesteps. The extremes corresponding to 90^th, 95^th and 98^th percentile thresholds are identified for both skew surge and rainfall time series, and the co-occurrence probability of the two extreme events is analysed for the historical data. The evolution of frequency and intensity of the potential compound flood days over the historical period is also investigated.

How to cite: Dutta, D., Srinivas, V. V., and Bala, G.: Trends in the Frequency and Intensity of Compound Coastal Flooding Events along the Indian coastline during 1980-2020, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17187, https://doi.org/10.5194/egusphere-egu23-17187, 2023.

As deep learning (DL) is gathering remarkable attention for its capacity to achieve accurate predictions in various fields, enormous applications of DL in geosciences also emerged. Most studies focus on the high accuracy of DL models by model selections and hyperparameter tuning. However, the interpretability of DL models, which can be loosely defined as comprehending what a model did, is also important but comparatively less discussed. To this end, we select thin section photomicrographs of five types of sedimentary rocks, including quartz arenite, feldspathic arenite, lithic arenite, dolomite, and oolitic packstone. The distinguishing features of these rocks are their characteristic framework grains. For example, the oolitic packstone contains rounded or oval ooids. A regular classification model using ResNet-50 is trained by these photomicrographs, which is assumed as accurate because its accuracy reaches 0.97. However, this regular DL model makes their classifications based on the cracks, cements, or even scale bars in the photomicrographs, and these features are incapable of distinguishing sedimentary rocks in real works. To rectify the models’ focus, we propose an attention-based dual network incorporating the microphotographs' global (the whole photomicrographs) and local features (the distinguishing framework grains). The proposed model has not only high accuracy (0.99) but also presents interpretable feature extractions. Our study indicates that high accuracy should not be the only metric of DL models, interpretability and models incorporating geological information require more attention.

How to cite: Zheng, D., Cao, Z., Hou, L., Ma, C., and Hou, M.: High accuracy doesn’t prove that a deep learning model is accurate: a case study from automatic rock classification of thin section photomicrographs, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-244, https://doi.org/10.5194/egusphere-egu23-244, 2023.

Starting from 2021 International Maritime Organization (IMO) introduced more demanding NO_x emission restrictions for ships operating in waters of the North and Baltic Seas. All methods currently used for ship compliance monitoring are financially and time-demanding. Thus, it is important to prioritize the inspection of ships that have a high chance of being non-compliant. 

TROPOMI/S5P instrument for the first time allows a distinction of NO₂ plumes from individual ships. Here, we present a method for the selection of potentially non-compliant ships using automated machine learning (AutoML) on TROPOMI/S5P satellite data. The study is based on the analysis of 20 months of data in the Mediterranean Sea region. To each ship, we assign a Region of Interest (RoI), where we expect the ship plume to be located. We then train a regression model to predict the amount of NO₂ that is expected to be produced by a ship with specific properties operating in the given atmospheric conditions. We use a genetic algorithm-based AutoML for the automatic selection and configuration of a machine-learning pipeline that maximizes prediction accuracy. The difference between the predicted and actual amount of produced NO₂ is a measure of inspection worthiness. We rank the analyzed ships accordingly. 

We cross-check the obtained ranks using a previously developed method for supervised ship plume segmentation.  We quantify the amount of NO₂ produced by a given ship by summing up concentrations within the pixels identified as a “plume”. We rank the ships based on the difference between the obtained concentrations and the ship emission proxy.

Ships that are also ranked as highly deviating by the segmentation method need further attention. For example, by checking their data for other explanations. If no other explanations are found, these ships are advised to be the candidates for fuel inspection.

How to cite: Kurchaba, S., van Vliet, J., Verbeek, F. J., and Veenman, C. J.: Detection of anomalous NO2 emitting ships using AutoML on TROPOMI satellite data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1183, https://doi.org/10.5194/egusphere-egu23-1183, 2023.

Compaction of agricultural soil negatively affects its hydraulic proprieties, leading to water erosion and other negative effects on the quality of the environment. This study focused on the effect of compaction on soil hydrodynamic properties under unsaturated and saturated conditions using the Hydraulic Property Analyzer (HYPROP) system. We studied the impact of five levels of compaction among loam sand soils collected in a potato crop field in northern Québec, Canada. Soil samples were collected, and the soil bulk densities of the artificially compacted samples were developed by increasing the bulk density by 0% (C0), 30% (C30), 40% (C40), 50% (C50), and 70% (C70). First, the saturated hydraulic conductivity of each column was measured using the constant-head method. Soil water retention curve (SWRC) dry-end data and unsaturated hydraulic conductivities were obtained via the implementation and evaluation of the HYPROP evaporation measurement system and WP4-T Dew Point PotentioMeter equipment (METER group, Munich, Germany). Second, the soil microporosity was imaged and quantified using the micro-CT-measured pore-size distribution to visualize and quantify soil pore structures. The imaged soil microporosity was related to the saturated hydraulic conductivity, air permeability, porosity and tortuosity measured of the same samples.  Our results supported the application of the Peters–Durner–Iden (PDI) variant of the bimodal unconstrained van Genuchten model (VGm-b-PDI) for complete SWRC estimation based on the root mean square error (RMSE). The unsaturated hydraulic conductivity matched the PDI variant of the unconstrained van-Genuchten model (VGm-PDI) well. Finally, the preliminary results indicated that soil compaction could strongly influence the hydraulic properties of soil in different ways. The saturated conductivity decreased with increasing soil compaction, and the unsaturated hydraulic conductivity changed very rapidly with the ratio of water to soil. Overall, the HYPROP methodology performed extremely well in terms of the hydraulic behavior of compacted soils.

How to cite: Mbarki, Y. and Gumiere, S. J.: Study of the effect of compaction on the hydrodynamic properties of a loamy sand soil for precision agriculture, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1583, https://doi.org/10.5194/egusphere-egu23-1583, 2023.

Scientific publications in space science contain valuable and extensive information regarding the links and relationships between the data interpreted by the authors and the associated observational elements (e.g., instruments or experiments names, observing times, etc.). In this reality of scientific information overload, researchers are often overwhelmed by an enormous and continuously growing number of articles to access in their daily activities. The exploration of recent advances concerning specific topics, methods and techniques, the review and evaluation of research proposals and in general any action that requires a cautious and comprehensive assessment of scientific literature has turned into an extremely complex and time-consuming task.

The availability of Natural Language Processing (NLP) tools able to extract information from scientific unstructured textual contents and to turn it into extremely organized and interconnected knowledge, is fundamental in the framework of the use of scientific information. Exploitation of the knowledge that exists in the scientific publications, necessitates state-of-the-art NLP. The semantic interpretation of the scientific texts can support the development of a varied set of applications such as information retrieval from the texts, linking to existing knowledge repositories, topic classification, semi-automatic assessment of publications and research proposals, tracking of scientific and technological advances, scientific intelligence-assisted reporting, review writing, and question answering.

The main objectives of TACTICIAN are to introduce Artificial Intelligence (AI) techniques to the textual analysis of the publications of all ESA Space Science missions, to monitor and evaluate the scientific productivity of the science missions, and to integrate the scientific publications’ metadata into the ESA Space Science Archive. Through TACTICIAN, we extract lexical, syntactic, and semantic information from the scientific publications by applying NLP and Machine Learning (ML) algorithms and techniques. Utilizing the wealth of publications, we have created valuable scientific language resources, such as labeled datasets and word embeddings, which were used to train Deep Learning models that assist us in most of the language understanding tasks. In the context of TACTICIAN, we have devised methodologies and developed algorithms that can assign scientific publications to the Mars Express, Herschel, and Cluster ESA science missions and identify selected named entities and observations in these scientific publications. We also introduced a new unsupervised ML technique, based on Nonnegative Matrix Factorization (NMF), for classifying the Planck mission scientific publications to categories according to the use of the Planck data products.

These methodologies can be applied to any other mission. The combination of NLP and ML constitutes a general basis, which has proved that it can assist in establishing links between the missions’ observations and the scientific publications and to classify them in categories, with high accuracy.

This work has received funding from the European Space Agency under the "ArTificiAl intelligenCe To lInk publiCations wIth observAtioNs (TACTICIAN)" activity under ESA Contract No 4000128429/19/ES/JD.

How to cite: Giannakis, O., Demiros, I., Koutroumbas, K., Rontogiannis, A., Antonopoulos, V., De Marchi, G., Arviset, C., Balasis, G., Daglis, A., Vasalos, G., Boutsi, Z., Tauber, J., Lopez-Caniego, M., Kidger, M., Masson, A., and Escoubet, P.: TACTICIAN: AI-based applications knowledge extraction from ESA’s mission scientific publications, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1902, https://doi.org/10.5194/egusphere-egu23-1902, 2023.

Identification of constituent grains in carbonate rocks is primarily a qualitative skill requiring specialist experience. A carbonate sedimentologist must be able to distinguish between various grains of different ages, preserved in differing alteration stages, and cut in random orientations across core sections. Recent studies have demonstrated the effectiveness of machine learning in classifying lithofacies from thin section, core and seismic images, with faster analysis times and reduction of natural biases.  In this study, we explore the application and limitations of convolutional neural network (CNN) based object detection frameworks to identify and quantify multiple types of carbonate grains within close-up core images. Nearly 400 images of carbonate cores we compiled of high-resolution core images from three ODP and IODP expeditions. Over 9,000 individual carbonate components of 11 different classes were manually labelled from this dataset. Using transfer learning, we evaluate one-stage (YOLO v3) and two-stage (Faster R-CNN) detectors under different feature extractors (Darknet and Inception-ResNet-v2). Despite the current popularity of one-stage detectors, our results show Faster R-CNN with Inception-ResNet-v2 backbone provides the most robust performance, achieving nearly 0.8 mean average precision (mAP). Furthermore, we extend the approach by deploying the trained model to ODP Leg 194 Sites 1196 and 1190, developing a performance comparison with human interpretation. 

How to cite: Dawson, H. and John, C.: Deep learning based identification of carbonate rock components in core images, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2388, https://doi.org/10.5194/egusphere-egu23-2388, 2023.

Water used in agricultural crops can be managed by irrigation scheduling based on plant water stress thresholds. Automated irrigation scheduling limits crop physiological damage and yield reduction. Knowledge of crop water stress monitoring approaches can be effective in optimizing the use of agricultural water. Understanding the physiological mechanisms of crop responding and adapting to water deficit ensures sustainable agricultural management and food supply. This aim could be achieved by analyzing stomatal conductance, growth rate, leaf water potential, and stem water potential. Calculating thresholds of soil matric potential, and available water content improves the precision of irrigation management by preventing water limitations between irrigations. Crop monitoring and irrigation management make informed decisions using geospatial technologies, the internet of things, big data analysis, and artificial intelligence. Remote sensing (RS) could be applied whenever in situ data are not available. High-resolution crop mapping extracts information through index-based methods fed by the multitemporal and multi-sensor data used in detection and classification. Precision Agriculture (PA) means applying farm inputs at the right amount, at the right time, and in the right place. RS in PA captures images in different spatial, and spectral resolutions through in-field, satellites, aerial, and handheld or tractor-mounted such as unmanned aerial vehicles (UAVs) sensors. RS sensors receive the electromagnetic signals of plant responses in different spectral domains. Optical satellite data, including narrow-band multispectral remote sensing techniques and thermal imagery, is used for water stress detection. To process and analysis RS data, cloud storage and computing platforms simplify the complex mathematical of incorporating various datasets for irrigation scheduling. Machine learning (ML) algorithms construct models for the regression and classification of multivariate and non-linear crop mapping. The web-based software gathered from all different datasets makes a reliable product to reinforce farmers’ ability to make appropriate decisions in irrigating agricultural crops.

Keywords: Agricultural crops; Crop water stress detection; Irrigation scheduling; Precision agriculture; Remote Sensing.

How to cite: Koohi, E., Gumiere, S. J., and Bonakdari, H.: Artificial Intelligence Models for Detecting Spatiotemporal Crop Water Stress in schedule Irrigation: A review, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3997, https://doi.org/10.5194/egusphere-egu23-3997, 2023.

Wetlands provide invaluable services for ecosystems and society and are a crucial instrument in our fight against climate change. Although Earth Observation satellites offer cost-effective and accurate information about wetland status at the continental scale; to date, there is no universally accepted, standardized, and regularly updated inventory of European wetlands <100m resolution. Moreover, previous satellite-based global land cover products seldom account for wetland diversity, which often impairs their mapping performances. Here, we mapped major wetland types (i.e., peatland, marshland, and coastal wetlands) across Europe for 2018, based on high resolution (10m) optical and radar time series satellite data as well as field-collected land cover information (LUCAS) using an ensemble model combining traditional machine learning and deep learning approaches. Our results show with high accuracy (>85%) that a substantial extent of European peatlands was previously classified as grassland and other land cover types. In addition, our map highlights cultivated areas (e.g., river floodplains) that can be potentially rewetted. Such accurate and consistent mapping of different wetland types at a continental scale offers a baseline for future wetland monitoring and trend assessment, supports the detailed reporting of European carbon budgets, and lays down the foundation towards a global wetland inventory.

How to cite: Kovács, G. M., Oehmcke, S., Horion, S., Gominski, D., Tong, X., and Fensholt, R.: Satellite-based continental-scale inventory of European wetland types at 10m spatial resolution, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6696, https://doi.org/10.5194/egusphere-egu23-6696, 2023.

Lagoons are highly valued coastal environments providing unique ecosystem services. However, they are fragile and vulnerable to natural processes and anthropogenic activities. Concurrently, climate change pressures, are likely to lead to severe ecological impacts on lagoon ecosystems. Among these, direct effects are mainly through changes in temperature and associated physico-chemical alterations, whereas indirect ones, mediated through processes such as extreme weather events in the catchment, include the alteration of nutrient loading patterns among others that can, in turn, modify the trophic states leading to depletion or to eutrophication. This phenomenon can lead, under certain circumstances, to harmful algal blooms events, anoxia, and mortality of aquatic flora and fauna, or to the reduction of primary production, with cascading effects on the whole trophic web with dramatic consequences for aquaculture, fishery, and recreational activities. The complexity of eutrophication processes, characterized by compounding and interconnected pressures, highlights the importance of adequate sophisticated methods to estimate future ecological impacts on fragile lagoon environments. In this context, a novel framework combining Machine Learning (ML) and biogeochemical models is proposed, leveraging the potential offered by both approaches to unravel and modelling environmental systems featured by compounding pressures. Multi-Layer Perceptron (MLP) and Random Forest (RF) models are used (trained, validated, and tested) within the Venice Lagoon case study to assimilate historical heterogenous WQ data (i.e., water temperature, salinity, and dissolved oxygen) and spatio-temporal information (i.e., monitoring station location and month), and to predict changes in chlorophyll-a (Chl-a) conditions. Then, projections from the biogeochemical model SHYFEM-BFM for 2049, and 2099 timeframes under RCP 8.5 are integrated to evaluate Chl-a variations under future bio-geochemical conditions forced by climate change projections. Annual and seasonal Chl-a predictions were performed out by classes based on two classification modes established on the descriptive statistics computed on baseline data: i) binary classification of Chl-a values under and over the median value, ii) multi-class classification defined by Chl-a quartiles. Results from the case study showed as the RF successfully classifies Chl-a under the baseline scenario with an overall model accuracy of about 80% for the median classification mode, and 61% for the quartile classification mode. Overall, a decreasing trend for the lowest Chl-a values (below the first quartile, i.e. 0.85 µg/l) can be observed, with an opposite rising fashion for the highest Chl-a values (above the fourth quartile, i.e. 2.78 µg/l). On the seasonal level, summer remains the season with the highest Chl-a values in all scenarios, although in 2099 a strong increase in Chl-a is also expected during the spring one. The proposed novel framework represents a valuable approach to strengthen both eutrophication modelling and scenarios analysis, by placing artificial intelligence-based models alongside biogeochemical models.

How to cite: Zennaro, F., Furlan, E., Melaku Canu, D., Aveytua Alcazar, L., Rosati, G., Aslan, S., Solidoro, C., and Critto, A.: Evaluation of lagoon eutrophication potential under climate change conditions: A novel water quality machine learning and biogeochemical-based framework., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8409, https://doi.org/10.5194/egusphere-egu23-8409, 2023.

Marine coastal ecosystems (MCEs) are of vital importance for human health and well-being. However, their ecological condition is increasingly threatened by multiple risks induced by the complex interplay between endogenic (e.g. coastal development, shipping traffic) and exogenic (e.g. changes in sea surface temperature, waves, sea level, etc.) pressures. Assessing cumulative impacts resulting from this dynamic interplay is a major challenge to achieve Sustainable Development Goals and biodiversity targets, as well as to drive ecosystem-based management in marine coastal areas. To this aim, a Machine Learning model (i.e. Random Forest - RF), integrating heterogenous data on multiple pressures and ecosystems’ health and biodiversity, was developed to support the evaluation of risk scenarios affecting seagrasses condition and their services capacity within the Mediterranean Sea. The RF model was trained, validated and tested by exploiting data collected from different open-source data platforms (e.g. Copernicus Services) for the baseline 2017. Moreover, based on the designed RF model, future scenario analysis was performed by integrating projections from climate numerical models for sea surface temperature and salinity under the 2050 and 2100 timeframes. Particularly, under the baseline scenario, the model performance achieved an overall accuracy of about 82%. Overall, the results of the analysis showed that the ecological condition and services capacity of seagrass meadows (i.e. spatial distribution, Shannon index, carbon sequestration) are mainly threatened by human-related pressures linked to coastal development (e.g. distance from main urban centres), as well as to changes in nutrient concentration and sea surface temperature. This result also emerged from the scenario analysis, highlighting a decrease in seagrass coverage and related services capacity, in both 2050 and 2100 timeframes. The developed model provides useful predictive insight on possible future ecosystem conditions in response to multiple pressures, supporting marine managers and planners towards more effective ecosystem-based adaptation and management measures in MCEs.

How to cite: Bianconi, A., Furlan, E., Simeoni, C., Pham, V., Vascon, S., Critto, A., and Marcomini, A.: Evaluating the risk of cumulative impacts in the Mediterranean Sea using a Random Forest model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8702, https://doi.org/10.5194/egusphere-egu23-8702, 2023.

The standard way for earth observation experts or users to retrieve images from image archives (e.g., ESA's Copernicus Open Access Hub) is to use a graphical user interface, where they can select the geographical area of the image they are interested in and additionally they can specify some other metadata, such as sensing period, satellite platform and cloud cover.

In this work, we are developing the question-answering engine EarthQA that takes as input a question expressed in natural language (English) that asks for satellite images satisfying certain criteria and returns links to such datasets, which can be then downloaded from the CREODIAS cloud platform. To answer user questions, EarthQA queries two interlinked knowledge graphs: a knowledge graph encoding metadata of satellite images from the CREODIAS cloud platform (the SPARQL endpoint of CREODIAS) and the well-known knowledge graph DBpedia. Hence, the questions can refer to image metadata (e.g., satellite platform, sensing period, cloud cover), but also to more generic entities appearing in DBpedia knowledge graph (e.g., lake, Greece). In this way, the users can ask questions like “Find all Sentinel-1 GRD images taken during October 2021 that show large lakes in Greece having an area greater than 100 square kilometers”.

EarthQA follows a template-based approach to translate natural language questions into formal queries (SPARQL). Initially, it decomposes the user question by generating its dependency parse tree and then automatically disambiguates the components appearing in the question to elements of the two knowledge graphs.  In particular, it automatically identifies the spatial or temporal entities (e.g., “Greece”, “October 2021”), concepts (e.g., “lake”), spatial or temporal relations (e.g., “in”, “during”), properties (e.g., “area”) and product types (e.g., “Sentinel-1 GRD”) and other metadata (e.g., “cloud cover below 10%”) mentioned in the question and maps them to the respective elements appearing in the two knowledge graphs (dbr:Greece, dbo:Lake, dbp:area, etc). After this, the SPARQL query is automatically generated.

How to cite: Punjani, D., Tsalapati, E., and Koubarakis, M.: EarthQA: A Question Answering Engine for Earth Observation Data Archives, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10681, https://doi.org/10.5194/egusphere-egu23-10681, 2023.

In the current stage of scientific research, it is necessary to break the barriers between traditional disciplines and promote the cross integration of various related disciplines. As one of the important carriers of research achievements of various disciplines, maps can be superimposed and integrated to more intuitively display the results of multidisciplinary integration, promote the integration of disciplines and discover new scientific problems. Traditional geological mapping is often based on different scales for single scale mapping, aiming at the mapping mode of paper printing results. It is difficult to read maps between different scales at the same time. To solve this problem，an integrated platform named Global Layer is being built under the support of Deep-time Digital Earth (DDE). Global Layer is embedded with several core databases such as Geological Map of the World at a scale 1/5M, Global Geothermal Database etc. These databases presented in form of electronic map which enables the results of different scales to be displayed and browsed through one-stop hierarchical promotion. In addition, Users can also upload data in four ways: local file, database connection, cloud file and arcgis data service, and data or maping results can be shared to Facebook, Twitter and other platforms in the form of links, widgets, etc. Construction of Global Layer could provide experience and foundation for integrating global databases related to geological map and constructing data platforms.

How to cite: Yang, X., Song, Y., Wu, Z., and Wu, C.: Global Layer——An integrated, fully online, cloud based platform, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11527, https://doi.org/10.5194/egusphere-egu23-11527, 2023.

Streams and ditches are seldom identified on current maps due to their small dimensions and sometimes intermittent nature. Estimates point out that only 9% of all ditches are currently mapped, and the underestimation of natural streams is a global issue. Ditches have been dug in European boreal forests and some parts of North America to drain wetlands and increase forest production, consequently boosting the availability of cultivable land and a national-scale landscape modification. Target 6.6 of the Agenda 2030 highlights the importance of protecting and restoring water-related ecosystems. Wetlands are a substantial part of this, having a high carbon storage capability, the property of mitigating floods, and purifying water. All things accounted for, the withdrawal of anthropogenic environment alterations can be on the horizon, even more because ditches are also strong emitters of methane and other greenhouse gases due to their anoxic water and sediment accumulation. However, streams and ditches that are missing from maps and databases are difficult to manage.

The main focus of this study was to develop a method to map channels combining deep learning and national Aerial Laser Scans (ALS). The performance of different topographical indices derived from the ALS data was evaluated, and two different Digital Elevation Model (DEM) resolutions were compared. Ditch channels and natural streams were manually digitized from ten regions across Sweden, summing up to 1923km of ditch channels and 248km of natural streams. The topographical indices used were: high-passing median filter, slope, sky-view factor and hillshade (with azimuths of 0°, 45°, 90° and 135°); while 0.5m and 1m were the DEM resolutions analysed. A U-net model was trained to segment images between ditches and stream channels: all pixels from each image were labelled in a way that those with the same class display similar attributes.

Results showed that ditches can be successfully mapped with this method and it can generally be applied anywhere since only local terrain indices are required. Additionally, when the natural streams are present in the dataset the model underperformed in predicting the location of ditches, while a higher resolution had the opposite effect. Streams were more challenging to map, and the model only indicated the channels, not whether or not they contained water. Further research will be required to combine hydrological modelling and deep learning.

How to cite: Dos Santos Toledo Busarello, M., Ågren, A., and Lidberg, W.: Mapping streams and ditches using Aerial Laser Scanning, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12373, https://doi.org/10.5194/egusphere-egu23-12373, 2023.

There is a soaring demand for up-to-date and spatially-explicit soil information to address various environmental challenges. One of the most basic pieces of information, essential for research and decision-making in multiple disciplines is soil classification. Conventional soil maps are often low in spatial resolution and lack the complexity to be practical for hands-on use. Digital Soil Mapping (DSM) has emerged as an efficient alternative for its reproducibility, updatablity, accuracy, and cost-effectiveness, as well as the ability to quantify uncertainties.

Despite DSM’s growing popularity and increasingly wider areas of application, soil information is still rare in forested areas and remote regions, and the integration with high-resolution data on a country scale remains limited. In Sweden, quaternary deposit maps created by the Geological Survey of Sweden (SGU) have been the main reference input for soil-related research and operation, though most parts of the country still warrant higher quality representation. This study utilizes machine learning to produce a high-resolution surficial deposits map with nationwide coverage, capable of supporting research and decision-making. More specifically, it: i) compares the performance of two tree-based ensemble machine learning models, Extreme Gradient Boosting and Random Forest, in predictive mapping of soils across the entire country of Sweden; ii) determines the best model for spatial prediction of soil classes and estimates the associated uncertainty of the inferred map; iii) discusses the advantages and limitations of this approach, and iv) outputs a map product of soil classes at 2-m resolution. Similar attempts around the globe have shown promising results, though at coarser resolutions and/or of smaller geographical extent. The main assumptions behind this study are: i) terrain indices derived from digital elevation model (DEM) are useful predictors of soil type, though different classification algorithms differ in their effectiveness; ii) machine learning can capture major soil classes that cover most of Sweden, but expert geological and pedological knowledge is required when identifying rare soil types.

To achieve this, approximately 850,000 labeled soil points extracted from the most accurate SGU maps will be combined with a stack of 12 LiDAR DEM-derived topographic and hydrological indices and 4 environmental datasets. Uncertainty estimates of the overall model and for each soil class will be presented. An independent dataset obtained from the Swedish National Forest Soil Inventory will be used to assess the accuracy of the machine learning model. The presentation will cover the method, data handling, and some promising preliminary results.

How to cite: Lin, Y., Lidberg, W., Karlsson, C., and Ågren, A.: Mapping Swedish Soils with High-resolution DEM-derived Indices and Machine Learning, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13099, https://doi.org/10.5194/egusphere-egu23-13099, 2023.

As human activities continue to expand and evolve, their impact on the planet is becoming more evident. These past years Murmuration has been studying one of the most recent and destructive trends that has taken off: mass tourism. In Malta, tourism has been on the rise since before the Covid-19 pandemic. Now that travel restrictions are beginning to lift, it's likely that this trend will go back to increasing in the coming years. While Malta’s economy is mostly based on tourism, it's essential that this activity does not alter the areas in which it takes place. To address these issues and ensure sustainable development, governments and organizations have developed a set of guidelines called Sustainable Development Goals (SDG). SDGs are a set of 17 goals adopted by the United Nations in 2015 to provide a framework to help countries pursue sustainable economic, social and environmental development. They include objectives for mitigating climate change, preventing water pollution and degradation of biodiversity, as well as providing economic benefits to local communities.

In order to help territories like the islands of Malta to cope with these environmental issues, Murmuration carries out studies on various ecological, human and economic indicators. Using the Sentinel satellites of the European Copernicus program for earth imagery data makes possible the collection of geolocated, hourly values on air quality indicators such as NO2, CO and other pollutants but also water quality and vegetation through the analysis of the vegetation health. Other data sources give access to land cover values at meter resolution, tourism infrastructures locations and many more human activity variables. This information is processed into understandable indicators, aggregated indexes which take international standards and SDGs in their design and usage. An example of these standards are the WHO air quality guidelines providing thresholds quantifying the impact on health of the air pollution in the area of interest. The last step is to gather all the data, maps and correlations computed and design understandable visualizations to make it usable by territory management instances, enabling efficient decision making and risk management. The goal here is to achieve a link between satellite imagery, internationally agreed political commitment  and ground level decision-making.

This meaningful aggregation comes in the shape of operational dashboards. A dashboard is an up-to-date, interactive, evolving online tool hosting temporal and geographical linked visualizations on various indicators. This kind of tool allows for a better understanding of the dynamic of a territory in terms of environmental state, human impact and ecological potential.

How to cite: Plantec, M. and Castel, F.: From satellite data and Sustainable Development Goals to interactive tools and better territorial decision making, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14519, https://doi.org/10.5194/egusphere-egu23-14519, 2023.

Karst aquifers cover a significant portion of the global water supply. However, a proper representation of groundwater recharge in karst areas is completely absent in the state-of-art global hydrological models. This study, based on the new version of the global hydrological model WaterGAP, (1) presented the first modeling of diffuse groundwater recharge (GWR) in all karst regions using the global map of karstifiable rocks; and (2) adjusted the current GWR algorithm with the up-to-date databases of slope and soil. A large number of ground-based recharge estimates on 818 half degree cells including 75 in karst areas were compared to model results. GWR in karst landscapes assuming equal to the runoff from soil leads to unbiased estimation. The majority of simulated mean annual recharge ranges from 0.6 mm/yr (10^th percentile) to 326.9 mm/yr (90^th) in nonkarst regions, and 7.5 mm/yr (10^th) to 740.2 mm/yr (90^th) in karst regions. The recharge rate ranges from 2% to 66% of precipitation according to ground-based estimates in karst regions, while the simulated GWR produces global recharge fractions between 4% (10^th) to 68% (90^th) in karst areas while that in nonkarst areas rarely exceeds 25%. Unlike the previous studies that claimed global hydrological models consistently underestimate recharge, we observed underestimation only in the very humid regions where recharge exceeds 300 mm/yr. These very high recharge estimates are likely to include preferential flow and adopt a finer spatial and temporal scale than the global model. In karst landscapes and arid regions, we demonstrate that WaterGAP incorporating karst algorithm gives a worthy performance.

How to cite: Wan, W. and Döll, P.: Karst integration into groundwater recharge simulation in WaterGAP, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15656, https://doi.org/10.5194/egusphere-egu23-15656, 2023.

In Marine Geology, scientists persistently perform extensive experiments to measure diverse features across the globe, hence to estimate environmental changes. For example, Mass Accumulation Rate (MAR) and Sedimentation Rate (SR) are measured by marine geologists at various oceanographic locations and are largely reported in research publications but have not been compiled in any central database. Furthermore, every MAR and SR observation normally carries i) exact locational information (Longitude and Latitude), ii) the method of measurement (stratigraphy, 210Pb), iii) a numerical value and units (2.4 g/m²/yr), iv) temporal feature (e.g. hundred years ago). The contextual information attached to MAR and SR observations is heterogeneous and manual approaches for information extraction from text are infeasible. It is also worth mentioning that MAR and SR are not denoted in standard international (SI) units.

We propose the comprehensive end-to-end framework GEOTEK (Geological Text to Knowledge) to extract targeted information from marine geology publications. The proposed framework comprises three modules. The first module carries a document relevance model alongside a PDF extractor, capable of filtering relevant sources using metadata, and the extraction module extracts text, tables, and metadata respectively. The second module mainly comprises of two information extractors, namely Geo-Quantities and Geo-Spacy, particularly trained on text from the Marine Geology domain. Geo-Quantities is capable of extracting relevant numerical information from the text and covers more than 100 unit variants for MAR and SR, while Geo-Spacy extracts a set of relevant named entities as well as locational entities, which are further processed to obtain respective geocode boundaries. The third module, the Heterogeneous Information Linking module (HIL), processes exact spatial information from tables and captions and forms links to the previously extracted measurements. Finally, the all-linked information is populated in an interactive map view.

How to cite: Suryani, M. A., Beth, C., Wallmann, K., and Renz, M.: GEOTEK: Extracting Marine Geological Data from Publications, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16252, https://doi.org/10.5194/egusphere-egu23-16252, 2023.

Climate change is arguably the greatest environmental challenge facing humankind in the twenty-first century. The United Nations Framework Convention on Climate Change (UNFCCC) facilitates multilateral action to combat climate change and its impacts on humanity and ecosystems. To make decisions on climate change mitigation and adaptation, the UNFCCC requires systematic observations of the global climate system.

The objective of the ESA’s climate programme, currently delivered via the Climate Change Initiative (CCI), is to realise the full potential of the long-term, global-scale, satellite earth observation archive that ESA and its Member States have established over the last 35 years, as a significant and timely contribution to the climate data record required by the UNFCCC.

Since 2010, the programme has contributed to a rapidly expanding body of scientific knowledge on >22 Essential Climate Variables (ECVs), through the production of Climate Data Records (CDRs). Although varying across geophysical parameters, ESA CDRs follow community-driven data standards, facilitating inter- and cross-ECV research of the climate system.

In this work, we highlight the use of artificial intelligence (AI) in the context of the ESA CCI. AI has played a pivotal role in the production and analysis of these Climate Data Records. Eleven CCI projects - Greenhouse Gases (GHG), Aerosols, Clouds, Fire, Ocean Colour, Sea Level, Soil Moisture, High Resolution Landcover, Biomass, Permafrost, and Sea Surface Salinity - have applied AI in their data record production and research or have identified specific AI usage for their research roadmaps.

The use of AI in these CCI projects is varied, for example - GHG CCI algorithms using random forest machine learning techniques; Aerosol CCI algorithms to retrieve dust aerosol optical depth from thermal infrared spectra; Fire CCI algorithms to detect burned areas. Moreover, the ESA climate community has identified climate science gaps in context to ECVs with the potential for meaningful advancement through AI.

We specifically focus on showcasing the use of AI for data homogenization and super-resolution of ESA CCI datasets. For instance, both the land cover and fire CCI dataset were generated globally in low resolution, while high resolution data only exists for specific geographical regions. By adapting super-resolution algorithms to the specific science use cases, we can accelerate the generation of global, high-resolution datasets with the required temporal coverage to support long-term climate studies. 

How to cite: Jungbluth, A., Pechorro, E., Albergel, C., and Mecklenburg, S.: The Use of Artificial Intelligence in ESA’s Climate Change Initiative, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16813, https://doi.org/10.5194/egusphere-egu23-16813, 2023.

Coastal ecosystems provide vital services, including water storage, carbon sequestration, biodiversity, and coastal protection. Human disturbances, however, cause massive losses. The most direct impact is habitat destruction through infrastructure development, restricting the space available to coastal ecosystems and impeding their capacity to adapt to sea level rise by landward retreat – a phenomenon called ‘coastal squeeze’. While shoreline retreat is intensively studied, coastal congestion through infrastructure remains unquantified. Here we calculated the distance to the nearest human-made structure along 263,900 transects worldwide to show that infrastructure occurs at a 560-meter median distance from the shoreline. Moreover, we find that 18% of global sandy shores harbour less than 100 m of infrastructure-free space, and that 14-17% of the unimpacted space may drown by 2100 according to sea level rise projections. Further analyses show that population density and gross domestic product explain 40-44% of observed squeeze variation, emphasizing the intensifying pressure imposed as countries develop and populations expand. Encouragingly, we find that nature reserves relieve squeezing by 3-5 times, illustrating their effectiveness. Yet, at present only 16% of world’s sandy shores has a protected status. We therefore argue that expansion of nature reserves could be key to preserving coastal resilience to sea level rise.

How to cite: Lansu, E., Reijers, V., Höfer, S., Luijendijk, A., Rietkerk, M., Wassen, M., Lammerts, E. J., and van der Heide, T.: A global analysis of how human infrastructure squeezes sandy coasts, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-67, https://doi.org/10.5194/egusphere-egu23-67, 2023.

Coastal ecosystems are vulnerable to climate change, with rising sea levels and increased anthropogenic pressure constraining space for natural processes. Nature based solutions using sediments rather than hard surfaces in coastal defense may offer an alternative that both creates new habitats and offer a flexible protection against flooding.

In contrast to hard infrastructure, the topography of dunes depends on the highly dynamic processes of wind and waves and the resistance to them offered by dune vegetation. Perennial grass species such as marram grass (Ammophila arenaria) and sand couch (Elytrigia juncea) play a key-role for topographic stability and the development and shape of coastal dune forms. This is usually attributed to their dense cover which effectively traps sand as well as their positive growth response to burial by sediments. Therefore, species like marram grass have been used as ecosystem engineers in both past and recent coastal dune restoration projects.

Whether this solution will be applicable in the future depends on climate change. Coastal vegetation is vulnerable to climate change due to its susceptibility to changes in growing conditions (e.g. Temperature, Precipitation). Especially at the dry-beach section where the influence of groundwater is limited, a change in growing season precipitation could potentially affect the cover of dune grasses. Past research was already able to establish a general link between dune development and growth in function of precipitation. However, to this date direct responses of dune vegetation to precipitation has not been quantified.

We explored the response of dune building grasses to summer precipitation and its implication for the future dune building in a two-step approach. We used a greenhouse-experiment to derive species growth relationships with water availability for marram grass and sand couch. In a second step we used these relationships to explore the impact of potential changes in summer precipitation on the growth of these species. We found that both marram grass and sand couch were equally sensitive to changes in water availability and responded positively to an increase in it. Comparing soil moisture from the field to the greenhouse, showed that field water availability tended to be on the lower end of ranges in the greenhouse. This suggests that dune vegetation in the field is susceptible to drought effects. Exploring these results further using climate scenarios, we found that plant growth was increased by 1.3 % (experimental period) – 1.8 % (extrapolated) under the most recent RCP 4.5 IPCC projection and by 9.6 – 13 % for an extremely wet year. In contrast, for an extremely dry year plant growth could decrease by 6.2 – 8.2 %.

While changes of < 2 % in plant growth might have limited implications for dune development and stability, years of extreme climate conditions show a bigger range in plant growth (- 8 % - + 13 %) which is more likely to also have direct consequences for dune growth and development. Incorporating these relationships between plant growth and climate in models of coastal dune development should improve predictions of climate change impacts.

How to cite: Homberger, J.-M., Lynch, A., Limpens, J., and Riksen, M.: Response of dune-building grasses to summer precipitation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1076, https://doi.org/10.5194/egusphere-egu23-1076, 2023.

Around the Arctic Ocean there are many stretches of coastline composed of ice-rich sediments. With the dramatic climatic, oceanic and terrestrial changes that are currently underway, there is considerable concern over the stability of these coasts and how they impact coastal communities. Unfortunately, there is still relatively little research that has been done the processes at work in these environments. Being able to effectively model coastal erosion in a permafrost setting is highly desirable. With the complexity that ice-rich permafrost conditions add to the coastal setting, modelling erosion involves a more detailed understanding of the physical and thermal conditions as well as the sedimentological and wave action processes. This research examines the rate and character of coastal erosion of ice-rich terrain and role that re-sedimentation has on the shallow water energy balance in preserving sub-bottom massive ice. It also addresses it implications to secondary sea bottom disturbance as the water depth increases.

The study area was Peninsula Point which is approximately 10 km west of Tuktoyaktuk, Northwest Territories, Canada. The massive ice and retrogressive thaw flows at this location are some of the more dramatic examples of the impact of ice-rich permafrost on coastal processes in the Arctic. Through a three decade long program of remote sensing, geophysical and ground-based monitoring, long-term changes were investigated. The character of coastal retreat above, and below, the waterline in an area where a massive ice body extends to depths below sea level were revealed. Airphoto, satellite imagery and drone data revealed the complexity of erosion with the retreating headwall of retrogressive thaw flow more rapidly eroding the landscape than the observed lateral changes of the waterline. Ground-penetrating radar (GPR) imaged the top and base of the massive ice body as well as providing a delineation of the subsurface sedimentary architecture. In winter, the GPR was pulled behind a snowmobile along transects on land, across the shoreline and out onto the near shore area of the Beaufort Sea. This provided the stratigraphic continuity between the terrestrial and sub-sea settings. The roles of erosion, re-sedimentation and shallow-water thermodynamics in the degradation and preservation of massive ground ice were revealed. The results of this study demonstrate how coastal erosion is much more complex that just the inland movement of the waterline.

How to cite: Moorman, B., Clark, A., and Whalen, D.: Ice-rich permafrost coastline erosion processes, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1493, https://doi.org/10.5194/egusphere-egu23-1493, 2023.

How to cite: van Rosmalen, S., Homberger, J.-M., Riksen, M., and Limpens, J.: Recreation impact on establishment of dune building species, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2771, https://doi.org/10.5194/egusphere-egu23-2771, 2023.

In front of the traditional dike at Oosteroever, Belgium, a new 120x20 m² artificial dune with planted marram grass of different densities and patterns was built in January 2021. This man-made dune was constructed to reduce the local aeolian sand nuisance on the dike. The complex interaction between aeolian sand transport and vegetation will ensure future morphological development of a dune body strengthening the local coastal protection. For an optimal design of these planted dunes, a fundamental knowledge of morphological changes is required. This study is twofold: 1) Investigate dune growth by exploring a multi-monthly field dataset of wind characteristics and high-resolution topographic data, 2) Development and assessment of the AeoLiS model for simulation of this new planted coastal dune field. The performance of AeoLiS is analyzed by comparing observed and simulated results of erosion and deposition patterns and cross-shore bed level changes. The total volume of sand in the dune has increased significantly since the plantation of marram grass, resulting in 15 m³ m^-1 due to aeolian sand transport from the surrounding beach. Most dune growth occurred during the first year. Dune growth during the second year was less pronounced and was attributed to the influence of supply limitations, vegetation characteristics, and sediment erosion by wind and storm events. The results of the model simulations demonstrate that AeoLiS can replicate the spatial patterns and profile development inside the artificial dune area to some extent. However, to adequately account for the interaction between vegetation and aeolian sand transport, the model's treatment of vegetation dynamics needs to be improved.

How to cite: Strypsteen, G. and Rauwoens, P.: Morphological changes in a planted coastal dune field: measurements and modelling, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4119, https://doi.org/10.5194/egusphere-egu23-4119, 2023.

How to cite: Choi, K., Kim, D., Jo, J., Sohn, S., and Nam, S.-I.: Morphodynamic evolution of paraglacial spit complexes on a tide-influenced Arctic fjord delta (Dicksonfjorden, Svalbard), EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4579, https://doi.org/10.5194/egusphere-egu23-4579, 2023.

Located in north-eastern Canada near Shippagan, New Brunswick, the Shippagan Gully Conservation Offsetting Project is leveraging salt marsh creation and sand motor techniques to create Piping Plover habitat while increasing resiliency of the Chaisson Office Spit and surrounding communities to climate change. The spit has been altered by more than a century of human activity and is increasingly impacted by climate change and sea-level rise. The project, which employs a holistic approach to improve marine navigation through the Gully and install nature-based solutions for coastal protection and habitat creation, is the first sand motor in Atlantic Canada and the most northern created marsh with sill to date. Extensive modeling was undertaken by NRC prior to the commencement of baseline data collection and design in 2017. Several monitoring and research initiatives are associated with the project, including a fifteen-year monitoring program (regulatory requirement), five-year post-graduate scientific research program, and a 3-year research project which will augment and build on the NRC-led Nature-Based Infrastructure for Coastal Resilience project. Construction began on the sand motor in 2020, with the marsh and marsh sill scheduled to be built in winter 2023 from on-site materials and planted in  spring 2023. The final stages of the implementation will include dune and wetland restoration following the removal of old infrastructure, returning nearly the entire spit to a more natural state and restoring natural processes. The first two years of monitoring following the sand motor implementation have shown a shift in conditions to those more closely matching a nearby control site, as well as the first successful nesting and fledging of Piping Plover (Federally Endangered Species) on the site in over 20 years. The project is the result of a collaborative effort that includes federal and provincial government departments, private industry, academia, and environmental NGOs.

How to cite: Graham, J., vanProosdij, D., Ellis, K., Bowron, T., and Thomas, J.: Restoring Piping Plover Habitat and Building Coastal Resilience with Nature-based Solutions in Atlantic Canada, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4673, https://doi.org/10.5194/egusphere-egu23-4673, 2023.

The Bering and Chukchi Seas are important oceanic regions of carbon sequestration, owing to enhanced gas solubility in cold surface waters and the rapid uptake of carbon dioxide (CO₂) during intense spring blooms. The biogeochemical impacts of decreasing sea ice extent and earlier onset of spring ice melt in this region are yet uncertain. As these marginal seas of the western Arctic Ocean are quite shallow, mostly <60m depth, there is extensive interaction across air-sea-sediment boundaries, but the transformations and fluxes of inorganic carbon in Bering and Chukchi Sea sediments have not been directly quantified. In May-June 2021, we collected water column samples at 14 stations and sediment cores at 5 stations spanning the eastern Bering Sea and southern and eastern Chukchi Sea. Duplicate cores were incubated for several days at in situ temperature, and core-top water was sampled to estimate inorganic carbon and nutrient fluxes. The stations spanned a range of surface ice coverage history, from greater than one month to less than one day of ice-free conditions. In the Chukchi Sea, salinity-normalized bottom water nutrient and dissolved inorganic carbon (DIC) concentrations increased northward, indicating a net input of remineralization products, although effluxes of these parameters from the sediments decreased northward. Moving northward in the Chukchi Sea, the surface water had greater sea ice concentrations, inhibiting surface productivity and air-sea exchange. This may have reduced the rain of labile carbon to the seafloor, resulting in the decreased benthic remineralization. The combination of increasing northward ice coverage and the northward flow of nutrient and IC-rich Pacific-sourced waters influences the bottom-water concentration of remineralization products and sediment-water fluxes. We expect our northeastern Chukchi Sea flux observations are representative of baseline low wintertime sediment-water flux conditions, while the more southerly stations represent at least one month post-ice melt benthic fluxes when surface water productivity is high and the air-water-sediment system openly interacts. We note that some duplicate core measurements were highly heterogeneous, especially in the Bering Sea, illustrating the dynamic nature of this macrofauna-dominated benthic environment and the range of possible fluxes under different rates of bioturbation. While these observations may serve as a seasonal reference, they may also demonstrate how sedimentary fluxes will evolve under future conditions that are expected when sea ice retreats earlier in the season. Here we present our sediment-water flux and water column DIC and nutrient measurements and place them in context with previous work in the region.

How to cite: Barrett, L., Vlahos, P., and Hammond, D.: Early melt-season nutrient and inorganic carbon sediment-water fluxes in the Bering and Chukchi Seas, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7877, https://doi.org/10.5194/egusphere-egu23-7877, 2023.

While most global coasts suffer from a loss of landmass due to sea-level rise and coastal transgression, the Arctic and Sub-Arctic coastlines of Hudson Bay and James Bay witness a reverse phenomenon due to post-glacial rebound. The carbon-rich peatlands Hudson Bay Lowland, that emerged from the retraction of the Tyrell Sea, are witnessing the highest rate of vertical upliftment on the planet. The continual reshaping of the coastline by multiple physical forcings is readily visible by the contiguous and recurrent pattern of raised beach ridges imprinted on the rising land far from the present-day coastline. These beach ridges, formed through the interplay of coastal sea ice dynamics and then preserved above sea-level by uplift, hold back terrestrial runoff and are thus critical to the extensive wetland-saltmarsh ecosystems that provide important habitats for waterfowl and wildlife. This study examines the intricate process behind the formation and modification of these geomorphological units using remote sensing techniques. The study includes the use of various remote sensing products to determine ice duration (Canadian Ice Service- Ice Charts), change detection of ridge dimensions and vectors (Landsat Images), elevation (SRTM and ICESat-2), rate of vertical upliftment (glacial isostasy models) and ice motion in the nearshore zone (GOES). Remote sensing observations reveal that the beach ridges originate offshore on mudflats due to ice scouring and gradually, pushed by sea ice, move shoreward, and often merge and build up existing ridges but sometimes initiating a new beach ridge sequence. The current study documents the impact of changing ice regime on the landward movement of beach ridges on the tidal flats. We find that the seaward point of origin on the tidal flats, and the rate at which the ridges expand and finally merge with the coastline vary greatly across the coastline. The slope of the coast and the dynamics of the sea ice in the nearshore zone are key factors leading to this variability.

How to cite: Ehn, J., Gupta, K., Reeves, A., and Mukhodpadhyay, A.: Beach ridge formation and landward migration along the isostatically rising coastlines of Hudson Bay, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9832, https://doi.org/10.5194/egusphere-egu23-9832, 2023.

The Arctic coast is facing rapid, irreversible changes mainly caused by Climate Warming, e.g., melting sea ice, permafrost thaw, glacial retreat, land uplift and sea level rise. These processes are leading to fundamental changes in the ecosystem structure and functioning, negatively impacting biological and human communities. Under this complex setting, more knowledge is needed to identify the hotspots of shoreline displacement at an Arctic scale. Thus, the goal of this study was to develop and describe a procedure for mapping long-term shoreline displacement in the Arctic that can provide local communities and environmental managers better opportunities to adapt to further coastal changes. Therefore, the procedure will need to be transferrable to diverse environments and able to handle pan-Arctic analyses at a 30-meter spatial resolution. In this study, the procedure was developed using two test areas: Tanafjorden in the low Arctic mainland Norway and Kongsfjorden in the high Arctic Svalbard. The presentation introduces the final procedure and validation results, and discusses its applicability to pan-Arctic shoreline displacement analyses.

The procedure was calibrated in the surroundings of Tanafjorden. It was built on a 40-year time-series of open Landsat and Sentinel multispectral satellite images, taken during the Arctic summer. Supervised random forest classification was used to identify land and water pixels, utilizing information from multiple infrared bands and spectral indices. Mountain shadow pixels were treated as their own class and then merged to the land class. Open spatial data were used for limiting the area-of-interest and for automated creation of training data. In total over 700 individual images were first classified separately to account for local environmental conditions and transient illumination conditions. Images were then summarized over 5-year time-steps. The classification results were examined against an independent validation dataset of 2000 land cover observations and manually digitized shoreline, and the supervised classification results were compared to single-band classifications based on Otsu’s thresholding. The final procedure was then validated in the Kongsfjorden environment. The process was built on Google Earth Engine’s image collections and cloud computing infrastructure to minimize computing times.

The results indicate that it is possible to transform open satellite imagery into 40-year pan-Arctic shoreline displacement information, with a 30-meter resolution and an overall accuracy of more than 95 %. Data fusion is needed in most processing steps: to limit the area-of-interest, save computing power and reduce errors, provide information that complements multispectral satellite data and reduce the impact of short-term atmospheric and water-level effects. Summarizing dozens of images efficiently removes data gaps and the impact of noise, but this efficiency is sensitive to the number of summarized images. The single-image classification approach is flexible and seems to make the procedure transferable to different locations. Cloud image collections are needed to remove the bottleneck of reading and writing satellite data, and potentially allows the promising procedure to be applied at a pan-Arctic scale in the future.

How to cite: Nylén, T., Gonzales-Inca, C., and Calle Navarro, M.: Procedure for examining long-term Arctic shoreline displacement from multispectral satellite data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11651, https://doi.org/10.5194/egusphere-egu23-11651, 2023.

The consequences of accelerating climatic warming on Arctic landscape evolution are far-reaching. In Svalbard, glaciers are rapidly retreating after the Little Ice Age, which leads to exposing new coastal landscapes from marine-terminating glaciers. Precise quantification of these changes was limited until the complete dataset of Svalbard glacier outlines from 1930’s was made available. Here, we analyse the new Svalbard glacier change inventory data and demonstrate that glacier retreat was responsible for a major shift from marine-terminating towards land-terminating glaciers in the last century. This retreat also led to the formation of 922.9 km of new coastline since 1930’s (representing increase of 16.37% in coastline length) creating pristine landscapes governed by paraglacial processes and sediment-rich nearshore fjord environments. Recent palaeogeographical reconstructions suggest that such a mode of coastal evolution was dominant over the extended periods of the Holocene. Transitions from marine-terminating to land-based glaciers had significant implications for fjord circulation, biological production, state of marine ecosystems, biogeochemical cycles between land and seas, and CO₂ budget in coastal waters. Still ongoing climate warming with associated further glacier retreat may lead to more coasts to be exposed in the future. Moreover, glacier retreat will likely cause collapse of Hornbreen-Hambergbreen glacier bridge leading to separation of Sørkappland and rest of Spitsbergen with severe consequences for regional ocean circulation and climate dynamics.

New bays, new straits, new peninsulas and new islands, that have appeared in the last decades of unprecedented warming and associated decay of marine-terminating glaciers in the Arctic are predominantely uncharted and unexplored territories which foreshadow ice-free Arctic and other cold regions of the warmer future. The importance of transdisciplinary research exploring those deglaciated oases has never been more important than at present.

Acknowledgement: The research leading to these results has received funding from the Norwegian Financial Mechanism  2014-2021: SVELTA - Svalbard Delta Systems Under Warming Climate (UMO-2020/37/K/ST10/02852) based at the University of Wroclaw.

How to cite: Kavan, J. and Strzelecki, M.: Glacier decay boosts formation of new Arctic coastal environments – lessons learned from Svalbard, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12412, https://doi.org/10.5194/egusphere-egu23-12412, 2023.

Increasing water levels at the shore can cause coastal erosion, wave overtopping and flooding that threaten communities and infrastructure. More frequent, longer and more severe storm events observed in the North Atlantic sector of the Arctic bring more energetic waves to beaches of western Svalbard. Decreasing extent and duration of the sea ice cover increases potential fetch which makes the waves higher and longer. At the shore, the number of ice-free days per year has increased and coasts that were protected from waves by ice are becoming exposed perennially or over longer time. Modelling suggests that in future the sea ice will continue to decrease while the storminess will further increase. Better understanding the role of sea ice conditions and nearshore wave transformations on wave energy at the Arctic shores is needed to predict coastal hazards under changing climate.

In this study we focus on wave energy delivery to the shores of Hornsund, a ~300 km² fjord of south-western Spitsbergen, Svalbard, and particularly to Isbjornhamna bay in northern Hornsund, where the Polish Polar Station infrastructure is located. We monitor continuously nearshore wind wave conditions and the state of the shore ice, and seasonally the wave run-up and beach morphology. We use three nested SWAN (Simulating WAves Nearshore) models that take low-resolution global wind and wave models and nearshore bathymetry to reconstruct wind wave conditions in the nearshore (~15 m depth) Isbjornhamna. Finally, we use Sentinel-1 SAR data to reconstruct sea ice conditions in Hornsund area which need be incorporated into the wave model. Here we show how our monitoring and modelling scheme facilitates the comprehensive understanding of the nearshore and coastal processes in Isbjornhamna.

How to cite: Swirad, Z., Moskalik, M., Herman, A., Johansson, M., and Rees, G.: Sea ice, wind waves and coastal erosion in Hornsund, Svalbard, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13052, https://doi.org/10.5194/egusphere-egu23-13052, 2023.

Rapid and unprecedented warming of high latitudes exposes Arctic coastal communities to greater vulnerability as they observe their territory changing through general permafrost degradation, episodes of flooding and accelerated coastal erosion threatening their infrastructure and livelihood. Local information is known for infrastructures mapping and coastal changes but consistency in the measurement is lacking as well as spatial coverage for large coastal areas. The need of a consistent circumpolar dataset is primordial in order to map risks and mitigate impacts for arctic coastal communities. Machine learning methods with Sentinel 1/2 imagery allow the circumpolar mapping of arctic coastal settlements (Bartsch et al. 2021a). Validation of recent updates are supported by high-resolution data from the Pleiades satellites, aerial and drone imagery. 

This study is part of the ESA EO4PAC project which aims to provide a range of satellite derived information, including coastal erosion/accretion and infrastructure in the proximity, for the next generation of the Arctic Coastal Dynamic Database (ACD; Lantuit, et al. 2012).  Previous results highlight the detection of 50% more human presence information than in OpenStreetMap especially in Russia with recent expansion of infrastructures related to expanding oil and gas industry. Recent updates of the SACHI dataset (Bartsch et al. 2021b) will be presented including additional attributes for roads and their validation. A preliminary categorization of settlements with respect to permafrost degradation (based on Permafrost_cci records) and coastal erosion based on the current ACD will be presented.

Bartsch, A., G. Pointner, I. Nitze, A. Efimova, D. Jakober, S. Ley, E. Högström, G. Grosse, P. Schweitzer (2021a): Expanding infrastructure and growing anthropogenic impacts along Arctic coasts. Environmental Research Letters. https://doi.org/10.1088/1748-9326/ac3176

Bartsch, Annett, Pointner, Georg, & Nitze, Ingmar. (2021b). Sentinel-1/2 derived Arctic Coastal Human Impact dataset (SACHI) (Version 1) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.4925911

Lantuit, Hugues; Overduin, Pier Paul; Couture, Nicole; Wetterich, Sebastian; Are, Felix; Atkinson, David; Brown, Jerry; Cherkashov, Georgy A; Drozdov, Dimitry S; Forbes, Donald Lawrence; Graves-Gaylord, Allison; Grigoriev, Mikhail N; Hubberten, Hans-Wolfgang; Jordan, James; Jorgenson, M Torre; Ødegård, Rune Strand; Ogorodov, Stanislav; Pollard, Wayne H; Rachold, Volker; Sedenko, Sergey; Solomon, Steve; Steenhuisen, Frits; Streletskaya, Irina; Vasiliev, Alexander (2012): The Arctic Coastal Dynamics Database: A New Classification Scheme and Statistics on Arctic Permafrost Coastlines. Estuaries and Coasts, 35(2), 383-400, https://doi.org/10.1007/s12237-010-9362-6

How to cite: Tanguy, R., Bartsch, A., Widhalm, B., von Baeckmann, C., Efimova, A., and Vieira, G.: Pan-Arctic remotely sensed observation of coastal settlements - recent updates, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14578, https://doi.org/10.5194/egusphere-egu23-14578, 2023.

Under current conditions, the Spanish Mediterranean coast is already presenting hotspots of extreme exposure to coastal hazards and recurrent damage, making it necessary to adopt disruptive adaptation strategies as opposed to the classic expectation of full protection. This situation is expected to worsen under the effect of sea level rise, which will increase existing erosion rates, with some areas being fully eroded due to the lack of accommodation space to allow natural adaptation to the new conditions.

In this context, nature-based solutions (NBS) are becoming one of the main type of measures to be favored in order to be more climate-resilient and thus support EU policy priorities. Although research on the effectiveness of most nature-based coastal protection methods is still limited, some of them such as dune systems and sand banks have been classified as essential for future coastal defense.

In highly-developed coastal zones, which are the most at risk, the lack of the sufficient space limits the viability of using NBS as they cannot be accommodated. Thus, the existence of accommodation space is the required condition to permit the beach migration and rebuilding under SLR, otherwise will progressively decline and ultimately disappear. It has to be stressed that the accommodation space is a relative concept, being related to the expected magnitude of the shoreline retreat at a given time horizon under a given climate forcing scenario. 

Within this context, this work presents a regional-scale framework to assess the accommodation space needed to adopt dune-based NBS planning as a coastal adaptation strategy, by integrating predictions of accommodation space needed to cope with coastal hazards under current and IPCC AR6 climate scenarios and for different time horizons relevant for planning purposes (up to 2100), and to enable dune development. The hazards considered are (i) long-term (decadal scale) coastline evolution; (ii) storm-induced erosion; (iii) SLR-induced erosion; (iv) permanent inundation due to SLR; and (v) storm-induced flooding. The framework applies to the Catalan coast, a 600 km long stretch of the Spanish Mediterranean coastline.

This work was supported by the Spanish Agency of Research in the framework of the CoastSpace project, TED2021-130001B-C21 (MCIN/AEI/10.13039/501100011033).

How to cite: Romero-Martín, R., Valdemoro, H., Ranasinghe, R., and Jiménez, J. A.: A framework for assessing the space needed for dune-based coastal adaption at multiple time scales., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14844, https://doi.org/10.5194/egusphere-egu23-14844, 2023.

The ongoing rise of atmospheric temperatures and sea level is exacerbating Arctic coastal permafrost thaw which leads to increased coastal erosion and input of permafrost soils into the Arctic Ocean. Permafrost soils hold vast amounts of organic carbon (OC) which is released into the coastal waters upon thawing. The fate of this OC with regards to its transport and degradation pathways is not yet fully understood - it could either be degraded within the water column and released into the atmosphere as CO2 or it could be buried at the sea floor. When settling onto the seafloor sediment-water interactions become crucial in the OC degradation process. These so-called flocculation layers at the sediment-water interface hold a high potential for sediment re-suspension and therefore represent an environment favouring the degradation of OC thus preventing burial. Yet, there is little data available from these flocculation (i.e. nepheloid) layers, particularly in the Arctic shelf seas.

To improve our understanding of OC degradation within these flocculation layers, we analysed samples from the flocculation layer and from the underlying surface sediments for organic geochemical parameters (TOC, C/N values, δ13C, Δ14C, sediment surface area). Samples within this study were collected along two cross-shelf transects in the Laptev and in the East Siberian Sea during ISSS-2020 expedition in late summer (Sept-Oct) of 2020 onboard R/V Akademik Msistlav Keldysh. First results show variations in OC composition in both shelf seas between the flocculation and surface sediment layers and also with increasing water depth and distance from shore, further emphasizing the degradation potential of this particular layer. With the collected data, we want to gain new insights into how transport and degradation processes of terrestrial OC vary across the vast Siberian shelves.

How to cite: Madaj, L., Keskitalo, K., Gustafsson, Ö., Tesi, T., Semiletov, I., Dudarev, O., Martens, J., Haghipour, N., Bröder, L., and Vonk, J.: Cross-Shelf Transport, Composition and Degradation of Terrestrial Permafrost Organic Matter at the Sediment-Water Interface in the Laptev and East Siberian Seas, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14978, https://doi.org/10.5194/egusphere-egu23-14978, 2023.

Nature-based solutions (NbS) are fast becoming the norm for multifunctional climate adaptation to the combined challenges of increased sea-level rise, coastal population densities, and erosion of sandy shores worldwide, delivering functions such as flood prevention, recreation, and biodiversity benefits. However, it remains a challenge to the research field to inform decision-makers well on the outcomes and trade-offs of designing, planning, and managing the multifunctional NbS. This study set out to identify the information requirements by decision-makers on NbS for coastal climate adaptation. Using the Sand Motor in The Netherlands as a case study, we applied a policy science framework to distinguish four stages of decision-making to quantitatively analyse the content of functions and indicators utilized per stage in public policy documents. These stages are the ambition, political, bureaucratic, and provisioning processes. This study is the first comprehensive empirical investigation distinguishing these crucial stages of decision-making to analyse NbS information requirements. Our results show, most notably, that as the project developed through the decision-making stages, the content of the functions and indicators changed from abstract to concrete. And, with it, the content of the information required shifted significantly. These results suggest that it is crucial for academic researchers to recognize the decision-making process their information will be used in and adapt its content and level of abstraction accordingly to increase its uptake in decision-making. This study lays the groundwork for future research into the multiple dimensions of NbS decision-making and for the increased understanding of the information requirements on evaluation and trade-offs in planning, designing, and managing NbS, to increase the ability of NbS to deliver multifunctional coastal climate adaptation for sandy shores worldwide.

How to cite: Geukes, H., Van Oudenhoven, A., and Van Bodegom, P.: Decision-Making on Nature-Based Solutions for Multifunctional Coastal Climate Adaptation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15297, https://doi.org/10.5194/egusphere-egu23-15297, 2023.

Nature-based solutions (NBS) are promising methods to enhance biodiversity and adapt to climate change in coastal areas. However, upscaling NBS to replace conventional methods requires knowledge about their performance from multiple perspectives, e.g., biodiversity, coastal safety, and economy. In recent years, great efforts have been put into researching NBS pilots of sandy solutions. Some of the most prominent examples are found in the Netherlands, e.g., the Sandmotor, the Hondsbossche Dunes, and the Prince Hendrik Sand dike. These are examples of large-scale interventions with nourished sand volumes of hundreds of thousands to millions of cubic meters. In contrast, this study focuses on small-scale NBS pilots of sandy solutions with nourishment volumes of hundreds to thousands of cubic meters. Two NBS pilots in Sweden are described and analysed, and the advantages and disadvantages of small-scale NBS are discussed in relation to larger-scale interventions.

The first pilot was installed in 2018 in the Furusund navigational fairway in the Stockholm Archipelago. A few hundred cubic meters of sand was nourished to a beach subject to erosion due to ship-generated waves. The nourishment protects an eroding bluff and prevents the loss of forest areas with high nature values. Compared to hard solutions, e.g., a rock revetment, the small-scale beach nourishment supplies sand to a small sandy beach down-drift used for recreational purposes. Since the implementation, a significant part of the nourishment has already been eroded, and the expected lifetime of this intervention is in the order of a few years.

The second pilot was installed at Fortuna beach, located in the narrow sound between Sweden and Denmark. The area has a low-energy wave climate, and the nourishment was designed to protect a beach in front of a residential area from storm erosion recurring with decadal frequency. The beach and dune area were nourished with approximately 3000 m³ of sediment. The area has limited offshore sand resources that can be extracted without adverse environmental impact. Therefore, the beneficial use of sediment dredged from local marinas and a mixture of sand and seaweed from nearby beach clean-ups was used to carry out the project. Within a year after the nourishment, a storm with a recurrence period of approximately 5-20 years hit the coast, but the dune volume still exceeded the volume before the measures.

Experiences from the small-scale sandy solutions are that the limited extent of the interventions facilitates financing and permitting processes, which can be a bottleneck in upscaling NBS. The limited volume of nourishments makes it easier for beneficial use of dredged material, which in many cases is viewed as a waste rather than a resource. Both the volume and timing can be adapted to nearby dredging operations, thus reducing the cost of maintenance of small marinas with a high cultural value. The short lifetime and low safety level of small-scale NBS can be a disadvantage but allow for more flexibility, and no-regret solutions compatible with adaptive pathway approaches to climate change adaptation.

How to cite: Hallin, C., Schmidt, E., and Almström, B.: Small-scale nature-based solutions for protection of sandy coasts, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15893, https://doi.org/10.5194/egusphere-egu23-15893, 2023.

Extreme temperatures, especially long-lasting heat and cold waves in urban areas, lead to thermal stress of the population and increase the number of weather-related health risks and deaths. The observed climate trend and the associated increase of extreme weather events are expected to continue in the future. Thus, the evaluation of urban thermal stress and the associated health effects becomes an important issue for urban planning and risk management. For Austrian cities, an information system for temperature warnings already exists (Weather warnings, ZAMG), which is based on the information of regional weather forecast models. However, this information does not have the required spatial resolution needed to resolve urban structure and thus to account for the urban heat island effect or cold stress situations in winter.

The aim of this project is to provide the basis for the improvement of extreme weather/thermal (dis)comfort warning systems in Austrian major cities by using high-resolution weather predictions (100 m). Therefore, the soil model SURFEX (developed by Météo France) coupled with the AROME numerical weather forecast model is applied to selected cities in Austria and used to determine the best model configuration to compute short-term forecasts (+60 hours). This method provides not a full dynamical model, but a way of pyhsical downscaling with height corrections and a high-resolution surface model.

In this project, land use parameterization will be updated and improved based on Pan-European High Resolution Layers (e.g. Urban Atlas) of the Copernicus Land Monitoring service in ECOCLIMAP (predefined land use classes for SURFEX). The model output will be verified with in-situ operational and crowd-sourced observations. Furthermore, the results will be compared to the micro-scale urban climate model MUKLIMO_3 from the German Weather Service (100 m) and various thermal infrared (TIR with 150 to 250 m) datasets. The novel modeling approach for simulating thermal stress in urban areas serves as the basis for improving the operational prediction system of extreme temperatures, for optimizing the future extreme weather warning system at the ZAMG, and for decision-making for the involved cities and their stakeholders.

How to cite: Oswald, S., Schneider, S., Zuvela-Aloise, M., Hahn, C., and Wastl, C.: Improvement and verification of urban extreme temperature predictions with satellite and ground observations in Austria (VERITAS-AT), EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-164, https://doi.org/10.5194/egusphere-egu23-164, 2023.

In the recent past, the increase in computational resources allowed researchers to run simulations at increasingly horizontal and time resolutions. One such project is the World Climate Research Program’s Coordinated Regional Downscaling Experiments Flagship Pilot Studies (FPS) on convective phenomena. This FPS encompasses a set of simulations driven by the ERA-Interim reanalysis for the period from 2000-2009 (hindcast) and by the Coupled Model Intercomparison Project Phase 5 Global models for the 1996-2005 period (historical). Most models feature a horizontal resolution of 2.2 to 3 km, nested in an intermediate resolution of 12-25 km. An extended Alpine domain is considered for the simulations, due to the complexity of the mountain system together with heavy precipitation events, a large observational network and the high population density of the area. This initiative aims to build first-of-its-kind ensemble climate experiments of convective-permitting models to investigate convective processes over Europe and the Mediterranean.

In this study, the Distribution Added Value metric is used to determine the improvement of the representation of all available FPS hindcast and historical simulations for the daily mean wind speed. The analysis is performed on normalized empirical probability distributions and considers station observation data as a reference. The use of a normalized metric allows for spatial comparison among the different altitudes and seasons. This approach permits a direct assessment of the added value between the higher resolution convection-permitting regional climate model simulations against their global driving simulations and respective coarser resolution Regional Model counterparts. Although the complexity of such simulations, those not always reveal an added value. In general, results show that models add value to their reanalysis or forcing global model, but the nature and magnitude of the improvement on the representation of wind speed vary depending on the model, the spatial distribution and the season.

How to cite: Molina, M. O., Careto, J., Gutiérrez, C., Sánchez, E., and Soares, P.: The added value of regional climate simulations at kilometre-scale resolution to describe daily wind speed: the CORDEX FPS-Convection multi-model ensemble runs over the Alps, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-353, https://doi.org/10.5194/egusphere-egu23-353, 2023.

Global Climate Models (GCMs) are fundamental for simulating future climate conditions. However, such tools have limitations like their coarse resolution, systematic biases, and considerable uncertainties and spread among the projections generated by different models. Thus, raw outputs from GCMs are insufficient for regional-scale studies, which can be solved using downscaling techniques. These methods are particularly relevant for South America (SA), given the continent's climate regimes and topographic complexity. Moreover, critical socio-economic activities developed in SA, such as rainfed agriculture and hydroelectric power generation, are highly dependent on climate conditions and susceptible to extreme events, which can lead to intense droughts or floods depending on the region. Given the background, this study aims to analyze the performance of the statistical downscaling technique Quantile Delta Mapping (QDM) applied to precipitation projections simulated by an ensemble composed of eight GCMs from the Coupled Model Intercomparison Project Phase 6 (CMIP6) for SA. In this manner, we evaluate both the original precipitation projections from the GCMs, and after applying the QDM statistical downscaling technique. Daily precipitation data from the Climate Prediction Center (CPC), with a horizontal resolution of 0.5°, and from the Multi-Source Weighted-Ensemble Precipitation version 2 (MSWEPV2), with a horizontal resolution of 0.1°, are used as a reference, so the final resolution of the GCMs (and the ensemble) projections after the QDM technique application is the same from the different validation databases. Preliminary results with CPC indicate a satisfactory performance of the technique on precipitation simulations over SA.

The authors thank the CAPES, the R&D Program regulated by ANEEL, and the companies Engie Brasil Energia and Energética Estreito for their financial support.

How to cite: de Souza Ferreira, G. W., Simões Reboita, M., and Martins Ribeiro, J. G.: Applying statistical downscaling to CMIP6 projections of precipitation for South America: Analysis of pre and post-processed simulations, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-617, https://doi.org/10.5194/egusphere-egu23-617, 2023.

Two main approaches to downscale global climate projections are possible: dynamical and statistical downscaling. Both families have been widely evaluated, but intercomparison studies between the two families are scarce, and usually limited to temperature and precipitation. In this work, we present a comparison between a Statistical Downscaling Model (SDM) based on Machine Learning and six Regional Climate Models (RCMs) from EURO-CORDEX, for five variables of interest: temperature, precipitation, wind, humidity and solar radiation. The study expands at a continental scale over Europe, with a spatial resolution of 0.11^o and daily data. Both the SDM and the RCMs are driven by the ERA-Interim reanalysis, and observations are taken from the gridded dataset E-OBS. Several aspects have been evaluated: daily series, mean values and extremes, spatial patterns and also temporal aspects. Additionally, in order to analyze the intervariable consistency, a multivariable index (Fire Weather Index) derived from the fundamental variables has been included. The SDM has reached better scores than the RCMs for all the evaluated aspects with only a few exceptions, mainly related to an underestimation of the variance. After bias correction, both the SDM and the six RCMs present similar results, with no significant differences among them. Results presented here, combined with the low computational expense of SDMs and the limited availability of RCMs over some CORDEX domains, should motivate the consideration of statistical downscaling at the same level as RCMs by official providers of regional information, and its inclusion in reference sites. Nonetheless, further analysis on crucial aspects such as the impact on long-term trends or the sensitivity of different methods to being driven by Global Climate Models instead of by a reanalysis, is needed.

How to cite: Hernanz, A., Correa, C., Domínguez, M., Rodríguez-Guisado, E., and Rodríguez-Camino, E.: Can Statistical Downscaling based on Machine Learning compete with Regional Climate Models? A comparison for temperature, precipitation, wind, humidity and radiation over Europe under present conditions, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1296, https://doi.org/10.5194/egusphere-egu23-1296, 2023.

The reliable prediction of flash flood relevant heavy precipitation events under climate change conditions remains a challenging task for the downscaling community. Therefore, a huge variety of downscaling approaches have been proposed and successfully applied, however, there is still potential for improvements. The conducted study aims to investigate potential improvements by circulation pattern (CP) trends conservation and their utilization for CP conditional statistical downscaling of daily summer precipitation in the (pre-)alpine region of Bavaria. The CPs have been created taking only atmospheric variables into consideration and the link to precipitation is established via CP conditional cumulative distribution functions (CDF) of the observed precipitation at selected measurement sites across the region. The derived CDFs allow for the sampling of CP conditional precipitation values at the station scale which are subsequently bias corrected by quantile mapping (QM) and parametric transfer functions (PTFs) as tested methods. The predicted precipitation values have been evaluated against obervations using different performance measures such as Kling-Gupta Efficiency (KGE), Root Mean Squared Error (RMSE) and Mean Absolute Error (MAE). In order to properly account for extreme events the evaluation has been conducted for the complete precipitation distribution and for the distribution above the 95th percentile seperately. The results show that the described CP conditional downscaling approach is capable of yielding more accurate daily precipitation values especially in the extremes compartment in which an average gain in prediction skill of + 0.24 and a maximum gain of + 0.6 in terms of KGE has been observed. This shows that the conservation of trends and atmospheric information through CPs and their utilization for downscaling can lead to improved precipitation downscaling results.

How to cite: Böker, B., Laux, P., Olschewski, P., and Kunstmann, H.: Accurate heavy precipitation prediction in an (pre-)alpine area: The benefit of trend conservation in circulation pattern conditional statistical downscaling., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2260, https://doi.org/10.5194/egusphere-egu23-2260, 2023.

While much effort has been devoted to analyzing long-term changes of temperature and precipitation in mean values and extremes, studies on changes in variability have been rather scarce. Trends in variability are, however, important, among others because their interaction with trends in mean values determines the degree with which extremes would change. The knowledge of long-term changes in temporal variability is essential for assessments of climate change impacts on various sectors, including hydrology (floods and droughts), agriculture, health, and energy demand and production.

SPAGETTA is a stochastic spatial daily weather generator (WG), which uses first-order multivariate (dimension = number of variables X number of gridpoints) autoregressive model to represent the spatial and temporal variability of surface weather variables (including precipitation and temperature). We consider the generator to be a suitable tool for assessing changes in the spatial and temporal variability of the weather series because of following reasons: (A) The inter-gridpoint lag-0 and lag-1(day) correlations included in a set of WG parameters may serve as representatives for spatial and temporal variability of input weather variables. (B) Statistical significance of changes in the lag-0 and lag-1 correlations derived from the input series may be easily assessed by comparing the changes with a variability of the lag-0 and lag-1 correlations related to the stochasticity in input weather series (the variability is assessed across a set of multiple realisations of the synthetic series). (C) Separate effects of changes in various statistical characteristics on any climatic characteristic may be easily assessed. Specifically, having analysed changes in the means, variability and inter-gridpoint correlations (e.g. based on RCM simulations of the future climate), we may modify only a selected (possibly only a single one) WG parameter(s) before producing the synthetic series and analysing effect of climate change on the climatic characteristics.

In the first part of the contribution, we employ SPAGETTA generator to analyse changes in interdiurnal variability of precipitation and temperature in 8 European regions (defined in Dubrovsky et al 2020, Theor Appl Climatol) using (a) gridded observational (last N years vs. first N years in available E-OBS times series) and (b) RCM-simulated surface weather series (2070-2099 vs 1971-2000; outputs from 19 RCMs available from the CORDEX database are analysed). In doing this, we assess the statistical significance of the detected changes. In the second part, we assess separate effects of changes in the means, variability and lag-0 & lag-1 correlations of temperature and precipitation (the changes based on a set of 19 RCM simulations are used to modify the corresponding WG parameters) on a set of climatic indices - including a set of compound precipitation-temperature characteristics representing spells of days with spatially significant extent of significantly non-normal weather (e.g. hot-dry spells).

How to cite: Dubrovský, M., Huth, R., Stepanek, P., Lhotka, O., Miksovsky, J., and Meitner, J.: Spatial and Temporal Variability of Precipitation and Temperature: Analysis of Recent Changes and Future Development with Use of the Weather Generator and RCM-Based Climate Change Scenarios, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2735, https://doi.org/10.5194/egusphere-egu23-2735, 2023.

Statistical and dynamical downscaling techniques are widely applied in the development of local climate change scenarios. This talk traces the conceptual development of downscaling as a decision-support tool for climate risk assessment, resilience and adaptation planning. Four epochs are identified: (1) early exploration of local changes in key climate variables, such as temperature and precipitation extremes; (2) application of downscaled scenarios to climate impacts modelling (such as for agriculture yield or water resource assessments); (3) advent of ensemble-based methods and more sophisticated handling of uncertainty in the downscaling-impacts workflow; and (4) use of downscaled scenarios to stress-test adaptation options under plausible ranges of climate and non-climatic conditions. Each phase is illustrated by and reflected in the development of the Statistical DownScaling Model – Decision Centric (SDSM-DC) over more than two decades. Questions around fitness for purpose and appropriate uses of the tool are explored. The talk concludes by considering: where next for downscaling?

How to cite: Wilby, R. and Dawson, C.: Conceptual development and use of downscaled climate model information, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3343, https://doi.org/10.5194/egusphere-egu23-3343, 2023.

An atmospheric single-column model (SCM) developed in the framework of the Canadian Regional Climate Model, CRCM, driven by NCEP-NCAR reanalyses is applied to study the non-linear interactions between the surface and the planetary boundary layer (Goyette et al., 2020). The approach to solve the model equations and the technique described may be implemented in any RCM system environment as a model option. The working hypothesis underlying this SCM formulation is that a substantial portion of the variability simulated in the column can be reproduced by processes operating in the vertical dimension and a lesser portion comes from processes operating in the horizontal dimension. This SCM offers interesting prospects as the horizontal and vertical resolution of the RCM is ever increasing. Due to its low computational cost, multiple simulations may be carried out in a short period of time. In this paper, a range of possible results obtained by changing the lower boundary from open water surface to land, and by varying model parameters are mainly shown for central Mediterranean but also for other applications. Results show that the model responded in a highly nonlinear but coherent manner in the lowest levels with changes in air temperature, moisture and windspeed profiles. The latter are consistent with those of the surface vertical sensible, latent heat and momentum fluxes. For example in the central Mediterranean, during a simulated year, air temperature is increased during all the seasons. Specific humidity is increased during the autumn and winter seasons but decreased by during the spring and summer seasons thus showing the contrasting influence of the land surface. The potential for further developments, as well as some guidance as to how to handle mixed land/open water coupling in RCMs, is also provided.

GOYETTE, Stéphane, FONSECA, Cédric, TRUSCELLO, Léonard. Assessment of nonlinear effects of a deep subgrid lake with an atmospheric single‐column model. In: International Journal of Climatology, 2020. doi: 10.1002/joc.6890

How to cite: Goyette, S. and Kasparian, J.: Numerical investigation with a coupled single-column surface-atmosphere model and an application to central Mediterranean, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3479, https://doi.org/10.5194/egusphere-egu23-3479, 2023.

Dynamical downscaling is a widely-used approach to generate regional projections of future climate extremes at a finer scale. Previous studies indicated that the global climate model (GCM) bias correction method prior to dynamical downscaling can improve the simulation of the climate extreme to a certain extent. Recently, a new bias correction method termed MVT was developed. Note that this method did not correct the GCM biases of the climate extreme event explicitly. In this study, we evaluate the MVT method in terms of various climate extreme events through three dynamical downscaling simulations over Asia-western North Pacific with 25 km grid spacing throughout 1980–2014, and further investigate to what extent and how this bias correction method can improve the simulation of downscaled climate extreme events. The dynamical downscaling simulations driven by the original GCM dataset derived from the MPI-ESM1-2-HR (hereafter WRF_GCM), the bias-corrected GCM (hereafter WRF_GCMbc) are validated against that driven by the European Centre for Medium-Range Weather Forecasts Reanalysis 5 dataset, respectively. The results suggest that compared with the WRF_GCM, the WRF_GCMbc shows more than 26% decrease in root mean square errors of the precipitation and temperature extreme indices, and even 61% out of seasonal extreme indices show more than 50% reduction. Such improvements in the WRF_GCMbc are primarily caused by the correct simulation of the large-scale circulation due to the GCM bias correction. The large-scale circulation in turn improves the simulation of the precipitation and cloud by the water vapor transport and further improves the simulation of the 2m temperature by the radiation process and the surface energy balance, which contribute to the better simulation of the precipitation and temperature extreme indices.

How to cite: Zhang, M.-Z., Han, Y., Xu, Z., and Guo, W.: Validation of MVT bias correction in dynamical downscaling simulations for climate extreme, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3741, https://doi.org/10.5194/egusphere-egu23-3741, 2023.

The gridded population, crucial for resource allocation and emergency support, is mainly downscaled from the census data with administrative divisions. A common dasymetric mapping approach is building a regression model between aggregated geospatial properties and population potential at the administrative level and then applying this model directly to the grid level. The aggregation of geospatial properties often relies on statistical methods like averaging. However, the difference in scale between the two levels can lead to the heterogeneity of geospatial properties, which causes a gap between the training domain and the target domain and makes these methods fail to preserve the physical meaning of geographic properties. To address this issue, we propose a deep learning-based approach, in which a sophisticated loss function involving tripartite elements, gridded geospatial properties, gridded population potential, and administrative population potential, is designed. In this way, scale heterogeneity both in aggregation and domains can be avoided. In this study, a 30-meter resolution population density map of Hong Kong is produced through the proposed approach. The validation result shows that compared with both the machine learning-based or the artificial neural network-based one, the proposed approach gets a lower RMSE and potentially provides a more accurate reference for detailed urban management.

How to cite: Lu, W. and Weng, Q.: Scale Heterogeneity Avoided Dasymetric Mapping for the Gridded Population, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3922, https://doi.org/10.5194/egusphere-egu23-3922, 2023.

Regional climate models, due to their systematic biases, are not usable for impact assessment and policy-relevant applications. It is common to post-process the regional model outputs with appropriate bias correction methodologies to provide reliable climate change information. We apply a distribution-based, trend-preserving quantile mapping procedure to bias correct the projections of daily precipitation and temperature from an ensemble of 5 RCMs driven by 5 GCMs, each at a resolution of 0.11°, chosen from the EURO-CORDEX initiative. The gridded observations from the German Weather Service, DWD-HYRAS, has been used as a reference for the bias correction. The impact of the bias correction is found to be more pronounced on precipitation than on temperature, as the precipitation biases are larger. The models are wetter and underestimate (overestimate) the daily maximum (minimum) temperature. The correction method eliminates large parts of these biases and maps the distributions of both the variables well with that of observations. The bias adjustment also leads to the narrowing down of the uncertainties in the projected changes of both the variables. The decomposition of total variance into model uncertainty and internal variability suggests that the bias correction acts mostly on the former component. The internal variability component does not seem, however, to undergo considerable changes following the bias correction. Due to the reduction of the uncertainty, we find a slight improvement in the signal-to-noise ratio in the projections. 

How to cite: Adakudlu, M., Xoplaki, E., and Luther, N.: Implications of statistical bias adjustment for uncertainties in regional model projections, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4495, https://doi.org/10.5194/egusphere-egu23-4495, 2023.

In this study, a multi-model ensemble of regional climate and air quality coupling model system was established to evaluate current climate and air pollution in China during 2010-2014. Meteorological initial and boundary conditions were obtained from the multi earth system models used in the Coupled Model Intercomparison Project Phase 6 (CMIP6) with a dynamical downscaling method and the National Centers for Environmental Prediction Final Analysis (NCEP-FNL) reanalysis data. These downscaling data under the historical scenario and FNL data were applied to driven the Weather Research and Forecasting model coupled to Chemistry (WRF-Chem) to simulate current climate and air quality. A comprehensive evaluation of the current five years was conducted against the ground-level meteorological and chemical observations. The performances for the 2 m temperature were very well and consistently overestimated the wind speed at 10 m by 0.8~1.2 m/s. PM_2.5 and ozone concentrations were underestimated by the downscaling data driven simulations compared with the FNL data. The model performance was relatively well and can be used to study the impacts of climate change on China's future air quality and pollution events in the context of carbon neutrality and clean air, which may shed light on policy formulation for medium and long-term air quality management and climate change alleviation.

How to cite: Cui, M.: Multi-model downscaling simulations of regional climate and air quality in China, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5321, https://doi.org/10.5194/egusphere-egu23-5321, 2023.

Floods are highly destructive natural hazards causing widespread impacts on socio-ecosystems. This hazard could be further amplified with the ongoing climate change, which will likely alter magnitude and frequency of floods. Estimating how flood-rich periods could change in the future is however challenging. The classical approach is to estimate future changes in floods from hydrological simulations forced by time series scenarios of weather variables for different future climate scenarios. The development of relevant weather scenarios for this is often critical. To be adapted to the critical space and time scales of the considered basins, weather scenarios are thus typically produced from climate models with downscaling models, either dynamical or statistical.

In this study, we assessed the ability of two typical simulations chains to reproduce over the last century (1902-2009) and from large-scale atmospheric information only observed temporal variations of river discharges and flood events of the Upper Rhône River (10,900 km²). The modeling chains are made up of (i) the atmospheric reanalysis ERA-20C, (ii) either the statistical downscaling model SCAMP (Raynaud et al., 2020) or the dynamical downscaling model MAR (Gallée and Schayes, 1994), and (iii) the glacio-hydrological model GSM-SOCONT (Schaefli et al., 2005).

The daily Mean Areal Temperature (MAT) and Precipitation (MAP) time series were compared to the observed ones over the period 1961-2009. The meteorological results highlight the need for a bias-correction for both downscaling models. To avoid irrelevant simulations of the snowpack dynamics, especially for high elevations, the bias-correction was needed not only for the precipitation and temperature scenarios but also for the lapse scenarios of the dynamical downscaling chain. Simulated discharges are globally in very good agreement with the reference ones in the bias-corrected simulations. Whatever the river basin considered, the multi-scale observed variations of discharges are well reproduced (daily, seasonal and interannual). The reconstruction power of the chains is lower for low frequency hydrological situations, namely low flow sequences and annual discharge maxima. Flood events tend to be underestimated by each simulation chain.

Flood activity was also estimated from the discharge time series using the Peak Over Threshold (POT) method. The results over the last century are very promising, and encourage us to continue towards simulations over the last millennium. Outputs from the PMIP4 experiments (CESM1 Last Millennium Ensemble) will be statistically downscaled with the SCAMP model (for reasons of computation costs) and used as forcings in the GSM-SOCONT model.

References: 
- Raynaud et al. (2020) HESS doi.org/10.5194/hess-24-4339-2020 
- Gallée and Schayes, 1994 MWR doi:10.1175/1520-0493(1994)122<0671:DOATDM>2.0.CO;2
- Schaefli et al. (2005) HESS doi.org/10.5194/hess-9-95-2005

How to cite: Legrand, C., Wilhelm, B., and Hingray, B.: Simulating river discharges variations and flood events from large-scale atmospheric information with statistical and dynamical downscaling models: Example of the Upper Rhône River, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6168, https://doi.org/10.5194/egusphere-egu23-6168, 2023.

Climate simulations often need to be adjusted before carrying out climate impact studies at regional scale in order to reduce the biases often present in climate models. To do that, bias adjustment methods are usually applied to climate output simulations and are calibrated over a reference period. This period ideally includes good observational coverage and is often defined as the 2 or 3 more recent decades. However, on these timescales, the climate state may be influenced by the low-frequency internal climate variability. There is therefore a risk of introducing a bias to the climate projections by bias-adjusting simulations with low-frequency variability in a different phase to that of the observations. We proposed here a new pseudo-reality framework using an ensemble of simulations performed with the IPSL-CM6A-LR climate model in order to assess the impact of the low-frequency internal climate variability of the North Atlantic sea surface temperatures on bias-adjusted projections of mean and extreme surface temperature over Europe. We show that adjusting a simulation in a similar phase of the Atlantic Multidecadal Variability to that of the pseudo-observations reduces the pseudo-biases in temperature projections. Therefore, for models and regions where low frequency internal variability matters, it is recommended to sample relevant climate simulations to be bias adjusted in a model ensemble or alternatively to use a very long reference period when possible.

How to cite: Bonnet, R., Vrac, M., Boucher, O., and Jin, X.: Sensitivity of bias adjustment methods to low-frequency internal climate variability over the reference period: an ideal model study, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6529, https://doi.org/10.5194/egusphere-egu23-6529, 2023.

The nature and severity of climate change impacts varies significantly from region to region. Consequently, high-resolution climate information is needed for meaningful impact assessments and the design of mitigation strategies. This demand has lead to an increase in the coupling of Empirical Statistical Downscaling (ESD) models to General Circulation Model (GCM) simulations of future climate. Here, we present a new open-source Python package (pyESD; github.com/Dan-Boat/PyESD) that implements several Perfect Prognosis ESD (PP-ESD) methods and the whole downscaling cycle. The latter includes routines for data preparation, predictor selection and construction, model selection and training, evaluation, utility tools for relevant statistical tests, visualisation, and more. The package includes a collection of well-established Machine Learning algorithms and allows the user to choose a variety of estimators, cross-validation schemes, objective function measures, hyperparameter optimization, etc., in relatively few lines of codes. The package is highly modular and flexible, and allows quick and reproducible downscaling of any climate information, such as precipitation, temperature, wind speed or even glacial retreat. We demonstrate the effectiveness of the new PP-ESD framework by generating station-based downscaling products of precipitation and temperature for complex mountainous terrain in Southwest Germany.

How to cite: Mutz, S. G. and Boateng, D.: pyESD: An open-source Python framework for empirical-statistical downscaling of climate information, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7470, https://doi.org/10.5194/egusphere-egu23-7470, 2023.

Numerous historical sources report on hazardous past climate and weather events that had considerable impacts on society. Studying changes in their occurrence or mechanisms behind such events is however hampered by a lack of spatial weather information. For Switzerland, we created a daily high-resolution (1x1 km²) reconstruction of temperature and precipitation fields for the years 1763 to 1960 using an analog resampling method based on observational data. The resampled fields are further post-processed by assimilating temperature observations and quantile mapping the precipitation fields. Together with the present-day meteorological fields, this forms a more than 250-year long gridded data set.

We use this data set to evaluate changes in spring weather impacts over the last 250 years. The spring season receives fewer attention since it has no extreme events in absolute terms. However, it is relevant since weather conditions in spring can delay vegetation onset and growth, and can create substantial vegetation damages due to for example late frost and snow events. We evaluate therefore the long-term changes of spring fresh snow days, late frost days, frost days, and warm days, and compare it to changes of spring onset and reconstructed phenological stages.

How to cite: Noemi, I. and Stefan, B.: Weather reconstruction and application for Switzerland: Long-term changes of spring weather impacts since 1763, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8234, https://doi.org/10.5194/egusphere-egu23-8234, 2023.

Knowing soil moisture conditions accurately is extremely important for natural hazards prediction, agriculture, and other water resources management practices. Remote sensing products have been used more and more in these contexts. Their main advantage is the spatial coverage, which allows one to obtain continental or even global products. Nevertheless, there are limitations associated with them, such as reduced penetrating depth, impact of cloudiness and snow/ice, and low spatial and temporal resolutions. To compensate for the low spatial resolution, downscaling techniques have been developed that combine different remote sensing products and/or other data considered to affect soil moisture redistribution. The main limitation in their development, is the lack of data to validate the techniques and the final product. Oftentimes in situ measurements are used for the calibration/training and for the testing/verification. These are very sparse, i.e., only available at few locations, and hard to compare directly, as both the satellite products and the downscaled estimates are volumetric and not point estimates.

Here, we create a soil moisture downscaling playground by generating soil moisture estimates with a physics-based hydrological model (ParFlow-CLM) at different resolutions, from a few kilometers to 100 meters. Having continuous gridded estimates of high- and low- resolution soil moisture with a reliable physics-based model, allows us to test and compare different downscaling techniques as well as the impact on the scaling of individual inputs/parameters. As an initial experiment, we model the East Taylor catchment (Colorado, USA) at 100m and 1000m resolution, by only changing the topography (i.e., all other inputs are resolved at 1000m), which is not only the best-known input even at high resolutions, but also the most impactful in soil moisture redistribution. The best performing downscaling technique will allow us, in an operational setup, to run the physics-based model at a coarser resolution but still have a high-resolution product in a computationally inexpensive manner. Beyond our application, the high- and low- resolution simulations generated in this work can be used for the validation of any downscaling technique also applicable with remote sensing products.

How to cite: Leonarduzzi, E. and Maxwell, R. M.: A soil moisture downscaling playground of multiple resolution physics-based simulations, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8829, https://doi.org/10.5194/egusphere-egu23-8829, 2023.

Copernicus Climate Change Service (C3S) integrates multiple seasonal forecast models of climate variables with multiple ensemble realizations. Assessing the risks of natural hazards with high impacts on human and natural systems and providing actionable services at the local scale require high-resolution predictions. We implement the AI-based approach proposed by Heidari et al. (2023) to address such needs and reach a kilometer scale. While downscaling seasonal forecasts, it is crucial to transfer the full range of the uncertainties given by the ensembles.

This study assesses how uncertainty is transferred by an AI-based downscaling approach. Quantile-based metrics are here used to measure the ensemble variability between seasonal forecasts and their downscaled products. On the other side, quantile-based metrics can also give an alternative description of the ensemble variabilities, which could replace the raw ensemble members in the downscaling process. In this study, the AI-downscaling system is tested by inputting (a) raw ensemble members and (b) quantile-based metrics. Transferred uncertainty and downscaling accuracy are then evaluated to develop and implement an optimal downscaling approach with hazard-dependent inputs being selected at  regional and local scales.

Heidari F., Lin Q., Espitia Sarmiento E.F., Toreti A., and Xoplaki E. (2023): A deep learning technique to realistically bias correct and downscale seasonal forecast ensembles of climate variables towards the development of an AI-based early warning system, EGU 2023 abstract

How to cite: Lin, Q., Heidari, F., Espitia Sarmiento, E. F., Adakudlu, M., Toreti, A., and Xoplaki, E.: Analysis of Ensemble Uncertainty Transfer in AI-Based Downscaling of C3S Seasonal Forecast, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9644, https://doi.org/10.5194/egusphere-egu23-9644, 2023.

The lack of long-term and consistent meteorological observations limits the application of land-surface simulators (e.g., of phenomena in hydrology, the cryosphere, ecology) at remote locations. For example, most permafrost areas are remote and lacking consistent meteorological time series, models that describe permafrost change over time cannot be driven for comparison with observations or for impact studies. Reanalysis-derived time series are valuable because they are available with global coverage, for a long time period, and for a broad set of physically consistent variables. Multiple reanalyses can be used to provide estimates of uncertainty. Practically, however, this data is difficult to use for several reasons: grid-scale reanalyses must be downscaled and interpolated horizontally (and vertically within the atmospheric column for mountains regions) to the site‑scale, differences in variables, units, and delivery between reanalyses must be reconciled, and large volumes of data need to be handled. Globsim is an open-source python library (available via GitHub) that was developed to handle these challenges and to facilitate a simulation workflow that takes advantage of the multiple reanalysis products available today. It outputs sub-daily meteorological time series that resemble meteorological stations for any location on the planet. Since the release of the first version of Globsim, we have improved usability, refactored code for maintainability and speed, and fixed a number of bugs. We also added support for ERA5 ensemble data, and added more sophisticated heuristic downscaling algorithms, including TOPOscale for elevation-adjusted radiative fluxes. We use Globsim as a core tool in a multi-model permafrost simulation workflow and, as a future step, we intend to use it as part of a debiasing routine to make predictions of permafrost using climate scenarios. We expect this tool to be broadly applicable to climate change impact modelers and other scientists using climate driven simulations working in (remote) locations that lack meteorological data of sufficient quality and duration for their application.

How to cite: Brown, N., Gruber, S., and Cao, B.: Globsim v.3 – Improvements to an open-source software library for utilizing atmospheric reanalyses in point-scale land surface simulation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10595, https://doi.org/10.5194/egusphere-egu23-10595, 2023.

Bias adjustment of climate models is today normally performed with quantile mapping methods that account for the whole distribution of the parameter. The bulk of the distribution is well described as long as sufficient data records are used (Berg et al., 2012), however, the extreme tails will always suffer from large uncertainties. These uncertainties stem from both the climate model and the reference data set, which prevents a robust and detailed identification of bias in the extreme tail. Empirical quantile mapping methods are therefore prone to overfitting, and may introduce substantial bias when applied outside the calibration period. Commonly, a constant adjustment is applied for values outside the range of the calibration period, but there is room for improvements of the extrapolation method.

While working with a climate service for Sweden, a clear offset was identified between data adjusted within and outside the calibration period for an extreme indicator of daily maximum precipitation. This study explores different extrapolation methods for the extreme tail of the distribution in the spline-based empirical quantile mapping method of the MIdAS bias adjustment method (Berg et al., 2022). By limiting the bias adjustment to the first 95% of the distribution, and thereafter applying a constant or a linear fit to the remaining 5% of data in the tail, the offset is strongly reduced and the adjusted extremes become more robust and plausible.

Berg, P., Feldmann, H., & Panitz, H. J. (2012). Bias correction of high resolution regional climate model data. Journal of Hydrology, 448, 80-92.

Berg, P., Bosshard, T., Yang, W., & Zimmermann, K. (2022). MIdASv0. 2.1–MultI-scale bias AdjuStment. Geoscientific Model Development, 15(15), 6165-6180.

How to cite: Berg, P., Bosshard, T., Bärring, L., Södling, J., Wilcke, R., Yang, W., and Zimmermann, K.: Reducing negative impacts of bias adjustment on the distribution tail and extreme climate indicators in MIdAS, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11124, https://doi.org/10.5194/egusphere-egu23-11124, 2023.

The long-existing double-ITCZ problem in GCMs affects not only the models' ability in simulating the current climate, but also implies limitations regarding the assessment of climate sensitivity and global climate change. Using a regional climate model (RCM) with explicit convection at a horizontal grid spacing of 12 km in a large computational domain covering the tropical and sub-tropical Atlantic, we develop a bias-correction downscaling methodology to remove the biases of a driving GCM. The methodology is related to the pseudo-global warming (PGW) approach. Normally this method is used to impress the climate-change signal to a reanalysis-driven RCM simulation, but it can also be used to modulate the lateral-boundary conditions of a GCM, such as to remove the large-scale biases. We show that the double ITCZ problem persists with classical dynamical downscaling (i.e. when driving the RCM directly by the GCM output), but with our bias-corrected downscaling the double ITCZ problem can be removed. Detailed analysis reveals that the main cause of the double ITCZ problem can be attributed to the GCMs' SST bias. Compared to the GCMs' AMIP simulations, RCMs with higher resolution allow explicit deep convection and enable a better simulation of tropical convection and clouds. By improving the corresponding radiative forcing, vertical motion is better simulated. Subsidence stronger to the south of the ITCZ pushes the ITCZ more north in the boreal spring, which is consistent with the observation of the ITCZ. The developed methodology provides an opportunity for better constraining climate sensitivity by removing double-ITCZ biases.

How to cite: Liu, S., Zeman, C., and Schär, C.: Understanding the double-ITCZ problem over the Atlantic with bias-corrected downscaling, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11595, https://doi.org/10.5194/egusphere-egu23-11595, 2023.

Enel, as most of the Energy Players, has an important exposure on weather risk due to the indirect effect of the power demand and to the direct effects on renewable production. A large component of such risk comes from the hydroelectric production, this is especially true in Southern America where, in some countries, it can represent up to 70% of the total production. We present a practical development of an operational chain to extract information from the seasonal forecasts produced by SEAS5. It works on some catchments in Colombia and Peru with the aim to provide an ensemble forecast of monthly precipitations at a high resolution from the fields at low resolution provided by Copernicus. To produce the high-resolution fields of precipitations we developed a procedure based on Lorenz et al. (2021); for our scope, the biases of the SEAS5 forecasts are corrected following a reference climatology obtained from the SEAS5 hindcasts that is calibrated over the cumulative distribution function calculated be mean of historical measurements of the IDEAM weather stations. The method and preliminary results as well as the validation will be shown in this work.

How to cite: Rea, G., Galuzzo, D., and Formenton, M.: Development of an operational seasonal forecast in Colombia and Peru by mean of statistical downscaling of the SEAS5-Copernicus data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12395, https://doi.org/10.5194/egusphere-egu23-12395, 2023.

In this study, the predictability of the coastal ocean is assessed in a downscaling exercise for the Adriatic Sea using NEMO 3.6 over a 19 years’ time window (2001-2019). Inspired by the perfect model approach (Denis et al. 2002, De Elia et al. 2002) using a dynamical downscaling setup, a high resolution (2 km) experiment (Big Brother – BB) for the entire Adriatic Sea is used as the “true” reference for a smaller domain, downscaling experiment (Little Brother – LB) in the Northern Adriatic subbasin. The LB experiment has the same horizontal resolution as the BB (2 km) and is downscaled from a low resolution parent model (6 km), in a ratio of 1/3 resolution jump. The 2 km horizontal resolution fits the purpose of reaching an eddy-permitting grid spacing in the Adriatic basin (Masina and Pinardi, 1994; Cushman-Roisin et al. 2002).

Power spectral density analysis is used to evaluate the kinetic energy variance on the frequency domain among the experiments and compare them with the BB experiment. Overall, the LB is more energetic than the parent model, and the timing of the peaks of energy coincides with the ones of the BB. The energy on the 1 year signal is higher in the LB than the BB. The LB can recover a significant amount of energy for all peaks, with special attention to the 6 months period, which is poorly captured by the parent model. The 4 months signal is equally represented in BB and LB, while there is an underestimation of the 6 months signal of LB with respect to BB. Energy in the LB does not deviate from BB more than ~20% in the low frequencies and ~10% in the high frequencies, while the parent model presents in a whole lower energy than the BB, with higher differences on the low frequencies.

The Northern Adriatic circulation is largely influenced by the surface buoyancy flux and the wind forcing (Cessi et al., 2014), which play a significant role in the energy budget and the anti-estuarine overturning circulation of the Adriatic basin. Differences between LB and parent model results may be associated with the energy cascade due to interactions of internal dynamic processes which are differently represented at different resolutions. Differences between LB and BB results are the effect of the downscaling method and the horizontal resolution ratio between the parent model and the nested LB.

Moreover, the analysis of the wavenumber spectra allows a clear overview of the energy distribution in the space domain among the experiments and the representation of small-scale features in the LB. Small scale features less than twice the grid spacing (~12 km) are absent in the low-resolution parent model outputs. Therefore, the comparison with the true reference, BB, reveals the energy spectrum of the parent model solves only the larger scales, while the downscaling LB can recover the smaller scales absent in the initial and lateral boundary conditions.

How to cite: Tatsch Eidt, R., Verri, G., Santos da Costa, V., Gunduz, M., and Navarra, A.: A downscaling exercise for the Adriatic Sea in a perfect model approach, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12899, https://doi.org/10.5194/egusphere-egu23-12899, 2023.

Climate change impacts assessment crucially relies on climate information at high temporal and spatial resolutions, not available from global climate models (GCMs) involved in the coupled model intercomparison project (CMIP). At the same time, dynamically downscaled regional climate model simulations do not provide global-scale coverage and in several cases are computationally too expensive.

For this reason, downscaling techniques are commonly applied to bridge the resolution gap between GCM simulations and impact studies. The most common methodology is the statistical downscaling approach. However, statistical downscaling fast computation comes at a price, it does not account for physical and dynamic processes potentially inflates temporal variability of the original simulations’ resolution. Given this limitation, the analogs technique may represent a valuable alternative since it considers both large and local scales dynamics balanced by a reasonable increase in computational costs.

The present study explores differences, added value, and limitations characterizing state-of-the-art bias adjustment/statistical downscaling based on a stochastic quantile mapping approach and the analogs technique. In particular, the comparison applies to the data computed in the inter-sectoral impact model intercomparison project (ISIMIP) and data obtained by applying the analogs method based on the same ISIMIP reference dataset. The two approaches are compared and evaluated in terms of the historical period observed statistics reproduction for a few climate variables over European regions.

This study is performed in the framework of GoNEXUS and NEXOGENESIS European projects.

How to cite: Peano, D., Sangelantoni, L., and Alvarez-Castro, C.: Evaluating state-of-art statistical downscaling and analogs approaches on historical climate statistics over European regions, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13876, https://doi.org/10.5194/egusphere-egu23-13876, 2023.

High resolution rainfall projections are useful for planning for climate change [1] but are expensive to produce using physical simulations. We make novel use of a state-of-the-art generative machine learning (ML) method, diffusion models [2], to more cheaply generate high resolution (8.8km) daily mean rainfall samples over England and Wales conditioned on low resolution (60km) climate model variables. The downscaling model is trained on output from the Met Office UK convection-permitting model (CPM) [3]. We then apply it to predict high-resolution rainfall based on either coarsened CPM output or output from the Met Office HadGEM3 general circulation model (GCM). The downscaling model is stochastic and able to produce samples of high-resolution rainfall that have realistic spatial structure, which previous methods struggle to achieve. It is also easy to train and should better estimate the probability of extreme events compared to previous generative ML approaches.

The downscaling model samples match well the rainfall distribution of CPM simulation output. We use as our conditioning variables We obtained further improvements by also including high-resolution, location-specific parameters that are learnt during the ML training phase. We will discuss the challenges of applying the model trained on coarsened CPM variables to GCM variables and present results about the method’s ability to reproduce the spatial and temporal behaviour of rainfall and extreme events that are better represented in the CPM than the GCM due to the CPM’s ability to model atmospheric convection.

References

[1] Kendon, E. J. et al. (2021). Update to the UKCP Local (2.2km) projections. Science report, Met Office Hadley Centre, Exeter, UK. [Online]. Available: https://www.metoffice.gov.uk/pub/data/weather/uk/ukcp18/science-reports/ukcp18_local_update_report_2021.pdf

[2] Song, Y. et al. (2021). Score-Based Generative Modeling through Stochastic Differential Equations. ICLR.

[3] Met Office Hadley Centre. (2019). UKCP18 Local Projections at 2.2km Resolution for 1980-2080, Centre for Environmental Data Analysis. [Online]. Available: https://catalogue.ceda.ac.uk/uuid/d5822183143c4011a2bb304ee7c0baf7

How to cite: Addison, H., Kendon, E., Ravuri, S., Aitchison, L., and Watson, P.: Downscaling with a machine learning-based emulator of a local-scale UK climate model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14253, https://doi.org/10.5194/egusphere-egu23-14253, 2023.

Statistical bias adjustment is now common practice when using climate models for impact studies, prior to or in conjunction with downscaling methods. Examples of widely used methodologies include CDFt (Vrac et al. 2016), ISIMIP3BASD (Lange 2019) or equidistant CDF matching (Li et al. 2010). Though common practice, recent papers (Maraun et al. 2017) have found fundamental issues with statistical bias adjustment. When multivariate aspects are not evaluated, improper use of bias adjustment is not detected. Fundamental misspecifications of the climate model, such as the displacement of large-scale circulation, cannot be corrected. Furthermore, results are sensitive to internal climate variability over the reference period (Bonnet et al 2022). If applied, bias adjustment methods should therefore be evaluated carefully in multivariate aspects and targeted to the use-case at hand.

However, good practice in the evaluation and application of bias adjustment methods is inhibited by what we frame as practical issues. If at all, published bias adjustment methods are often published as individual software packages across different programming languages (mostly R and Python) that do not allow users to adapt aspects of the method, such as the fit distribution, to their use-case. Existing open-source software packages, such as ISIMIP3BASD or CDFt, often do not offer an evaluation framework that covers multivariate (spatial, temporal, multi-variable) aspects necessary to detect misuse of methods, or user-specific impact metrics. Several of these issues apply to downscaling similarly.

To address some of these practical issues, we developed the open-source software package ibicus in collaboration with ECWMF (available on PyPi, extensive documentation https://ibicus.readthedocs.io/en/latest/index.html, published under Apache 2.0 licence). The package implements eight peer-reviewed bias adjustment methods in a common framework. It also includes an extensive evaluation framework covering multivariate aspects as well as the ETCCDI climate indices. The package thereby contributes to enhanced flexibility and ease-of-use of better evaluation practises in bias adjustment.

Our contribution presents three case studies using ibicus, highlighting a number of pitfalls in the usage of bias adjustment for climate impact modelling, and shows possible ways to address these issues. We investigate extreme indices of precipitation and compound extreme temperature-precipitation indices, modification of the climate change trend, and dry spell length as an example of a temporal index, over northern Spain and Turkey.

We evaluate how bias adjustment adds to the ‘cascade of uncertainty’ and how this can be made transparent in the different use-cases. We also demonstrate how some of the fundamental issues that can arise when applying bias adjustment can be detected and how evaluation of spatial and temporal aspects such as dry spell length can be made specific to the use-case at hand to detect improper use of bias adjustment. Lastly, we demonstrate how the ‘best’ bias adjustment method may depend on the metric of interest, and therefore a user-centric design of comparison and evaluation methods is necessary.

How to cite: Spuler, F., Wessel, J., Cagnazzo, C., and Comyn-Platt, E.: Case studies in bias adjustment: addressing potential pitfalls through model comparison and evaluation using a new open-source python package, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14254, https://doi.org/10.5194/egusphere-egu23-14254, 2023.

Global and regional climate models are constantly improving the quality of their outputs with increasingly fine spatial and temporal resolutions. These products, which comprise, for instance, reanalysis, reforecast and forecast, can be used for several applications, such as boundary conditions for climate simulations, initial conditions for local weather forecasting, and reference datasets for environmental and energy uses. Nevertheless, many authors have pointed out that such climate models are not suitable for direct use in local applications due to the presence of biases between the model results and the metered data. At this aim, several statistical methodologies have been proposed to correct and downscale the climate models outputs and make it available also for local purposes. Therefore, the purpose of this contribution is to analyse the current state-of-the-art statistical bias correction methods on different time aggregation to assess the capabilities of these methods from monthly to hourly temporal scale.

This study is carried out on the Trentino- Alto Adige, which is an alpine region in north Italy equipped with several measuring weather stations, around 300. The temperature and precipitation observations have been then used to produce a reference dataset through the geostatistical interpolation method called kriging. Instead, ERA5-Land, the reanalysis of ECMWF, has been adopted for the bias correction analysis. Several methods have been tested comprising of univariate and multivariate method including: linear scaling, variance scaling, local intensity scaling, local power transformation, quantile mapping, quantile delta mapping, and multivariate bias correction methods such as MBCn, MBCp, and MBCr. The time scale investigated are monthly, daily and hourly aggregations.

The results show a general decreasing of the performance of all the bias correction methods with the increase in the time-frequency of the weather variables. In particular, the mean absolute error of the corrected daily temperature is 50% larger than the monthly one, and the same 50% increase in error is found between daily and hourly corrected data. The increase in error with decreasing temporal resolution is even more pronounced for the precipitation variable, which is known to be discontinuous with respect to temperature. Multivariate bias correction methods seem to have difficulty maintaining dependencies between variables in the case of high-frequency data.

Although the results on the hourly data are not so scarce, it is evident that more depth analysis of temporal high-resolution climate data is needed, including sub-hourly data in the future, and therefore become crucial to develop new methodologies capable of correcting sub-daily bias. In conclusion, with this work, the authors seek to support research in the direction of providing high-frequency weather data for local applications, which are crucial, for example, in hydrological simulations for the assessment of hydrogeological risks and the management of renewable energy in the electricity market.

How to cite: Menapace, A., Dhawan, P., Dalla Torre, D., Larcher, M., and Righetti, M.: Analysis of the statistical bias correction of ERA5-Land on different time aggregations in Trentino-Alto Adige, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15537, https://doi.org/10.5194/egusphere-egu23-15537, 2023.

How to cite: Legasa, M. N., Thao, S., Vrac, M., Casanueva, A., and Manzanas, R.: Extending A Posteriori Random Forests for Multivariate Statistical Downscaling of Climate Change Projections, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17077, https://doi.org/10.5194/egusphere-egu23-17077, 2023.

Understanding urban ecohydrological interactions is crucial for the assessment of ecosystem responses to climate change and anthropogenic influences, especially in heavily urbanized environments. Urban water bodies can enhance local biodiversity, with urban blue water infrastructure providing valuable ecosystem services that contribute to healthier and more sustainable environments. Because the urban water cycle is less resilient to extreme climate events, there is a need to better understand how biological flow paths interact with climate and hydrological dynamics. To that end, synoptic sampling of environmental DNA (eDNA) was carried out on four major rivers in Berlin, Germany (Spree, Erpe, Wuhle, Panke) on a weekly basis over the course of one year. In conjunction with climate and hydrological data, the spatial and temporal variations in planktonic microbial communities were assessed in order to identify the differences in ecohydrological interactions among urban streams. Preliminary results indicate that while the rivers Wuhle and Erpe harbour similar bacterial communities, the more urbanized rivers Panke and Spree each had a different taxonomic composition. All rivers show a clear seasonal signal, although with varying intensity and directions of change. To further disaggregate the seasonal ecological changes, we determined the relative influence of climate as well as water chemistry, land use and stream flow conditions on bacterial community composition. In future, the integration of eDNA with other ecohydrological tracers such as stable water isotopes will provide even more insights into the ecological and hydrological functioning of urban environments. Such a combination of ecohydrological tracers has wider implications not only for future urban planning but for mitigating the negative effects of climate change in urban environments and assessing the resilience of urban water bodies to future extreme events.

How to cite: Warter, M., Monaghan, M. T., Ring, A.-M., Christopher, J., Kissener, H. L., Funke, E., Soulsby, C., and Tetzlaff, D.: Spatio-temporal variations in environmental DNA within heavily urbanized streams in Berlin, Germany, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-107, https://doi.org/10.5194/egusphere-egu23-107, 2023.

Building-integrated photovoltaic technology (BIPV) has been proven as an effective way to increase renewable energy in the urban environment. Without occupying any land resources, this technology has great potentials for achieving low carbon in the economically developed cities. Due to the lack of modelling tools, the impact of BIPV window in the street canyon is not well understood. To fill the gap, we developed a new parameterization scheme for BIPV window, and incorporated it into building energy simulations coupled with a single-layer urban canyon model. Model evaluation suggests that the coupled model is able to reasonably capture the diurnal profiles of BIPV window temperature and power generation, building cooling load, and outdoor microclimate. Canyon aspect ratio, window coverage, façade orientation, and power generation efficiency are found to be the most critical factors in maximizing the power generation of BIPV windows. Simulation results of an office floor in three Chinese cities under different climate backgrounds show that Beijing has the greatest solar potential in south orientation for power generation, which is 1.5 times the power generation in Shenzhen and Nanjing. Compared to clear window, BIPV window has positive benefits when window coverage is greater than 60% in open canyon. With lighting energy saving and power generation, BIPV window consistently has positive benefits than wall materials. The benefit of BIPV windows is larger in Beijing, followed by Shenzhen and Nanjing. Under future climate forcing of year 2050, the net electricity benefit of BIPV window will be larger than 15%. Findings in this study provide guidance for BIPV application in the built environment, and cast light on the construction of sustainable and low-carbon neighborhoods.

How to cite: yang, J., Chen, L., and Zheng, X.: Quantifying the BIPV window benefit in urban environment under climate change: a comparison of three Chinese cities, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-970, https://doi.org/10.5194/egusphere-egu23-970, 2023.

Knowledge creation is essential to encourage regional development especially for areas in regional transition. The positive impact of university-enterprise cooperation on the process has been extensively discussed in previous studies, but a comprehensive picture of the mechanism has not been fully described, such as how academic researches promote practical projects. Modeling this complex nature has been dealt and required by both academia and planner. Toward this target, we propose an empirically founded agent-based model to demonstrate the collaborative network. It uses the literature and patents created in each project as a basis for measuring regional knowledge production and innovation. Based on the analysis of the observed data, a conceptual model of network-based university-enterprise cooperation was constructed with the help of NetLogo. It will simulate the whole process from academic research to industrial practice to explore the driving mechanism of universities for regional knowledge production and innovation. The collaborative network of practice projects is constructed from the partnership data of every project recorded within the REVIERa platform, where each node in the network was classified into fields of research. Knowledge was quantified by metrics such as the quantity and quality of literature. Based on the various characteristics of nodes, networks and their path dependencies, the birth of innovative projects will be simulated and the impact of interdisciplinary on regional transformation will be quantified in the model. The model is based on the real world data and corroborated with it to capture the mechanism and characteristics of this complex process, showing its value to boost the scientific regional planning in the future.

How to cite: Feng, W. and Li, B.: A Network-Based Model for Simulating Regional Transformation Driven by University-Enterprise Collaboration, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1091, https://doi.org/10.5194/egusphere-egu23-1091, 2023.

Intercity patient mobility reflects the geographic mismatch between healthcare resources and the population, and has rarely been studied with big data at large spatial scales. In this study, we investigated the patterns of intercity patient mobility and factors influencing this behavior based on over 4 million hospitalization records of patients with chronic kidney disease in China. To provide practical policy recommendations, a role identification framework informed by complex network theory was proposed considering the strength and distribution of connections of cities in mobility networks. Such a mobility network features multiscale community structure with “universal administrative constraints and a few boundary breaches”. We discovered that cross-module visits which accounted for only 20 % of total visits, accounted for >50 % of the total travel distance. The explainable machine learning modeling results revealed that distance has a power-law-like effect on flow volume, and high-quality healthcare resources are an important driving factor. This study provides not only a methodological reference for patient mobility studies but also valuable insights into public health policies.

How to cite: Ding, J., Chao, Y., Wang, Y., Li, P., Wang, F., Kang, Y., Wang, H., Liang, Z., Zhang, J., Han, P., Wang, Z., Chu, E., Li, S., and Zhang, L.: Influential factors of intercity patient mobility and its network structure in China, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1770, https://doi.org/10.5194/egusphere-egu23-1770, 2023.

While there have been many studies on the diurnal and seasonal variations of urban heat islands (UHIs), the day-to-day variability of UHI remains largely unknown, despite being the key to explaining interactions between UHIs and extreme heat events or heatwaves. In this study, we aim to understand and quantify the persistence of urban/rural temperatures. Autocorrelation and spectral analyses are conducted on urban and rural temperatures simulated by a global land model to quantify the urban-rural difference of temperature persistence. A surface energy balance model is then derived to explain the simulation results, elucidating the key biophysical processes that contribute to the stronger persistence of urban surface temperature in certain areas.

How to cite: Li, D., Wang, L., and Sun, T.: Persistent urban heat, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1780, https://doi.org/10.5194/egusphere-egu23-1780, 2023.

Heating and cooling of buildings is one of the largest final energy uses and largest sources of greenhouse gas emissions. To reduce the impact of heating and cooling on our climate, more efficient strategies are needed. Coupling and centralizing the production of heat and cold in combination with underground seasonal thermal storage (UTES) can significantly reduce CO₂ emissions and costs. To plan and implement such strategies for heating and cooling, information on sources and sinks of heat and cold is essential for local authorities. However, spatial information on the cooling sector is rare and difficult to obtain. Often, the theoretical cooling demand of specific buildings and building types is modeled, but not met by air-conditioning equipment in reality. On the other hand, large-scale cooling demand models, which focus on entire countries, may use data from different countries as proxy or are not applicable below kilometer-scale.

In this study, we present a method to identify air-conditioning equipment on the rooftops of buildings and quantify their cooling capacity. Thus, air-cooled and hybrid evaporative condensers, cooling towers and packaged rooftop units are detected on aerial images. Using manufacturer data, regression analyses are created to estimate the cooling capacity based on the size of the units and the number of condenser fans. The unit locations and all required parameters are obtained by convolutional neural network-based pixel classification models, which are easily executable within a geographical information system (GIS) framework. The approach is successfully evaluated by testing the capability of the detection models and comparing our estimated cooling capacities to the actual installed cooling capacities of air-conditioners for different locations. The detection performance strongly depends on the resolution of the used aerial images. At a resolution of 8 cm/pixel, the model detects 93% of the units and the pixel classification overestimates the relevant parameters for the regression by 0.7%. Using the regression analyses, the overall capacity in the evaluated areas is overestimated by 7-21%. To demonstrate the capability of our approach, we map the cooling capacity of air-conditioners in parts of Manhattan. In the Manhattan financial district alone, a cooling capacity of over 2 GW is estimated, which is equivalent to 1.3% of the summer peak load demand of the energy grid of the entire state of New York.

The presented approach is a fast and easy to conduct method that requires little input data. It can detect individual air conditioners over large areas. The obtained information can support the creation of cooling cadastres and can serve as supplement or validation for other cooling demand models, such as building stock models, or example to include additional building types, such as industrial buildings.

How to cite: Barth, F., Schüppler, S., Menberg, K., and Blum, P.: Estimating installed cooling capacities on city scale, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1969, https://doi.org/10.5194/egusphere-egu23-1969, 2023.

Lead (Pb) exposure to residents of impacted communities depends on the environmental concentrations of lead that is potentially bioaccessible. Such concentrations are the result of a complex array of interactive factors that influence one another through direct or indirect linkages. Current models to predict the health impact of Pb exposure often do not consider the complexity from a system perspective. Therefore, there exists a great need to develop a holistic modeling strategy to simulate the risk to Pb exposure and resulting blood Pb concentrations based on bioaccessible Pb concentrations in the environment and how socioeconomic status, policy/ scholarly intervention, and collective community behaviors influence that concentration. Our study attempts to develop a grey system integrated system dynamics simulative modeling framework to simulate general Pb bioaccessibility in the environments and how it transmits from the soil, the water, the house, and the general environment to human bodies. The study aims to predicting risk of Pb exposure in the long run in a community/neighborhood, especially the risk to vulnerable populations, such as young children and the aged population. The model also aims to identify the most effective ways to curb human exposure to bioaccessible Pb. This is the first stage of a multi-stage research activity. In this stage, the study focuses on developing a theoretical and empirical modeling framework of the simulative model, and the data structure. In this study, we take a macro perspective to treat a neighborhood/community, a city, or a designated area as an integrated and dynamic system in that it is composed of many interrelated, feedback-linked components. Each component exists and acts because of its interaction with other system components, both observable and hidden. The integrated mutual interaction and multiple components collectively determine how the system will change and evolve in the future, manifested as the change and evolution of the various system components. Bioaccessible Pb in the neighborhood’s environment is our key system component. The grey system and system dynamic simulative model attempts to analyze the potential interactive co-variations among different system components, or the changing and/or evolving trend a system component demonstrates over time. By establishing the interactive feedback loops that connect all the observable system components with sufficient data, the model will be able to simulate the dynamics of the system to predict its behavior and manifestation in the future. Since the simulative model is built upon the interactive feedback loops among all system components, the model will produce simulated results for all observable system components as well. Our goals in later stages are to predict the potential damage to human bodies that will be the basis for Pb reduction and removal related policies at the macro management levels.

How to cite: Yu, D.: A system dynamics simulative modeling framework to assess bioaccessibility of lead and facilitate lead reduction in the urban environment, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2461, https://doi.org/10.5194/egusphere-egu23-2461, 2023.

Accurate and agile modelling of weather, climate, hydrology and air quality in cities is essential for delivering integrated urban services. SUEWS (Surface Urban Energy and Water balance Scheme) allows simulation of urban–atmospheric interactions by quantifying the energy, water and carbon fluxes.  SuPy (SUEWS in Python) provides the SUEWS computation kernel, a Python-based data stack that streamlines pre-processing, computation and post-processing to facilitate common urban climate modelling. This paper documents the recent developments in both SuPy and SUEWS, and the background principles of their interface, F2PY (Fortran to Python) configuration and Python front-end implementation. SuPy is deployed via PyPI (Python Package Index) allowing an automated workflow for cross-platform compilation on all mainstream operating systems (Windows, Linux and macOS). The online tutorials, using Jupyter Notebooks, allow users to become familiar with SuPy. A brief overview of other complementary SUEWS developments will be given, and include within canopy layer profiles of temperature, humidity, wind, and radiation that are supporting a wide range of applications; and database developments for obtaining model parameters.

How to cite: Sun, T., Grimmond, S., Backlin, O., Bluun, L., Hogan, R., Stretton, M., and Xie, X.: SuPy and SUEWS urban land surface modelling: new developments and capabilities, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2970, https://doi.org/10.5194/egusphere-egu23-2970, 2023.

In response to climate change and growing ecological threats, many cities are planning to increase the resilience of urban drainage systems, including the reduction of combined sewer overflows (CSOs) - one of the leading causes of surface water pollution. Blue green infrastructure (BGI) are growing in popularity to do so, and recent studies have made progress to evaluate the potential of BGI to eliminate CSOs. However, current research tends to consider a limited number of individual BGI elements and scenarios, often overlooking different combinations (e.g., bioretention basins combined with green roofs) and uncertainty in a future climate. The aim of this research is to evaluate the ability of a range of blue green infrastructure combinations to reduce CSOs under multiple future climate scenarios.
A hydrological simulation model, EPA SWMM, is used to simulate the performance of a 95-hectare combined sewer system near Zurich, Switzerland. Four types of BGI are evaluated, including bioretention basins, porous pavements, green roofs, and stormwater ponds. The potential surface availability for each BGI element was quantified using GIS and LiDAR data, yet scenarios include a range of different implantation rates for each type. Combinations of BGI element types are generated by combining different implementation surfaces to the share of the BGI type (e.g., 20% of the available surface with the same share of bioretention basins, porous pavements and green roofs, etc.).
Bioretention basins are assumed to be implemented on pervious surfaces (i.e., gardens, traffic islands), porous pavements on impervious surfaces (i.e., sidewalks, cycling lanes) and green roofs on flat roof buildings. Observed rainfall data (1990-2019) are used to simulated the baseline conditions, while more than five bias-corrected future rainfall timeseries (2070-2099) from EURO-CORDEX regional climate models (RCP 8.5) are used to represent a worst-case future climate. CSO Volume, duration and frequency are used to characterize system-wide CSO events across the seven outfalls.
Preliminary results show that in a current climate, bioretention basins are most effective at reducing CSO volume, followed by porous pavements and green roofs. BGI do not relevantly reduce the duration and number of CSO events. In one future scenario, future precipitation is concentrated into shorter duration events, which consistently leads to shorter, higher intensity CSO events at a frequency similar to the historical record. Overall, the only scenario that can avoid an increase in future CSO volume is an extensive implementation of bioretention basins. Porous pavement and green roofs are less effective in a future climate because they can store limited amounts of water compared to bioretention basins. As rainfall intensities increase, the ability to retain large amounts of water will be the most effective. These results point to strategies with higher storage capacities to account for high-intensity rainfall events that are expected in the future. Future work will evaluate additional BGI elements, including urban ponds, and a more comprehensive set of BGI scenarios, future climate scenarios, and case studies, enabling a definition of guidelines and BGI design requirements at an urban scale for Switzerland.

How to cite: Cavadini, G. B. and Cook, L.: Blue Green Infrastructure in a future climate: can we reduce combined sewer overflows?, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3148, https://doi.org/10.5194/egusphere-egu23-3148, 2023.

Urbanization has been shown to significantly increase the frequency and intensity of extreme weather events, i.e., extreme precipitation events, and heatwave events. Actually, the occurrence of compound extreme events, such as sequential flood-heatwave (SFH) events, can lead to more severe impacts than singular extreme events. However, the impact of urbanization on these compound extreme events is not well understood. In this study, we examine urbanization effects on the growth of SFH events in the Guangdong-Hong Kong-Macao Greater Bay Area from 1961 to 2017. We classify stations into urban and rural types based on dynamic land use data and define SFH events using the criteria from previous studies. We find that the frequency of SFH events in urban stations increased from 0.218 events per year before the 1990s to 1.401 events per year after the 1990s, while the frequency of SFH events in rural stations increased from 0.250 events per year to 0.920 events per year. The urban impact of 0.131 events per decade also shows that urbanization can promote the occurrence of SFH events. Our analysis also indicates that urbanization promotes the growth of SFH events mainly by increasing the frequency of heatwave (HW) events. These findings highlight the need for further research on the effects of urbanization on compound extreme events and the development of effective management strategies to reduce their risks.

How to cite: Liao, W. and Zheng, K.: Urbanization impacts on sequential flood-heatwave events in the Guangdong-Hong Kong-Macao Greater Bay Area , China, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3807, https://doi.org/10.5194/egusphere-egu23-3807, 2023.

Variability of building height induces flow heterogeneity and directly controls the depth of the roughness sub-layer, the strength of mutual sheltering, and the overlapping of urban canopy flow, which poses challenges for accurate modeling. Large-eddy simulations over 96 building arrays with varying density, height variability (standard deviation of building height), and horizontal arrangements were conducted to reveal the impact on the urban flow. Results demonstrate a strong non-local building effect on the flow due to height variability, where flow around high buildings possesses high wind speed, dispersive momentum flux, and other distinctive flow patterns, whereas around low buildings, the flow pattern is less unique. The complex flow behavior is beyond the capacity of the current multi-layer urban canopy model (MLUCM) where turbulent constants and drag effects were considered in a simplified way. The increased height variability and urban density also blur the interface of urban canopy, further making MLUCM estimates model constants heavily based on a clear urban canopy inappropriate. Based on the original model, we comprehensively tested potential contributing factors such as the estimation of displacement height, height-dependent drag coefficients, and the extended roughness sublayer. The modified model provides a better overall agreement with the LES results, especially above the mean building height where the prediction of the extended urban canopy layer is largely improved. 

How to cite: Lu, J., Nazarian, N., Hart, M., Krayenhoff, S., and Martilli, A.: Urban canopy parameterization of the non-local building effects with variable building height, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3877, https://doi.org/10.5194/egusphere-egu23-3877, 2023.

Buildings are common components in the urban environment whose 3D information is fundamental for urban hydrometeorological modeling and planning applications. In order to monitor building footprint and height across large areas on a regular basis, recent earth observation research has witnessed promising progress in mapping such information from publicly available satellite imagery by statistical methods using regression between multi-source remotely sensed data and target variables. However, most of them often involve tedious feature preprocessing, which constrains their capability to establish a comprehensive representation of an ever-changing and multi-scale urban system efficiently.

Considering this bottleneck, this work develops a deep-learning-based (DL) Python package-SHAFTS (Simultaneous building Height And FootprinT extraction from Sentinel Imagery) to estimate 3D building information at various scales. SHAFTS provides Convolutional Neural Networks (CNN) with the Multi-Branch Multi-Head (MBMH) structure to automatically learn representative features shared by building height and footprint mapping tasks from multi-modal Sentinel imagery and additional background DEM information. Besides, to leverage the power of big data infrastructures, SHAFTS offers essential functionality including automatically collecting potential reference datasets by web scraping and filtering appropriate input imagery from Google Earth Engine, which can effectively ease model upgrading and deployment for large-scale mapping.

To evaluate the patch-level prediction skills and city-level spatial transferability of developed models, this work performs diagnostic performance comparisons in 46 cities worldwide by using conventional machine-learning-based (ML) models and CNN with the Multi-Branch Single-Head (MBSH) structure as benchmarks. Patch-level results show that DL models successfully produce more discriminative feature representation and improve the coefficient of determination of building height and footprint prediction over ML models by 0.27-0.63, 0.11-0.49, respectively. Moreover, stratified error assessment reveals that DL models effectively mitigate severe systematic underestimation of ML models in the high-value domain. Additionally, within the DL family, comparison in spatial transferability demonstrates that the MBMH structure improves the accuracy of CNN and reduces the uncertainty of building height predictions in the high-value domain at the refined scale. Therefore, multi-task learning can be considered as a possible solution for improving the generalization ability of models for 3D building information mapping.

How to cite: Li, R., Sun, T., Tian, F., and Ni, G.: SHAFTS – A deep-learning-based Python package for Simultaneous extraction of building Height And Footprint from Sentinel Imagery, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4192, https://doi.org/10.5194/egusphere-egu23-4192, 2023.

Extreme high-temperature phenomena in urban areas are recognized as natural disasters. Asphalt roads and high-rise buildings are concentrated in megacities, and buildings and roads may contribute to extremely high air temperatures, especially in the summer season. For the improvement of the thermal environment in urban areas, green spaces are being created on building roofs. In this study, we investigated the effects of roof greening on street-canyon flows in the presence of roof and ground heating using a computational fluid dynamics (CFD) model. For validation, we compared the simulated street-canyon flows to the measured ones in a wind tunnel experiment. The CFD model used in this study reproduced the wind speeds, turbulent kinetic energies, and air temperatures in a street canyon in the presence of building roof and ground heating.

How to cite: Rho, J.-H., Mun, D.-S., Wang, J.-W., and Kim, J.-J.: Effects of roof greening and surface heating on street-canyon flows, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4807, https://doi.org/10.5194/egusphere-egu23-4807, 2023.

While road transport is a cornerstone of modern civilization bringing profound positive impacts to the economy and human well-being, it is also associated with several undesirable and unsustainable outcomes including urban air pollution, climate change, noise and congestion. Hence, sustainable road transport has been given significant attention in the past two decades. Digital twins of urban transport are promising digital assets to evaluate and improve the sustainability level of the transport systems.  Digital twins provides a testbed whereby the impacts of the current and future policies and strategies can be modelled and analysed in a digital environment, helping ensure that tax money spent delivers the expected results. However, the desperate shortage of spatiotemporal road data is the major challenge in establishing data flow between the digital and physical twins of road transport. Vehicle telematics data, typically collected from GPS-connected, can provide an excellent source of intormation with which to address the spatial and temporal aspects of transport data. This presentation will highlight how telematics data can be used within road transport digital twins.

In this study, we develop digital twins of road transport for Tyseley Environmental Enterprise District (TEED) a small area of east Birmingham, UK, using the newly-developed approach of GeoSpatial and Temporal Mapping of Urban Mobility (GeoSTMUM). GeoSTMUM uses vehicle telematics (location and time) data to estimate several road transport characteristics such as the average speed of traffic flow, travel time, etc., with high spatial and temporal resolutions of 15m and 2h, respectively. It also allows for evaluation of the average vehicle dynamic status as the speed-time-acceleration profile of the roads. Vehicle telematics data for this study were collected for the years 2016 and 2018 through the WM-AIR projct (www.wm-air.org.uk). We then use real-world fleet composition and exhaustive emission measurements to translate the vehicle dynamics status into the real-urban fuel consumption and CO₂ and NO_x emission factors. Results highlight the importance on fleet renovation, in terms of vehicle propulsion systems (EURO class, fuel type, etc.) upon real-urban emissions and fuel consumption. The presentation will end with example future use cases of telematics data within digital twins.            

How to cite: Ghaffarpasand, O. and Pope, F.: Using vehicle telematics data within a digital twin of urban transport systems; a case study in the West Midlands, UK, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5315, https://doi.org/10.5194/egusphere-egu23-5315, 2023.

Temperatures are rising and the frequency of heat waves is increasing due to anthropogenic climate change. At the same time, the population in urban areas is rapidly growing. As a result, an ever-larger part of humankind will be exposed to even greater heat stress from heat waves in urban areas in the future. In this research, we focus on studying the determinants of land surface temperature (LST) gradients in urban environments. We implement a spatial regression model that is able to predict with high accuracy (R² > 0.9 in the test phase of k-fold cross-validation) the LST of urban environments across 200 cities based on land surface properties like vegetation, built-up areas, and distance to water bodies, without any additional climate information. We show that, on average, by increasing the overall urban vegetation by 3%, it would be possible to reduce by 50% the exposure of the urban population that lives in the warmest areas of the cities for the average of the three summer months, achieving a reduction of 1 K in LST. By coupling the model information with the population layer, we show that an 11% increase in urban vegetation is necessary in order to obtain a reduction of 1 K in the most populated areas, where at least 50% of the population live. We finally discuss the challenges and the limitations of greening interventions in the context of available surfaces in urban areas.

How to cite: Massaro, E., Caporaso, L., Piccardo, M., Schifanella, R., Taubenböck, H., Cescatti, A., and Duveiller, G.: A spatial regression model to measure the urban population exposure to extreme heat, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6583, https://doi.org/10.5194/egusphere-egu23-6583, 2023.

The urban transport network is threatened by urbanisation and climate change-enhanced urban flood, leading to substantial impacts on economic activities, social well-being and the environment. By taking the flood propagation process into consideration, we developed a flood-impact-assessment method to comprehensively assessed the economic impacts of traffic disruption in terms of time delay, fuel consumption and pollutant emission. The flood is simulated with CADDIES-2D flood model and the traffic flow is simulated with a microscopic model (SUMO). We applied this method to Beijing and quantified the economic damage of various flood scenarios. Comparing the baseline traffic scenario with those of three flooded scenarios yields the impacts of floods on traffic. The study revealed three key findings: (a) a rain occurring at 7 a.m. induces four times more cost than the baseline scenario, while rain of the same intensity and duration occurring at 8 a.m. or 9 a.m. lead to a traffic cost increase for 37.33% and 13.21% respectively. (b) The central and southern parts of Beijing suffer more from flooding and should be given priority for adaptation planning. (c) There is no significant spatial correlation between flood depth and traffic cost increase on a census block level. The proposed framework has the potential to assist decision-makers in prioritizing flood mitigation investments and therefore increase the resilience of transport networks to flooding impacts.  

How to cite: Liu, Y., Yang, S., Dawson, R., Ford, A., and Tang, J.: Flood Impact to Urban Transport Networks Considering the Flooding Propagation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6611, https://doi.org/10.5194/egusphere-egu23-6611, 2023.

Road traffic is usually the most pervasive source of noise in an urban environment. Epidemiological studies conducted at the regional or national scale have shown associations of road traffic noise with sleep disturbance, cardiovascular diseases, and mental health problems. Strategic noise mapping (European Noise Directive) only covers populations living in large urban areas. The limited coverage of harmonised noise exposure data at a pan-European scale prevents us from studying the effect of road traffic noise on health in larger populations across Europe. Therefore, this study aims to develop models capturing within-city, intra-city and national variations in road traffic noise exposures across Europe to facilitate pan-European multi-cohort health studies. To estimate noise, we used a simplified version of CNOSSOS-EU (Common NOise aSSessment MethOdS) noise modelling framework.   

The CNOSSOS-EU model requires a range of input data, including a detailed road network, traffic intensity, traffic speed, and land use data including building footprints. We used OpenStreetMap (OSM) to define the road network and buildings. Because traffic intensity is not provided in OSM, we estimated Europe-wide annual average daily traffic (AADT) counts using random forest trained by observations collected in Austria, Switzerland, Germany, France, Italy, and the United Kingdom. Three random forest models were built separately for 1) motorway and trunk roads, 2) primary roads, and 3) secondary, tertiary, residential and unclassified roads defined in OpenStreetMap (OSM). Predictor variables included road length, sizes of residential areas, and population within different circular buffer (ranging from 100m to 200km). The models were validated using 5-fold cross-validation. The 5-fold root mean square errors of AADTs were 19646, 6589, 4005, 3824 and 3210 for highway (motorway and trunk roads), primary, secondary, tertiary, and residential roads. The traffic speed was approximated by the speed limit from OSM, and the missing speed limit data was replaced by the legal country-specific speed limit separated by inside and outside built-up areas, depending on the road type. Building height was approximated by using a morphological operation on the AW3D30 digital surface model (DSM). The road traffic noise was estimated at noisiest building façades (i.e., with shortest Euclidean distance to nearby roads within 100m with the highest AADT) using CNOSSOS-EU. The modelled noise level of L_Aeq16 with these input data ranged from 52.17 dB to 72.54 dB for points in the test city of Bristol in the United Kingdom. In conclusion, we developed the input data required for noise modelling, especially traffic intensity, at a European scale. Modelled noise will be used in Europe-wide studies of health effects of noise. We will also compare our Europe-wide noise estimates with national noise model estimates in the Netherlands and Switzerland.

How to cite: Shen, Y., de Hoogh, K., Schmitz, O., Gulliver, J., Karssenberg, D., Vermeulen, R., and Hoek, G.: Europe-wide road traffic noise modelling using a harmonized methodological framework (CNOSSOS-EU), EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9249, https://doi.org/10.5194/egusphere-egu23-9249, 2023.

It is predicted that over 60% of the urban population in sub-Saharan Africa (SSA) lives in slums, and this number is increasing in the coming years. However, issues on urban poverty and slum persistence in SSA cities are ignored and suffered from a paucity of robust evidence for a longtime. As reliable data on where are locations of urban deprived areas and slums, and on how these areas have evolved remain scarce, the scale of urban deprivation and challenges related to slums in SSA cities are underestimated.

This study explores to which extent urban morphology and accessibility of social services could explain urban poverty and slum locations, by using geospatial and socio-economic data, as well as machine learning techniques. Taking four African countries including Nigeria, Kenya, Ghana, and Malawi as examples, we mapped slum locations and demonstrate that urban building morphological variables only can explain up to over 78% of slum locations. Our results further showed a declining trend in slum growth in old towns that are compacted in space. However, slums are not representing the most deprived urban area, while outskirts of megacities, middle-sized and small cities showed the least economic well-being, demonstrated by lower GDP and wealth index value; poor road and water access services. Our proposed slum and urban poverty mapping methods and results will be accessible and instrumental for scientists, local communities, policy-makers, and city planners, which will accelerate the process of finding solutions for tackling poverty, better managing public health and infrastructure in developing countries.

How to cite: Li, C., Yu, L., and Hong, J.: Monitoring slum and urban deprived area in sub-Saharan Africa using geospatial and socio-economic data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-10872, https://doi.org/10.5194/egusphere-egu23-10872, 2023.

As cities continue to expand it has become crucial to describe their evolution in time and space. Building on analogies with biological systems, we propose a minimalist reaction-diffusion model coupled with economic constraints and an adaptive transport network, describing the co-evolution of population density with the transport system. Using a unique dataset, we reconstruct the evolution of London (UK) over 180 years and show that after an initial phase of diffusion limited growth, population has become less centralised and more suburban in response to economic needs and an expanding railway network. The coevolution of the rail system with a growing urban population has generated a transport network with hierarchical characteristics which have remained relatively constant over time. These results show that urbanisation patterns largely depend on the evolution of transport systems and population-transport feedbacks should be carefully considered when planning and retrofitting urban areas.

How to cite: Capel-Timms, I., Levinson, D., Bonetti, S., and Manoli, G.: Modelling the coevolution of London's population and railway system, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-12606, https://doi.org/10.5194/egusphere-egu23-12606, 2023.

The urban environment is made up of a complex and changing mosaic of territories. To what extent these spatial heterogeneities, between territories, determine the quality of life or social, economic or health inequalities? Within the framework of the ANR EGOUT project (egout.cnrs.fr), we assume that the spatial distribution and temporal evolution of tracers archived in the sediments of the Parisian sewerage networks can help deciphering the diversity of aboveground conditions and their temporal trajectory.

We compiled the geochemical results acquired before cleaning out operations on sediments accumulated in more than 100 silt traps (STs) that line the sewerage network of the City of Paris. These STs receive sediments that transit through the Parisian combined (wastewater and stormwater) sewer system. These analyses concern granulometry, metals, polychlorinated biphenyls (PCBs), but also 16 polycyclic aromatic hydrocarbons (PAHs). These regulatory analyses (which guide the nature of the treatment processes to be implemented) have been available since the year 2000, with cleaning out and thus measurement frequencies varying from one ST to another. They therefore allow addressing not only geochemical spatial disparities but also their temporal evolution.

In order to assign these results to the corresponding catchment areas for every ST, we first defined the catchment areas of each DT. The TIGRE 7 information system of the City of Paris was used to distinguish each sub-network draining the sediments to each ST. Two spatial scales of drainage (wet and dry weather), but also a sedimentary cascade system could be highlighted. The catchment areas of each ST were then defined by linking individual connections to individual addresses and cadastral parcels.

Here are the most striking results from the exploitation of existing data:

• The Haussmannian buildings, which are present for the most part in the city Centre, are major sources of zinc emissions (Gromaire et al., 2001). This element is found in the sediments of DT draining areas with a high density of historical buildings.
• Based on concentration ratios (Ayrault et al., 2008), PAHs mainly result from road traffic. The concentration of PAHs has been decreasing in ST sediments since 2000. This decrease could reflect the 59% decrease in the car traffic in Paris recorded between 2001 and 2018.
• PAH levels and types differ from DT to another, as already noted by Rocher et al. (2004). These differences probably indicate local specificities in PAH production of each catchment area. Cross-referencing our data with other spatialized data related to potential PAH sources (road traffic, heating, etc.) should allow us to better understanding the factors that control their presence in sewer sediments.

How to cite: Asselin, C., Jacob, J., and Moilleron, R.: Controls on the spatial distribution and temporal variation of anthropogenic tracers in the sediments of the Paris sewerage system., EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13043, https://doi.org/10.5194/egusphere-egu23-13043, 2023.

OpenStreetMap (OSM) is the largest crowd-sourced mapping effort to date, with an infrastructure network that is considered near-complete. The mapping activities started as any crowd-sourced information platform: the community expanded OSM anywhere there was a collective interest. Initial efforts were found around universities, hometowns of mappers and areas designated by organizations like the Humanitarian OSM Team (HOT). This resulted in a map that is of non-uniform completeness, with some areas having all building footprints in, while other areas remain incomplete or even untouched. Currently, with 530 million footprints, OSM identifies between a quarter and half of the total building footprints in the world, if we estimate that there are around 1-2 billion buildings in the world.

A global view on the completeness of buildings existing in OSM did not yet exist. Unlike other efforts, that only look at a subset of OSM building data (Biljecki & Ang 2020; Orden et al., 2020; Zhou et al., 2020), we have used the Global Human Settlement Layer (GHSL) to estimate completeness of the entire dataset. The remote sensing dataset is distributed onto a grid and in each tile of the grid, the built area of GHSL is compared to the total area of OSM building footprints. The computed ratio is measured against a completeness threshold that is calibrated using areas that were manually assessed.

Using information derived from remote sensing datasets can be problematic: GHSL does not only measure building footprints: it includes any human-built structures, including infrastructure and industrial areas. Next to that, due to circumstances like imperfect input data or failing algorithms, the dataset is not of the same quality as the crowd-sourced data in OSM in areas that are complete. False positives (i.e. rocky coasts) and false negatives (i.e. buildings missing in mountainous areas) exist in automatically generated data.

Even with these limitations, a comprehensive global completeness assessment is created. The assessment should not be used as ground truth, but rather as reflection on the OSM building dataset as is and as a guideline for priorities for the future. Statistics on regional completeness can be created and the quality of GHSL could be assessed on countries that are considered to be complete, such as France or the Netherlands.

How to cite: Oostwegel, L. J. N., Garcia Ospina, N., Evaz Zadeh, T., Shinde, S., and Schorlemmer, D.: Automatic global building completeness assessment of OpenStreetMap using remote sensing data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13160, https://doi.org/10.5194/egusphere-egu23-13160, 2023.

The complex nature of brownfield sites means that making decisions about their regeneration can be challenging and involve a wide range of stakeholders. To support these stakeholders, data-driven spatial decision support systems (DSSs) are often used. For these kinds of tools to be effective, it is crucial to have a comprehensive understanding of the problems and challenges faced by stakeholders. This research builds on previous large-scale stakeholder engagement (Hammond et al., 2023), critical review (Hammond et al., 2021), and detailed user requirements gathering. We present a framework for the development of a novel web-based DSS to support early-stage city region-scale brownfield planning and redevelopment. The DSS has four objectives: (1) improve the findability and visualisation of data, (2) support the assessment and understanding of ground risk posed by contamination and geotechnical instability, (3) provide better visualisation of data related to economic viability assessment, and (4) support evidence gathering for master planning through the modelling of land-use potential using a GIS multi-criteria method. We demonstrate the capabilities of the DSS framework through the implementation of a case-study focussing on the Liverpool City Region, a combined authority area in north-west England. Findings from user testing of the DSS and verification work are also presented.

How to cite: Hammond, E., Coulon, F., Hallett, S., Thomas, R., Dick, A., Hardy, D., and Beriro, D.: Development of a decision support system for regional planning and the assessment of brownfield sites: A case-study, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13747, https://doi.org/10.5194/egusphere-egu23-13747, 2023.

Air pollution affects the economy, the environment, and public health. This is particularly relevant in dense urban areas due to their urban built, high traffic activity, and near-the-source population exposure. In the city of Barcelona, the 40 ug/m³ nitrogen dioxide NO₂ annual limit value set up by the 2008/50/EC European Air Quality Directive (AQD) is systematically exceeded in traffic stations mainly due to the contribution of road transport. In the last Urban Mobility Plan (2019-2024), the city hall of Barcelona presented several traffic management strategies aiming to reduce on-road traffic emissions by both renewing and reducing the private motorized transport in the city. These measures include the application of tactical urban actions, green corridors and superblocks along with a Low Emission Zone, which together are expected to reduce the number of private vehicles circulating throughout the city by -25%. In parallel, the Port of Barcelona has recently announced a plan to electrify the docks and reduce emission from hotelling activities by -38%. To properly assess the impact of such measures, the AQD recommends the application of numerical models in combination with monitoring data. Following AQD recommendations, our study runs a coupled transport-emission model able to characterize traffic movement along the city and produce multiple scenarios that quantify the impact of the aforementioned measures on primary emissions. The resulting scenarios are then used to feed a multi-scale air quality modeling system to estimate NO₂ concentration values at very high resolution (20m, hourly). To reduce the uncertainty typically associated with modeling results, the estimated values are corrected with a data-fusion methodology using observations from the official monitoring network and several measurement campaigns. Our results show that the implementation of all mobility restrictions and electrification of the Port will allow Barcelona to comply with the current legislated NO₂ air quality standards at the traffic monitoring stations, with reductions up to -24.7% and -12 ug/m³. However, the resulting NO₂ levels achieved at these locations would still fail to meet the new 2021 WHO guideline (10 ug/m³) and the recent proposal for a revision of the EU AQD (20 ug/m³). Also, despite the estimated NO₂ reductions, several areas in the city would still be above the current legal limit of 40 ug/m³, including 16.7% of schools and 19.7% of hospitals and healthcare facilities. All in all, our results suggest the planned measures are steps in the right direction, yet still insufficient to ensure healthy AQ values across the entire city.

How to cite: Rodriguez-Rey, D., Guevara, M., Armengol, J. M., Criado, A., Enciso, S., Tena, C., Benavides, J., Soret, A., Jorba, O., and Pérez-Garcia Pando, C.: Challenges for achieving clean air - The case of Barcelona (Spain), EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14487, https://doi.org/10.5194/egusphere-egu23-14487, 2023.

In the context of climate change adaptation, sustainable urban development, and environmental justice, local civil services must meet a range of dynamic demands. To do so, municipalities require both qualitative and quantitative knowledge about the current state and the development of urban structures such as impervious surfaces and vegetated areas within their boundaries. However, obtaining this data through surveys is costly and time-consuming, and the frequency of these surveys is too low to capture changes consistently. As a result, data availability to support urban planning strategies often depends on financial priorities and is not consistently available to all local authorities in Germany. Remote sensing data, which has extensive spatial coverage and is regularly available, offers a more viable option for effective monitoring of urban structures. However, local authorities often lack knowledge about the benefits and limitations of using such data. The UrbanGreenEye project aims to bridge this gap by developing urban climate indicators based on Earth Observation data that meet the needs of local authorities. Drawing on the experience of nine partner municipalities, the project will demonstrate the use and implementation of these indicators in planning processes and strategies. It will also help create digital twins for urban planning applications and provide a free, regularly updated indicator-geodata foundation for Germany to support decision-making, particularly for climate change adaptation. The indicators will help identify locations experiencing high thermal and hydrological stress and quantify the relief provided by vegetated and pervious areas. Land surface temperature (LST) derived from satellite data from the US Landsat program will be used to monitor thermal stress, while the urban green volume and vegetation vitality indicators, derived from EU Copernicus Sentinel-2 satellite data, will contribute to thermal stress relief. The imperviousness indicator will also be derived from Sentinel-2 data using spectral models. Artificial intelligence algorithms, including Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) and attention-based transformer models, will be used to extract complex information from the urban surfaces, which require large amounts of reference data to capture necessary details. This reference data will be generated from high-resolution aerial images using CNN, supported by local ground truth data from municipal authorities and citizen science projects. It will then be upscaled and used as a reference for satellite-level models to provide nationwide consistent products. The satellite-based indicators will be validated for error ranges and at different spatial scales using the micro-scale climate model PALM-4U. Eventually, the indicators will be used to create a model for urban green volume deficiency to identify hot spots for adaptation measures and support planning strategies.

How to cite: Wagner, K., Frick, A., Stoeckigt, B., Gey, S., Lehmler, S., Scholz, N., Loeffler, F., Engnath, V., Heiland, S., Schubert, S., and Salim, M.: Monitoring Urban Areas for Climate Change Adaptation Using Remotely Sensed Indicators in the UrbanGreenEye Project, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15135, https://doi.org/10.5194/egusphere-egu23-15135, 2023.

Building energy use is one of the largest global demands,  accounting for 36% of final energy use and 39% of energy and process-related carbon dioxide (CO2) emissions. Green plans are the recommended planning technique for reducing the energy demand of buildings without changing the current built environment. This study investigates the effect of neighbourhood features on the energy performance of buildings. On the basis of building age and tree density, four typical Dublin city centre neighbourhoods are chosen to generate simulations. Surface Urban Energy and Water Balance Scheme (SUEWS) was used to generate the forcing climate data surrounding the neighbourhood, which was then fed into Integrated Environmental Solutions Virtual Environment (IES VE) as the meteorological data for conducting building energy simulations. The results showed that the fraction of trees plays an important role in wind speed in neighbourhoods. Incorporating the missing neighbourhood signature into the forcing data for building energy modelling improves the simulation's efficiency and precision. This study illustrates the importance of considering the local climate while simulating building energy efficiency.

How to cite: Ren, Z., Mills, G., and Pilla, F.: Urban meteorological forcing data for building energy simulations at a neighbourhood scale, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15145, https://doi.org/10.5194/egusphere-egu23-15145, 2023.

According to the WHO, in 2019, 23% of global mortality was attributable to environmental risk factors with significant gradients related to social factors, at the origin of significant health inequalities in the cities. Air pollution is one main environmental risk in urban spaces, and must therefore also be understood by taking into account the issue of socio-spatial inequalities.

Realistic modelling of population exposure to air pollution in large cities requires taking into account air quality at the level of the individual, as well as individual spatial dynamics (mobility and realization of daily activities) that shape each person's risk of exposure. These requirements call for the development of interdisciplinary tools combining the representation of urban space, traffic simulation, emission calculation, advanced air quality models, and the consideration of behavioral and socio-economic dimensions in the modeling process.

We present here a socially and spatially differentiated modeling study of the factors and behaviors that build the exposure of individuals to air pollution in the Greater Paris. This study was carried out using an integrated urban modeling platform including the OLYMPUS emissions model and the CHIMERE chemistry-transport model. The OLYMPUS tool, developed at LISA, is an innovative emission model based on the activity of individuals, making it possible to simulate the socio-differentiated mobility of individuals, for the construction of a pollutant emission inventory adapted to a given urban area. The use of CHIMERE then makes it possible to cross air quality, individuals and mobility, and address the issue of individual exposure to air pollution in a dynamic and integrated manner.

We simulated the year 2009 in the Greater Paris region, and calculated the exposure of individuals taking into account their mobility and their social characteristics, activity and place of residence. Our results are interpreted with regard to the main scientific and societal questions that arise on the subject: Are some individuals more exposed than others? Are these inequalities in exposure linked to the places where people live? To mobility practices? Can they be dependent on socio-professional categories? Do they affect socially vulnerable populations in the same way?

Beyond access to an assessment of exposure inequalities in the current situation, this work makes it possible to support the reflection on the impact of public action on the reduction of environmental inequalities.

Acknowledgments

This research received funding from the French National Agency for Research (ANR-14-CE22-0013), the French Environment and Energy Management Agency (ADEME) and the Île-de-France region (DIM QI²). It was granted access to the HPC resources of TGCC under the allocation A0090107232 made by GENCI. We acknowledge AIRPARIF for data supply.

Références

Elessa Etuman, A., & Coll, I. (2018). OLYMPUS v1.0: Development of an integrated air pollutant and GHG urban emissions model-methodology and calibration over greater Paris. Geoscientific Model Development, 11(12), 5085–5111. https://doi.org/10.5194/gmd-11-5085-2018

Elessa Etuman A., Coll I., Makni I., Benoussaid T., Addressing the issue of exposure to primary pollution in urban areas: Application to Greater Paris, Atmospheric Environment, Volume 239, 117661, ISSN 1352-2310, https://doi.org/10.1016/j.atmosenv.2020.117661, 2020

How to cite: Benoussaïd, T., Coll, I., Charreire, H., and Elessa Etuman, A.: A socio-spatial analysis of air pollution exposure in the Greater Paris, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15999, https://doi.org/10.5194/egusphere-egu23-15999, 2023.

The urban environment has a large population and therefore a large number of polluting activities. Urban organization and the organization of road traffic are levers for controlling energy consumption, transport demand, air quality and the exposure of city dwellers, but these initiatives have social repercussions that do not impact all citizens in the same way, which may create a form of social and environmental injustice. Modeling of urban development scenarios should make it possible to investigate these questions. The challenges of such scenarios lie in the difficulty of modeling a given situation at the level of individuals by taking into account their socio-demographic characteristics, their places of life and their mobility behaviors.

In 2017, a Low Emission Zone (LEZ) has been implemented in the Paris Metropolis, including the city of Paris and the nearby municipalities. The objective is to reduce pollutant emissions and improve air quality in the territory. Recent studies have assessed the average health impact expected from such measures, but they did not consider socio-environmental outcomes. In our work, we thus decided to build a complete LEZ scenario considering the mobility of individuals differentiated by their geographic, demographic and socio-professional specificites.

The urban modeling platform that we use is centered on the OLYMPUS tool, which makes it possible to design mobility scenarios and the associated pollutant emissions by taking into account the urban form, the transport offer, the constraints of land use planning as well as than individual and socially differentiated mobility. With this tool we built different forms for the implementation of the Greater Paris LEZ, and we used the resulting emissions in the CHIMERE air quality model.

We present here the methodology implemented to transcribe the LEZ scenario in our platform as well as the first results obtained on the socio-differentiation of the LEZ constraints (mobility for individuals) and impacts (exposure on individuals).

Aknowledgments

This work was supported by the EUR-LIVE program at Université Paris Est Créteil (French National Research Agency fundings).

References

Elessa Etuman, A., & Coll, I. (2018). OLYMPUS v1.0: Development of an integrated air pollutant and GHG urban emissions model-methodology and calibration over greater Paris. Geoscientific Model Development, 11(12), 5085–5111. https://doi.org/10.5194/gmd-11-5085-2018

Elessa Etuman A., Coll I., Makni I., Benoussaid T., Addressing the issue of exposure to primary pollution in urban areas: Application to Greater Paris, Atmospheric Environment, Volume 239, 117661, ISSN 1352-2310, https://doi.org/10.1016/j.atmosenv.2020.117661, 2020

Host, Sabine, Cécile Honoré, Fabrice Joly, Adrien Saunal, Alain Le Tertre, et Sylvia Medina. « Implementation of Various Hypothetical Low Emission Zone Scenarios in Greater Paris: Assessment of Fine-Scale Reduction in Exposure and Expected Health Benefits ». Environmental Research 185 (juin 2020): 109405. https://doi.org/10.1016/j.envres.2020.109405.

How to cite: Costes, M., Elessa Etuman, A., Benoussaïd, T., and Coll, I.: Building a LEZ scenario and its social and environmental impacts in the Greater Paris area, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16298, https://doi.org/10.5194/egusphere-egu23-16298, 2023.

Urban water security levels will be threatened during the next few years due to new development pressures combined with the climate emergency and increasing population growth in cities. In the UK, London’s planning authorities have a target of more than half a million households for the next 10 years. This new housing will increase the current impacts on urban consumer demand, flood risk, and river water quality indicators. In our previous work, we developed a new concept for urban Water Neutrality (WN) inside an integrated urban planning sustainability framework called CityPlan to deal with water stress and urban complexity issues. This framework integrates the UK’s planning application process with systemic design solutions and evaluation, all being spatially represented in a GIS platform. With the new digital era, there is a constantly increasing number of spatial datasets that are openly available from different sources, but most of them are disaggregated and difficult to understand by key urban stakeholders such as Local Planning Authorities, housing developers, and water companies. Moreover, there are several Multi-Criteria Decision Support Tools (MCDST) that address water management challenges in the literature; but there is still little evidence of one that evaluates the impacts and opportunities to allocate water neutral urban developments.

In this work, we expand the CityPlan framework and present an innovative fully data-driven approach to test WN indicators at different urban scales. WaNetDST integrates GIS spatial data with a series of rules for development impact and offset opportunity based on the current properties of the urban land. This integration is linked to a new scoring system from expert advice that maps strategic areas for water neutral interventions and links the most impactful zones with others more prone to be intervened. The tool connects different urban scales with a series of case study areas: from city (i.e., London), to borough (i.e., Enfield), and to urban development scale (i.e., Meridian Water Development). In the end, WaNetDST visually compares the need for housing vs. green spaces and the trade-offs between new housing vs. retrofitting existing infrastructure, providing a series of maps that guide the planning decision-making process in an integrated way. The results from CityPlan might potentially change the decision-making process for LPAs and housing developers and open a new dialogue between boroughs inside the same city, providing a novel and automated system for WN trade-offs and linking data-driven design with future planning decisions

How to cite: Puchol-Salort, P., Boskovic, S., Dobson, B., Krivtsov, V., Rico-Carranza, E., van Reeuwijk, M., Whyte, J., and Mijic, A.: Integrated Urban Planning Decision-Making Process Towards Water Neutral Solutions, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16910, https://doi.org/10.5194/egusphere-egu23-16910, 2023.

The potential to deliver better, more efficient and sustainable cities has motivated recent research on Digital Twins (DTs) that seek to support planning decisions by displaying analytical evidence to inform collaborative design actions. Prior research identifies different types of DT tools and gives recommendations for their use, but it has not been grounded in engaged research that co-designs DTs with planning users from the context description and problem formulation stage. We report on a research project to co-create DTs with local government planners to visualize interactions between water resources and housing development. We describe co-creating two different DTs starting from the context and problem for water management and assess the steps against these three categories. While engaging with the field we built on prior studies to identify a set of categories relevant to co-creating and assessing DTs: context, governance, spatial and technical definition.  By recording the steps of the DT design process, and contrasting the results with the theoretical proposals, we develop a three-step framework for the co-creation of DTs. This step-by-step framework, illustrated by examples, provides a contribution to the literature on the co-design of DT, and we conclude by discussing implications for practitioners and areas for further research.  

How to cite: Rico Carranza, E., Mijic, A., and Whyte, J.: Co-Creating Digital Twins for Planning of Water Resources and Housing Development, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17338, https://doi.org/10.5194/egusphere-egu23-17338, 2023.

This study shows the unintended tradeoffs of water conservation strategies in the City of Las Vegas. We used the Weather and Research Forecasting model coupled with a multilayer Urban Canopy Model, and forced with the Local Climate Zones from WUDAPTv2, to carry out cloud-resolving simulations aiming to estimate the impacts of city-wide turf removal. Results show that removing the turf, which removes most of non-functional urban irrigation needs, significantly warms up surface temperature via surface energy rebalancing by reducing latent heat and increasing sensible and ground heat fluxes. A striking result is that the increase in sensible heat also increases boundary layer instability favoring more and longer lasting clouds and invigorating afternoon storms. The enhanced afternoon storms tend to cool the surface temperature, but the turf removal net warming impact remains. We also used the model to show how climate intervention scenarios based on cool roofing and pavement strategies can ameliorate the underlying turf removal consequences.

How to cite: Mejía Valencia, J. F., Henao, J., and Saher, R.: The effect of removal of all non-functional turf in Las Vegas: tradeoffs between water conservation, excessive heat, and storminess, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-17481, https://doi.org/10.5194/egusphere-egu23-17481, 2023.

Precipitation nowcasting (short-term forecasting ahead for 0 to 6 hours) is crucial for decision-making of weather-dependent industries in order to mitigate socio-economic impacts. Accurate and trustworthy precipitation nowcasting can serve as an early warning of massive floods, as well as a guide for water-related risk management. Although precipitation nowcasting is not a novel concept, it is challenging and complicated due to the extreme variability of precipitation. The traditional theory-driven numerical weather prediction (NWP) methods confront numerous obstacles, including an insufficient understanding of physical processes, enormous initial conditions impacts on predictions and requiring substantial computing resources. On the other hand, data-driven deep learning models establish a relationship between input and output data to predict future precipitation without taking into account the underlying physical processes. The framework of universal multifractal (UM) is also presented to describe the variability of precipitation nowcasting and compared to the radar observations. In this study, the convolutional long short-term memory (ConvLSTM) model is used to perform precipitation nowcasting over Metropolitan France. The study employs radar data collected every 5 minutes with a spatial resolution of 1km from Meteo-France. The preliminary results show that the structure of the field is reasonably forecast, as well as the somewhat moderate rain rates, but not the most intense ones. We discuss how to improve the methodology.

How to cite: Zhou, H., Schertzer, D., and Tchiguirinskaia, I.: Deep Learning and Universal Multifractal for Nowcasting Precipitation in Urban Geosciences, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-768, https://doi.org/10.5194/egusphere-egu23-768, 2023.

During the lockdown of the COVID-19 pandemic, most countries imposed mobility restrictions such as physical distancing from others, “self-isolation”, and/or “quarantine”. Although these “isolation” measures had been effective in hindering the spread of the virus effectively, people’s mental health can be strongly affected by these “isolation” measures. Some researchers found that the mental health problem indeed increased during the lockdown. In order to reduce the negative effects of isolation on mental health, some studies suggested that people should be brought closer to nature spaces during the lockdown. In this regard, policymakers are paying more attention to Nature-based Solutions (NBS) (e.g. green roofs, gardens, and urban parks), which can potentially improve people's physical and mental health via interactions between people and nature.

In order to enhance the connectivity of the landscape to improve the ecosystem services and reduce health risks with the help of NBS in a post-COVID world, it is significant to consider the heterogeneity of the spatial distribution of green spaces. A number of studies have found that estimates of green space areas are scale-dependent, it is therefore important to investigate the intrinsic complexity of the heterogeneity of the green spaces across a range of scales. This could be achieved with the help of the universal multifractal (UM) framework (Schertzer and Lovejoy, 1987), a stochastic approach widely used to quantify the variability of geophysical fields across a range of scales. This study aims to improve the landscape connectivity of the green spaces in Paris across scales with the help of the UM cascade model.

To achieve the aim of this study, we first quantified the heterogeneous spatial distributions of green spaces of the selected areas by using the fractal dimension. Then, a distance analysis is performed for non-green spaces to green spaces, and a series of NBS scenarios are created based on integrating potential NBS into the current landscape by using the UM cascade model. Finally, the spatial distributions of the NBS combined with the original green spaces are quantified by the fractal dimension and distance analysis. The results indicate that NBS can effectively improve the connectivity of the landscape and has the potential to reduce the physical and mental risks caused by COVID-19. More specifically, this study proposes a scale-independent approach for enhancing the multiscale connectivity of the NBS network in urban areas and provides quantitative suggestions for cities in a post-COVID world.

How to cite: Qiu, Y., Tchiguirinskaia, I., and Schertzer, D.: Enhancing landscape connectivity by Nature-based Solutions for cities in a post-COVID world: a case study in Paris, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-820, https://doi.org/10.5194/egusphere-egu23-820, 2023.

How to cite: Thomas, A. J., Kurths, J., and Schertzer, D.: Multifractals, Climate Networks and the extreme variability of precipitation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-841, https://doi.org/10.5194/egusphere-egu23-841, 2023.

Urban green space - the presence of vegetation-covered area within cities' boundaries - is an increasingly relevant indicator for evaluating sustainable cities. This is because besides providing a set of local services such as mitigating the urban heat island effect and reducing the impact of extreme precipitation events, urban green space has been widely associated with increasing well-being of urban dwellers. Here we present a global analysis of recent trends in urban green space based on modelling of multi-spectral satellite imagery data to reproduce street-based vegetation presence indicators. We estimate local to continental trends over the 2016-2022 period and estimate global scale (considering a large set of the most populated cities) urban green space change statistical trends. We examine heterogeneities in the direction and magnitude of trends across cities and regions, while also analysing within-city inequalities. Our analysis provides an updated picture of urban green space across world cities and an open-source and open data-driven, spatially cross-validated approach to assess changes in near-real-time.

How to cite: Falchetta, G. and Hammad, A. T.: Tracking global urban green space trends, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1989, https://doi.org/10.5194/egusphere-egu23-1989, 2023.

Dengue fever is a mosquito-borne disease caused by the dengue virus bringing huge health burdens in tropical regions. With global warming, rapid urbanization, and mosquito species introductions, the range of dengue fever is expected to expand to subtropical regions and increase potential health risks for local populations. To reduce dengue fever transmission, relevant risk map is one of the most effective tools for public health management. Though there is abundant literature about mapping the dengue fever risks in endemic regions, few studies in contrast have investigated dengue fever risks for non-endemic regions; hindering the development of preparedness planning.

In this study, the spatial hazard-exposure-vulnerability assessment framework proposed by the Intergovernmental Panel on Climate Change was applied in to detect the dengue fever risk in Hong Kong, which is a typical high-density city located within a subtropical region. Firstly, the spatial distribution of the habitat suitability for Aedes albopictus, a mosquito species common in Hong Kong and proxy for the potential dengue fever hazard, was predicted using MaxEnt models relying on the surveillance data and a list of variables related to urban morphology, landscape, land utilization, and local climate. Secondly, the bivariate local Moran’s I was measured to identify urban areas with both high dengue hazard and high human population exposure. Then, vulnerable groups among the human population were identified from the 2016 Hong Kong census data. Finally, dengue risks were assessed at the community scale by overlapping the results of hazard, exposure, and vulnerability analysis.

In the optimal MaxEnt model predicting the presence possibility of Aedes albopictus, the normalized difference vegetation index, frontal area index, and the aggregation index of public residential land ranked the top three among all predictors, with permutation importance of 31.8%, 22.8%, and 17% respectively. Three components were generated after principal component analysis on the vulnerable groups. Lastly, this approach allowed the identification of 17 high-risk spots within Hong Kong. In addition, the underlying factors behind each hot spot were investigated from the aspects of hazard, exposure, and vulnerability respectively, and specific suggestions for dengue prevention were proposed accordingly.

The findings provide a useful reference for developing local dengue fever risk prevention measures, with the proposed method easily exportable to other high-density cities within subtropical Asia and elsewhere.

This study was funded by the Health and Medical Research Fund of the Food and Health Bureau (No. 20190672).

How to cite: Yin, S., Hua, J., Ren, C., Guénard, B., Wang, R., Weemaels, A. I., Shi, Y., Lee, T.-C., Yuan, H.-Y., Chong, M. K., and Tian, L.: Mapping dengue fever risk for a non-endemic high-density city in subtropical region, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-2994, https://doi.org/10.5194/egusphere-egu23-2994, 2023.

This presentation focuses on the governing equations of incompressible and compressible flow in fractional time and multi-fractional space as developed recently by the authors in DOI: 10.1038/s41598-022-20911-3. Mathematical differentiation has found many applications in real-life problems in the last two decades, before which it was mainly utilized by mathematicians and theoretical physicists. The proposed fractional governing equations for fluid flow may be interpreted as the general forms of the classical Navier–Stokes equations; as they reduce to the classical ones when integer values are replaced with their fractional powers in space and time. Due to their nonlocal structure, proposed governing equations in factional time/space can reflect the initial conditions for long times, and the boundary conditions for long distances. Results of numerical applications are presented for flow due to a wall suddenly set into motion. It is found that the proposed equations have the potential to model both sub-diffusive and super-diffusive flow cases.

How to cite: Ercan, A. and Kavvas, M. L.: Navier–Stokes equations in Fractional Time and Multi‑Fractional Space, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6138, https://doi.org/10.5194/egusphere-egu23-6138, 2023.

Limited knowledge on current and future processes as well as data scarcity pose a major challenge when it comes to the evaluation of adaptation strategies towards flooding. Current simulation approaches often lack the flexibility do deal with the inherit dynamics of future development (land use, urban growth, re-development of slum areas, infrastructure construction, etc.) in coastal cities and the resulting changes in flood hazard, exposure and vulnerability whilst facing a lack of sufficient data. Therefore, we developed a modelling approach which is able to integrate future dynamics in the three risk components, hazard, exposure and vulnerability under the uncertainties arising from lacking data as well as limited knowledge. We used Mumbai, India as a first case study to combine Urban Structure Types with Bayesian Networks (BN) and to assess pluvial flooding. BN structures are defined by process understanding supported by existing models, literature and expert evaluations. The quantification of the BNs is done by using urban structure types as proxies for relevant parameters/nodes where data is not available, like the distribution and capacity drainage infrastructure and its condition or the degree of imperviousness of certain areas. This is justified by the assumption that the appearance and the processes in urban structure types are similar. However, the probabilistic definition of nodes in a BN allows to account for the variability within an urban structure type class. As a first step, the approach was set up for the hazard component of risk. Here first results of the simulation of pluvial flooding are shown and validated against flood hotspots reported by the government of Mumbai. The simulation approach reproduced the flooding hotspots, however it has a great sensitivity towards certain parameters, especially towards the digital elevation model and the condition of the drainage infrastructure. In a next step BNs for multi-hazard evaluation and vulnerability assessment will be developed and linked, i.e. fluvial and coastal flooding as well as social vulnerability. The integration of different risk components and the flexibility of the approach help to assess the effect of individual and combinations of soft and hard adaptation measures on future flood risk.

How to cite: Zwirglmaier, V. and Garschagen, M.: Towards an integrated assessment of future flooding in dynamic and data-scarce urban environments by linking Urban Structure Types with Bayesian Network modelling, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6276, https://doi.org/10.5194/egusphere-egu23-6276, 2023.

A significant number of papers focusing on the relationships between COVID-19 diffusion and geographic factors is available in literature. The same applies to the use of geographic techniques (e.g., spatial tools and mapping) for the study of the pandemic. Although the literature on these topics is already abundant, a detailed and comprehensive review is still lacking.
In this context, the purpose of this paper is to fill the existing gap by presenting a literature review of geographical studies dealing with the COVID-19 pandemic. The review is aimed at: i) understanding the role of geographic/territorial determinants (e.g., geographic location of confirmed cases, climatic and environmental characteristics, urbanization) in the spread of COVID-19; ii) identifying common approaches, materials, and methods used in the study of the COVID-19 outbreak from a geographical perspective; iii) recognising possible research gaps to address future in-depth analyses.
To achieve these goals a literature review was made concerning the application of geographical approaches for the study of one or more geographical factors/variables, as well as socioeconomic factors in relation to the outbreak and diffusion of the COVID-19 pandemic. The main academic literature databases were inquired. More than 80 papers were reviewed and categorized according to different criteria, e.g., considered variables, investigated period, spatial and temporal resolution and applied methodologies.
This research is part of an interdisciplinary project (“DISCOV19”) funded by the University of Modena and Reggio Emilia and aiming at identifying the main vulnerability and risk factors related to COVID-19 outbreak and at formulating prevention and management schemes with a focus on the Province of Modena (Northern Italy). The investigation crosses different disciplines: i) public health epidemiology, investigating the contagion modalities and health and socio-demographic predisposing factors; ii) economic-statistical methodology, pointing out the structural characteristics of the networks that convey the contagion and the main social, technological and management vulnerabilities with respect to COVID-19 spread; iii) geography and geomorphology, for thematic mapping and spatial analysis of COVID-19 outbreak and understanding the role of environmental and physical-geographical factors on COVID-19 incidence. The review here presented fits into this context being one of the first outputs of the project implementation.

How to cite: Coratza, P., Ghinoi, A., Palandri, L., Righi, E., Rizzi, C., Soldati, M., and Vandelli, V.: The influence of geographical factors on COVID-19 outbreak: A literature review, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-8235, https://doi.org/10.5194/egusphere-egu23-8235, 2023.

The improved ability of imaging sensors to capture very high resolution (VHR) remote sensing images has been boosted by recent enhancement in the data processing algorithms. This improvement raises the potential of providing precise scene understanding for many applications. For instance, the task of semantic segmentation is an important one in the context of 3D building modelling. In this work, the main objective is to show the potential of using images of a stereo pair as inputs to a neural network trained for semantic segmentation into different urban classes. Actually, instead of increasing artificially the amount of training data, the use of stereo pair images can be seen as a realistic data augmentation, where the model will be trained to see the same object from different acquisition angles. From an experimental point of view, the results show that the method achieves better performances and gives a greater ability to generalize compared to the use of a single view. 

The segmentation process is performed using an encoder-decoder network architecture, namely the U-net network which includes an EfficientNet for the encoder part and a RefineNet for the decoder stage. The model is trained on Pleiades images involving different sources of ground truth (OpenStreetMap, IGN databases and in-house LCLU AI4GEO hierarchical labelled data). Additionally to the spectral information, height information is also considered to enhance the segmentation accuracy. This latter information is obtained using digital surface model (DSM). Indeed, classes identifying urban areas (building class for example) can be more easily discerned according to their height information. 

Furthermore, since Pleiades images are used as inputs of the proposed model, some geometrical issues need to be handled. To remove this complexity, a simulated imaging geometry of a perfect instrument is designed with no optical distortion and no high attitude perturbations. Resulting geometry is commonly called perfect sensor geometry. Since then, to avoid problems of geometric offsets between different data sources (satellite images in perfect geometry), terrain geometry of DSM/DTM and various ground truth databases, several tools have been developed to allow conversion between different geometries. The ortho-rectification is a commonly used geometrical correction that aims at presenting images as if they had been captured from the vertical. Therefore, this correction requires the availability of a HR digital terrain model (DTM) and may result on some distortions. In particular, some area may be occluded and others may arise a spreading effect of buildings.

 To address this issue, and preserve the native image information of the perfect sensor geometry, one key contribution of this work is to map the DSM data and ground truth image into the perfect sensor geometry. By doing this geometric processing and object positioning during the training process, better overlays between different data sources (stereo pair images, DSM model and ground truth data) are ensured and geometrical distortions and offsets can be avoided. In addition, the inferences can be done directly on the perfect sensor geometry without having to go through terrain geometries which requires high resolved DSM/DTM models.

How to cite: Akodad, S. and Lassalle, P.: Automatic Land Cover Segmentation from Perfect Sensor Stereo Images with Height Information, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9983, https://doi.org/10.5194/egusphere-egu23-9983, 2023.

Malaria is a widespread, mosquito-borne, potentially lethal infectious disease that affects humans and other animals. Its prevention and treatment have been targeted in science and medicine for hundreds of years. During the 20th century it was widespread in the Middle East, including Cyprus, and was one of the most important health hazards worldwide. In 1967, the World Health Organisation (WHO) declared Cyprus as malaria-free, an impressive feat considering malaria had plagued the island since the Roman period (Demetrios, 2009). We conducted a critical analysis of anti-malarial campaigns on urban and rural districts in Cyprus, under British colonial rule (1878 – 1960), in the context of malarial disease knowledge in health surveillance and care policy. Under the HIGH-PASM (High-resolution palaeoclimate records and social vulnerability for the last Millennium) project, we present a methodology for constructing database tools relative to heterogeneously distributed historical sources. The aim of this research is to study the impact of the anti-malarial works on the Cypriot landscape as the social-political situation and the methods implemented did not follow the stringent protocols that exist today. Main issues are the complexity regarding British, Ottoman and French social and political roles, ground truth data extracted from historical sources - that need critical analysis, and the complex phenomena under scope. Primary sources are annual medical reports, written by British medical officers, published from 1913 to 1953. The main focus of these actions linked geography to healthcare issues by eliminating the newly identified malaria-vector by directly influencing the mosquito’s habitat, thus indirectly affecting the Cyprus landscape. The assertion, verification and evaluation of the before mentioned actions requires the medical reports to be contextually placed alongside secondary sources (for example correspondences, journal articles, conference proceedings, etc), which were produced or disseminated during this time period by different actors or groups of actors. We aim to apply methodologies used in digital humanities and conceptual modelling within geosciences to verify and understand spatial-temporal information that may be found within archival references. Raw data are extracted from a small corpus to produce meta-data (data sense ((Hui, 2015)) using existing cultural heritage vocabularies (CIDOC CRM base and extensions) relative to different fields and objects to model spatial-temporal events and align these data with authority databases using W3C (World Wide Web Consortium) semantic web standards (technologies). Given the British colonial role in the governance of the island, there is a lack of empirical evidence on the choices of techniques or actions employed. Thus, this meta-data conceptual modelling and raw data collection, as a data management approach, offers a syntactic, semantic and pragmatic understanding of archival sources. This methodology ultimately aims to study the impact on the landscape of the anti-malarial campaigns by bridging gaps in existing literature by the digitisation of physical reports and digitalisation of a healthcare system from the late 19^th to 20^th century.

How to cite: Danielkutty, L. E., Krummeich, R., Couillet, A., Charalambidou, I., and Eliot, E.: Constructing an event centred modelling process to produce spatial and temporal data for the critical study of the impact of British colonial rule (1878 – 1960) on the Cypriot landscape through their anti-malarial campaigns, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11346, https://doi.org/10.5194/egusphere-egu23-11346, 2023.

The remediation of brownfield is vital to sustainable place-making and levelling up across the country. It provides an improved local environment that can unlock regeneration and the social, economic and ecological revitalisation of communities. However the total benefits of remediation are not fully understood or utilised in decision making. As a result, sites can remain derelict for years and opportunities to optimise value from public and private investment are missed.

Jacobs and BGS undertook research for the Environment Agency in England to evaluate the feasibility of developing a tool, which included:

• A virtual workshop using MURAL to enable digital interaction and collaboration to refine scope, define data requirements and map project stakeholders;
• Primary benefit and user requirements research, including looking at the potential impact of a tool through the development of a Theory of Change model and focussed interviews with key stakeholders to understand user requirements.
• Review of academic and grey literature;
• Accelerated design sprint to frame the problem/opportunity, explore technology agonistic solutions for the tool and develop into a storyboard.
• Develop a low fidelity prototype as a blueprint of how a tool might look.

The outcome of the work indicated there is both a need and demand for such a tool. It was also demonstrated to be technically feasible through the literature review and design sprint. Such a tool would have an extremely positive impact on the perceptions of brownfield, shifting it from a constraint to an opportunity. The presentation will provide a summary of the methods, an overview of the results and a demonstration of a prototype digital tool. Our disucssion will focus on the opportunities presented by using systems thinking combined with design thinking to influence the approach taken to planning and redeveloping brownfield sites.  

How to cite: Beriro, D., Macklin, Y., Thrasher, J., Griggs, D., and Haslam, A.: A feasibility study for a novel remediation and sustainable growth digital tool for the Environment Agency, England, UK, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11851, https://doi.org/10.5194/egusphere-egu23-11851, 2023.

Understanding how the exposure and vulnerability to floods and other climate hazard varies between different groups of urban residents is an urgent prerequisite for guiding urban climate change adaptation policy and action. To be most effective, adaptation measures need to be designed specifically in relation the exposure and vulnerability profiles of different groups. Index approaches have since long been used to cluster households according to different levels of exposure and vulnerability. However, two main gaps remain in current research: First, while indices are typically based on either survey or statistical data, approaches transcending both levels of resolution through proxy variables are rare. Second, profiling is typically not linked to spatial categories such as urban morphology types.

The approach presented here contributes to bridging both gaps. We use original household survey data from Ho Chi Minh City to generate an exposure and vulnerability index for the city. We then test the validity of that index, which is based on detailed data, in comparison to an index which builds on rougher statistical data. In a third step, we test how well vulnerability and exposure profiles from the index map against urban morphology types which can be used in risk modeling in order to analyse in how far valid exposure and vulnerability profiles can be linked to such morphology types. Our results show an existing yet limited link between exposure and vulnerability profiles on the one side and urban morphology types on the other. The results are essential for advancing urban risk modeling in an integrated manner and rolling such modeling out to larger spatial areas.

How to cite: Tu, J., Reimuth, A., Katzschner, A., Yang, L. E., and Garschagen, M.: Profiling households through a combined vulnerability and flood exposure index in Ho Chi Minh City, Vietnam, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13059, https://doi.org/10.5194/egusphere-egu23-13059, 2023.

Keywords: Urban sustainability, Earth observation, 3D change detection, Deep Learning, Dataset

Nowadays, remote sensing products can provide useful and consistent information about urban areas and their morphological structures with different spatial and temporal resolutions, making it possible to perform long term spatiotemporal analyses of the historic development of the cities and in this way to monitor the evolution of their urbanization patterns, a goal strictly related to the United Nations (UN) Sustainable Development Goals (SDGs) concerning the sustainability of the cities (SDG 11 - Sustainable Cities and Communities).

In this context, Change Detection (CD) algorithms estimate the changes occurred at ground level and are employed in a wide range of applications, including the identification of urban changes. Most of the recently developed CD methodologies rely on deep learning architectures. Nevertheless, the CD algorithms currently available are mainly focused on generating two-dimensional (2D) change maps, where the planimetric extent of the areas affected by changes is identified without providing any information on the corresponding elevation (3D) variations. These algorithms can thus only identify planimetric changes such as appearing/disappearing buildings/trees, shrinking/expanding structures and are not able to satisfy the requirements of applications which need to detect and, most of all, to quantify the elevation variations occurred in the area of interest (AOI), such as the estimation of volumetric changes in urban areas.

It is therefore essential to develop CD algorithms capable of automatically generating an elevation (3D) CD map (a map containing the quantitative changes in elevation for the AOI) together with a standard 2D CD map, from the smallest possible amount of information. In this contribution, we will present the MultiTask Bitemporal Images Transformer (MTBIT) [1], a recently developed network, belonging to the family of vision Transformers and based on a semantic tokenizer, explicitly designed to solve the 2D and 3D CD tasks simultaneously from bitemporal optical images, and thus without the need to rely directly on elevation data during the inference phase. 

The MTBIT performances were evaluated in the urban area of Valladolid on the modified version of the 3DCD dataset [2], comparing this architecture with other networks designed to solve the 2D CD task. In particular, MTBIT reaches a metric accuracy equal to 6.46 m – the best performance among the compared architectures – with a limited number of parameters (13,1 M) [1].

The code and the 3DCD dataset are available at https://sites.google.com/uniroma1.it/3dchangedetection/home-page.

References

[1] Marsocci, V., Coletta, V., Ravanelli, R., Scardapane, S., and Crespi, M.: Inferring 3D change detection from bitemporal optical images. ISPRS Journal of Photogrammetry and Remote Sensing. 2023 - in press.

[2] Coletta, V., Marsocci, V., and Ravanelli, R.: 3DCD: A NEW DATASET FOR 2D AND 3D CHANGE DETECTION USING DEEP LEARNING TECHNIQUES, The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, XLIII-B3-2022, 1349–1354, 2022.

How to cite: Marsocci, V., Coletta, V., Ravanelli, R., Scardapane, S., and Crespi, M.: New trends in urban change detection: detecting 3D changes from bitemporal optical images, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13357, https://doi.org/10.5194/egusphere-egu23-13357, 2023.

COVID-19 severely affected Italy from the beginning of the pandemic. The number of cases can be analysed within statistical methods, to understand its spread over time, in conjunction with factors as meteorological and environmental conditions, socio-economic conditions and urban settings [1]. The role of spatial aggregation of data is seldom investigated in detail, as the information available to the public is often limited to administrative boundaries.

We investigated the number of infections stratified by spatial location and time, analyzing each of the 107 provinces of Italy during the two infection waves in 2020. Infections were greater in urban areas such as Rome, Milan and Naples. We further investigated the role of urban areas by considering specific spatial aggregations that explicitly included indicators of human presence [2].

We used the hhh4 endemic-epidemic model to study the spatial-temporal pattern of COVID-19 [3]. The model includes three components, representing autoregressive effects (transmission of disease within a single province), neighborhood effects (transmission between provinces) and endemic effects (sporadic events by unobserved sources of infection).  Covariates included daily temperature, humidity, employment rate and number of high-care hospital beds for each province. In addition, we considered specific covariates to account for urban indicators: population, population density, the proportion of urban area, average area of cities, and number of cities. Covariates were considered on both the autoregressive and neighbourhood components to determine the effect of transmission within and between provinces. To simulate the spread between provinces on the neighbourhood component, we considered the spatial adjacency between provinces, and considered decreasing importance with increasing distance.

Outputs from the model included the risk ratios (RRs) of the covariates, with resulting RR of 0.89 on the autoregressive component and RR of 0.83 on the neighbourhood component. An existing study found that higher temperatures were related to a decline in daily confirmed COVID-19 case counts with a corresponding RR of 0.80 [4].

We specifically looked at covariates related to urban settings, as an existing study showed positive correlation between population density and COVID-19 transmission rate [5]. Our results showed a RR of 1.23 (autoregressive component) and RR of 1.48 (neighbourhood component), suggesting that larger population density leads to more infections, and that movement of people across provinces could lead to a higher risk of COVID-19 cases.  Province area, average city area and number of cities were not statistically significant.

Eventually, we explicitly considered the role of urban settings by aggregating spatial-temporal data within individual urban areas [2], instead of administrative boundaries. As COVID-19 data itself were available at the province level, we distributed them to urban areas proportionally to the area occupied within each province; other data was actually aggregated within urban polygons. We argue that study of the spatial-temporal transmission of infection using urban areas may provide reliable results and help selecting characteristics in urban settings that may favour or prevent the spread of diseases.

[1] M. Agnoletti et al. DOI: 10.1016/j.landurbplan.2020

[2] M. Alvioli. DOI: 10.1016/j.landurbplan.2020.103906

[3] S. Meyer et al. DOI: 10.1214/14-AOAS743

[4] J. Liu et al. DOI: 10.1016/j.scitotenv.2020.138513

[5] K.T.L. Sy et al. DOI: 10.1371/journal.pone.024927

How to cite: Alvioli, M., Fowler, D., and Lim, S.: Spatial-temporal modeling of COVID-19 areas in Italy and the role of urban settings, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13960, https://doi.org/10.5194/egusphere-egu23-13960, 2023.

Expansion of built-up areas has consumed large areas of natural ecological patches in many cities around the world, affecting environmental and living quality of urban residents.  Managing urban landscape and urban trees has gained a special attention in Vietnam in recent years. The green space development plan for Hanoi to 2030 and a vision to 2050 has targeted to reach 62% of green spaces. However, there is lack of detailed green space development plan at the district and commune/ward levels. This study aims to assess urban green space area and quality in Hanoi at the commune/ward levels using remote sensing and population data. The study uses combined remote sensing data from Google Earth, Sentinel-2, and Normalized Difference Vegetation Index (NDVI) to analyze urban green quality and space for Hanoi. The urban green space will be combined with population data at commune/ward level to estimate urban tree cover per person. The research results can contribute to improve the credibility and scientifically of green space construction so that urban planning can adapt and serve the city and its residents and achieve green development.

How to cite: Reimuth, A., Nong, D. H., and Ngo, S. T.: Assessing urban green space area and quality using remote sensing and population data: A case study of Hanoi urban districts, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14511, https://doi.org/10.5194/egusphere-egu23-14511, 2023.

The cooling impact of green roofs is highlighted in the context of urbanisation and urban heat island (UHI) effect. And it is usually described and quantified by evapotranspiration (ET) processes. Understanding ET process is the key to optimize cooling effect. ET estimation can be achieved either directly (weighing lysimeters) or indirectly (e.g., Penman-Monteith equation). Micro-meteorological approaches have been developed in recent years. Among which scintillometer can evaluate ET by its measurement parameter C_n² (which corresponds to the fluctuations of air refractive index n ) in combination with surface energy balance (SEB) and Monin-Obukhov similarity theory (MOST) . Hence, C_n² improvement in Cn2 data would result in better ET estimation. But it is often overlooked and very little research has focused on it. In this project, the research area lies on the top of the Carnot and Bienvenüe buildings in Ecole des Ponts Paristech. Covering an area of ​​1 ha, it is a wavy and vegetated large green roof, known as the Blue Green Wave (BGW). Data from a large aperture scintillometer (LAS) with 10-minute time step during December 2019 and January 2020 on BGW is used in this study. Three estimates of C_n²(Cn2_UCn2, Cn2_PUCn2 and Cn2_Var) were analysed with structure function and universal multifractal model (UM). Such framework has been widely use to characterize geophysical fields extremely variable across wide range of space-time scales. There are two relevant parameters in an UM model, the mean codimension of intermittency C₁≥0and multifractality index 0≤α≤2. α=0, indicates monofractal; α=2, indicates log-normal model. Data in UM framework is analysed by Trace Moment (TM) method and Double Trace Moment (DTM) method. All of estimates demonstrated scale invariance, which could be used for upscaling and downscaling. Cn2_Var performed well even during measurement malfunction, but UM analysis showed it was contradictory to the hypothesis of lognormality. It implies the way it calculates Cn2_Var need some revisions and an assessment of the scintillometer could be achieved by analysing C_n². This research provides a complete grasp of the properties of C_n² and sets the stage for its future application in precise ET estimates.

How to cite: Zhu, S., Gires, A., Maksimovic, C., Tchiguirinskaia, I., and Schertzer, D.: Multifractal analysis of Cn2 scitillometer data and consequences for evapotranspiration estimates in urban areas, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14618, https://doi.org/10.5194/egusphere-egu23-14618, 2023.

The Einstein Telescope (ET) is a proposed underground infrastructure to host a third-generation, gravitational-wave observatory. There are currently two candidate sites to host it: one of this is located in Sardinia region (Italy), in a favourable geological context, the other one in the Meuse-Rhine Euregion. Site-characterization studies are under way towards the site selection, which is expected for 2024. The scope work of this research is to evaluate the surface deformation of this site by integration of remote sensing techniques with geological and geophysical data. In this framework the PSI (Persistent Scattered Interferometry) technique with SAR data is the proposed approach for the analysis of a long time-series imagery. Although recent crustal movements in the study area are supposed to be very small (≃ - 0.5 mm/years from 2014 as measured by EUREF Permanent Network https://epnd.sgo-penc.hu), ESA Sentinel-1 data from Copernicus program, represents an effective tool to update this knowledge and monitor the phenomenon. A first analysis has been performed in the study area using the Snap2Stamps methodology. During the first assessment of this research, this methodology has been tested to a dataset of 94 images from Sentinel-1. The radar data (SLC, Single Looking complex) acquired from January 2021 to July 2022 for both descending and ascending orbits on an area of 250 sq km has been managed. The applied methodology requires a long-time for the processing in order to derive vertical velocities, and we considered the opportunity to use a cloud service. So, we exploited the possibility offered by SNAPPING service provided by Terradue (https://www.terradue.com/portal/), a cloud on-demand computing service for Sentinel-1 Multi-Temporal DInSAR processing, based on integrated SNAP and StaMPS chain. In this service, the dataset can be improved, considering a longer time of acquisition, exploiting the complete Sentinel revisiting time, starting from 2014.

The first results of this analysis have been calibrated with the existing GNSS measures provided by EUREF using the data of the Nuoro station. The ground vertical displacement calculations, composing data from both acquisition orbits, confirm the existing evaluations and extend the current information to the whole study area. Moreover, it will be possible to consider also future acquisition with a continuous monitoring process.

These results can be considered an important value for the proposed Italian site and the ET infrastructure realization.

How to cite: Dessì, F., Melis, M. T., Da Pelo, S., and Funedda, A.: Monitoring land deformation through PSI technique for Einstein Telescope site characterization of Sos Enattos (Sardinia, Italy), EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-14807, https://doi.org/10.5194/egusphere-egu23-14807, 2023.

Urbanization induced changes have attracted widespread attention. Key challenges arise from the inherent uncertainties in attribution models diagnosing the driving mechanisms and the interrelationships of the attributes given by the complexity of interactions within a city. Here, we investigate urbanization dynamics from nighttime light signals before analyzing their driving mechanisms from 2014 to 2020 on both provincial and regional scale and a flat versus mountainous urbanization comparison. Model uncertainties are discussed comparing the contribution results from Geodetector and the Gini importance from Random Forest analyses. The method is applied to Shaanxi Province, where flat urban land is located mainly in its center and mountainous urban land is situated in the North and South. The following results are noted: i) Employing the Geodetector based maximum contribution method for urban region extraction of night time light reveals a notable accuracy improvement in flat urban land compared with the closest area method. ii) Geographical factors attain high contribution for mountainous urban land of Shannan, while for flat urbanization land dynamics, economic factors and community factors prevail. iii) The most obvious driving mechanisms are economic factors which, associated with local urban development strategies, show highest contribution values in 2014 (2018) over the flat (mountainous) urban land of Guanzhong Plain (Northern Shaanxi Plateau or Shanbei region) linked with an early (late) development. iv) Population factors achieve high contribution values in the initially low populated urban land of the northern mountainous land initiating huge migration. v) The contributions resulting from Geodetector are in agreement with the Gini importance from Random Forest in agriculture, geographical and population factors (R² > 0.5) but not in economy, community and climatic factors (R² < 0.5). The dynamics of driving mechanisms for urbanization provides insights in connecting urban geographical expansion with multi-factors and thus to assist municipal governments and city stakeholders to design a city with geographical, climatic and social-economic changes and interactions in mind.

How to cite: Huang, S., Yu, L., Cai, D., Zhu, J., Liu, Z., Zhang, Z., Nie, Y., and Fraedrich, K.: Driving mechanisms of urbanization: Evidence from Geographical, Climatic, Social-economic and Nighttime Light data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16718, https://doi.org/10.5194/egusphere-egu23-16718, 2023.

To characterize a community-scale urban functional area using geo-tagged data and available land-use information, several supervised and semi-supervised classification models are presented and evaluated in Hong Kong for comparing their uncertainty, robustness and sensitivity. The following results are noted: (i) As the training set size grows, models’ accuracies are improved, particularly for multi-layer perceptron (MLP) or random forest (RF). The graph convolutional network (GCN) (MLP or RF) model reveals top accuracy when the proportion of training samples is less (greater) than 10% of the total number of functional areas; (ii) With a large amount of training samples, MLP shows the highest prediction accuracy and good performances in cross-validation, but less stability on same training sets; (iii) With a small amount of training samples, GCN provides viable results, by incorporating the auxiliary information provided by the proposed semantic linkages, which is meaningful in real-world predictions; (iv) When the training samples are less than 10%, one should be cautious using MLP to test the optimal epoch for obtaining the best accuracy, due to its model overfitting problem. The above insights could support efficient and scalable urban functional area mapping, even with insufficient land-use information (e.g., covering only ~20% of Beijing in the case study).

How to cite: Deng, R., Guan, Y., Cai, D., Yang, T., Fraedrich, K., Zhang, C., Tang, J., Liao, Z., Wei, Z., and Guo, S.: Supervised versus semi-supervised urban functional area prediction: uncertainty, robustness and sensitivity, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16766, https://doi.org/10.5194/egusphere-egu23-16766, 2023.

Most science and engineering problems are modeled by time-dependent and parametrized nonlinear partial differential equations. Their resolution with traditional computational methods may be too expensive, especially in the context of predictions with uncertainty quantification or optimization, to allow for rapid predictions.  In this talk, we will overview data-driven methods aimed at representing high-fidelity computational models by means of reduced-dimension surrogate ones.  Different approaches will be presented for the uncertainty quantification for reliable predictions and forecasts in inundation problems.

Particularly, a non-intrusive reduced-order model based on convolutional autoencoders is proposed as a data-driven tool to build an efficient nonlinear reduced-order model for stochastic spatiotemporal large-scale physical problems. The method uses two-level autoencoders to reduce the spatial and temporal dimensions from a set of high-fidelity snapshots collected from an in-house high-fidelity numerical solver of the shallow-water equations. The encoded latent vectors, generated from two compression levels, are then mapped to the input parameters using a regression-based multilayer perceptron. The accuracy of the proposed approach is compared to the linear reduced-order technique-based artificial neural network (POD-ANN) on benchmark tests (the Burgers and Stoker's solutions) and a hypothetical dam-break flow problem over a complex bathymetry river. The numerical results show that the proposed nonlinear framework presents strong predictive abilities to accurately approximate the statistical moments of the outputs for complex stochastic large-scale and time-dependent problems, with low computational cost during the predictive online stage.

The caveat that remains is the long-term temporal extrapolation for problems marked by sharp gradients and discontinuities. Our study explores forecasting convolutional architectures (LSTM, TCN, and CNN) to obtain accurate solutions for time-steps distant from the training domain, on advection-dominated test cases. A simple convolutional architecture is then proposed and shown to provide accurate results for the forecasts. To evaluate the epistemic uncertainties in the solutions, the methodology of deep ensembles is adopted.

REFERENCES

• Bhatt, Y. Kumar and A. Soulaïmani. Deep Convolutional Architectures for Extrapolative Forecast in Time-dependent Flow Problems, DOI: 10.48550/arXiv.2209.09651.
• Abdedou and A. Soulaïmani. Reduced-order modeling for stochastic large-scale and time-dependent problems using deep spatial and temporal convolutional autoencoders.
  arXiv:2208.03190[physics.flu-dyn].
• Jacquier, A. Abdedou, V. Delmas and A. Soulaimani. Non-intrusive reduced-order modeling using uncertainty-aware Deep Neural Networks and Proper Orthogonal Decomposition: Application to flood modeling. Journal of Computational Physics. Volume 424, 1 January 2021, 109854.
• Abdedou and A. Soulaïmani. A non-intrusive reduced-order modeling for uncertainty propagation of time-dependent problems using a B-splines Bézier elements-based method and proper orthogonal decomposition: Application to dam-break flows. Computers & Mathematics with Applications. Volume 102, 15 November 2021, Pages 187-205.
• Chaudhry and A. Soulaimani. A Comparative Study of Machine Learning Methods for Computational Modeling of the Selective Laser Melting Additive Manufacturing Process. Appl. Sci. 2022, 12(5), 2324; https://doi.org/10.3390/app12052324.
• Delmas and A. Soulaimani. Parallel high-order resolution of the Shallow-water equations on real large-scale meshes with complex bathymetries. Journal of Computational Physics. Volume 471, 15 December 2022, 111629

How to cite: Soulaïmani, A., Abdedou, A., Kumar, Y., and Bhatt, P.: Deep Convolutional Architectures for Uncertainty Quantification and Forecast in Inundation Problems, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-39, https://doi.org/10.5194/egusphere-egu23-39, 2023.

Floods are one of the most devastating natural disasters in the world causing loss of human lives and property across the world. These losses can be minimized by accurate prediction of floods well in advance. However, 2D hydrodynamic models which are used for flood inundation modelling require high computational time and hence are unsuitable for development of real-time flood monitoring system in most cases. Therefore, a surrogate machine learning model named XGBoost Regressor (XBGR) is developed for flood inundation modelling. The developed model overcomes the constraint of high computational time required by 2D hydrodynamic models. The XGBR is developed to predict maximum flood depth map and is evaluated with the LISFLOOD-FP hydrodynamic model. The training data for the XGBR model is generated using the LISFLOOD-FP model. The surrogate model is trained on 21 flood events, tested on 4 and validated for 1 flood event. For better development of the surrogate model, physical characteristics of the study area are considered in the form of nine indices referred here as topographic variables along with the flood characteristic variables. However, to refrain the XGBR model from overfitting and decrease the training time, a feed forward feature selection method is used to select the best predictive topographic variables. Four topographic variables are selected after which there is no significant improvement in the model was found. Number of trees and learning rate parameters of XGBR model are parameterized which are having highest impact on the model performance. Mean absolute error (MAE) and root mean square error (RMSE) are used for evaluating model accuracy. For testing period, the average MAE and RMSE are 0.433 m and 0.780 m, respectively and for the validation event MAE and RMSE are 0.595 m and 0.960 m respectively. For evaluating the accuracy of the surrogate model on flood inundation extent, F1 score is used which is the harmonic mean of precision and recall. The F1 score is 0.908 for the testing events and is 0.931 for validation events. The higher value of F1 score (>0.9) indicates good accuracy of the XGBR model when validated using the hydrodynamic model.

How to cite: Sasanapuri, S. K., Chadrika Thulaseedharan, D., and Ashwini Kumar, G.: Using machine learning to emulate the hydrodynamic model for flood inundation modelling, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-554, https://doi.org/10.5194/egusphere-egu23-554, 2023.

Physics-based deep learning experienced a major breakthrough a few years ago with the advent of neural operators. Beyond the traditional use of deep neural networks to predict the solution to a fixed Partial Differential Equation (PDE), these novel methods are able to learn the operator solution to a class of PDEs.

Comparisons and analyses of popular neural operators such as Fourier Neural Operator and DeepONet have been conducted for numerical case studies. However, they are still lacking for more realistic problems in complex settings.

In this study, we compare several neural operators to predict the propagation of seismic waves in heterogeneous media. Our database is composed of more than 12 million ground motion timeseries generated from 50,000 media. We quantify the accuracy of the neural operators, their memory requirements, and their dependence towards both the initial condition and the PDE parameters. We also propose insights on their possible extension to 3 dimensions.

How to cite: Lehmann, F., Gatti, F., Bertin, M., and Clouteau, D.: Advantages and promises of deep neural operators for the prediction of wave propagation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5068, https://doi.org/10.5194/egusphere-egu23-5068, 2023.

In the last decades, geophysicists have developed numerical simulators to predict earthquakes and other natural catastrophes. However, the more precise the model is, the higher the computational burden and the time to results. In addition, even if we could reproduce the phenomenon with more complex and more representative models, the underlying uncertainty would remain significantly high, affecting the reliability of the final prediction. In response to this challenge, we adopted a hybrid strategy, consisting into mixing physics-based numerical simulations and machine-learning. The goal is to transform synthetic earthquake ground motion, obtained via physics-based simulation, accurate up to a frequency of 5 Hz, into a broader-band prediction that mimics the recorded seismographs. In doing so, we factorize the latent representation of the seismic signal, by forcing an encoding that splits features into two parts: a low frequency one (0-1 Hz) and a high frequency one (1-20 Hz). In the following, we train a convolutional U-Net neural network and apply two different signal-to-signal translation techniques: pix2pix and BiCycleGAN. The latter strategies are compared with the prior work of Gatti et al., 2020, on the Stanford Earthquake Dataset (STEAD) showing their capability of mimicking recorded seismographs. We finally tested the two strategies on the synthetic time-histories obtained for the 2019 Le Teil earthquake (France).

How to cite: Jacquet, G., Clouteau, D., and Gatti, F.: Hybrid generation based on machine learning to enhance numerical simulation for earthquake, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-5252, https://doi.org/10.5194/egusphere-egu23-5252, 2023.

The fragile geomorphological context of Italy sets a variety of natural challenges, ranging from seismic to hydrogeological risk. In such a complex territory, documenting the conditions of infrastructures is crucial for planning adequate strategies of maintenance through 3D modelling for structural analysis and digital twins’ implementation of structures like dams (Pagliari et al., 2016) or bridges (Gaspari et al., 2022). Geomatics, through periodical surveys using state-of-the-art technologies, reconstruct accurate 3D models of structures that results in the generation of dense pointclouds from which polygon meshes can be derived as well as in the model integration in Building Information Modeling (BIM) or Finite Element Method (FEM) environments for the computation of simulations and deformation monitoring or structural health assessment analysis in support of decision making.

Such data are generated through different approaches. A traditional methodology first implies the materialization and measurement of a topographic network in a local system with a total station and its subsequent georeferencing in a global coordinate reference system through a roto-translation based on Global Navigation Satellite System observations of ground control points. In the same framework, scans for the acquisition of dense pointclouds are defined through the adoption of a terrestrial laser scanner (TLS). Hence, the execution of planned drone flights, with nadiral and side view of the structure and its surrounding environment, serving as input for the generation of photogrammetric cloud through a robust Structure from Motion data processing.

Implementing open-source WebGL solutions like Potree supports the digital twin and data sharing with audiences of different technical backgrounds, committers concerned with the adoption of a monitoring platform for integrating products in different format as well as experts with non-geomatics expertise interested in further analysis of collected data through computer vision and deep learning approches that enrich the existent documentation. With a user-friendly interactive web platforms users are able to access the 3D model, make measurements and execute simple processing operation like cross-sections and clipping (e.g. https://labmgf.dica.polimi.it/piacenzacs/lugagnano/).

Since 2019, the dams of the Sila mountains in the Calabria region represented the case study for testing the described integrated approach. The present work concerns the integration of data from different sensors (TLS for indoor and outdoor environment, photogrammetric images and lidar from drone) for the generation of the digital twin of the arcuate-plan gravity dam of Trepidò. The dam digital twin of the dam and adjacencies consists of a pointcloud of 2594370 points, with adaptive density and average accuracy of 1-2 cm for the structure and 10 cm for the downstream vegetated sediment. It can be used to increase knowledge of the structure (built in 1930) and for structural analysis.

Bibliography:

Gaspari, F., Ioli, F., Barbieri, F., Belcore, E., and Pinto, L. (2022): Integration of UAV-LiDAR and UAV-photogrammetry for infrastructure monitoring and bridge assessment, Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XLIII-B2-2022, 995–1002, doi.org/10.5194/isprs-archives-XLIII-B2-2022-995-2022.

Pagliari, D., Rossi, L., Passoni, D., Pinto, L., de Michele, C., and Avanzi, F. (2016). Measuring the volume of flushed sediments in a reservoir using multi-temporal images acquired with UAS, Geomatics, Natural Hazards and Risk, 8(1), 150–166, doi.org/10.1080/19475705.2016.1188423

How to cite: Gaspari, F., Barbieri, F., Ioli, F., Pinto, L., and Valgoi, P.: Integration of 3D surveying approaches for critical infrastructure digital twins in natural hazard-prone scenarios, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-7773, https://doi.org/10.5194/egusphere-egu23-7773, 2023.

Early warning systems protect and support lives, jobs, land and infrastructure. DAKI-FWS, a German national project, aims at developing an early warning system to protect the German society and economy against extreme weather and climate events such as floods, droughts and heatwaves. With a seasonal temporal horizon, DAKI-FWS requires high resolution and bias corrected seasonal forecast of daily minimum and maximum temperatures, daily precipitation and wind speed. To derive such information, we have developed a deep neural network (DNN) approach to downscale and bias correct coarse resolution seasonal forecast ensembles on a 1 degree grid to a 1 arc minute grid.

The proposed DNN approach is here analyzed and compared with other machine learning approaches. Results show that such a deep learning technique can generate realistic, temporally consistent, and high-resolution climate information. The statistical and physical properties of the generated ensembles are analyzed using spatial correlation, cross validation and SVD. The DNN predicts extreme values that are very close to the observed values while preserving the physical relationships in the system as well as the trends in the variables.

How to cite: Heidari, F., Lin, Q., Espitia Sarmiento, E. F., Toreti, A., and Xoplaki, E.: Towards the development of an AI-based early warning system: a deep learning approach to bias correct and downscale seasonal climate forecasts, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-9555, https://doi.org/10.5194/egusphere-egu23-9555, 2023.

Usually, the earthquake catalog for a given region represents a collection of all detected and localized earthquakes and, thus, contains not only the main shocks, but also fore- and aftershocks. In order to perform an independent seismic event and seismic hazard analysis we require a catalog that, ideally, contains only mainshocks. Thus, the removal of dependent fore- and aftershocks from an earthquake catalogby declustering is a crucial step in seismic hazard analysis. Machine learning methods can potentially offer improvements in speed and accuracy in comparison to classical declustering approaches.

Here, we propose a hybrid approach to identify the temporal clusters of earthquakes from the catalogs of California (USGS) and Japan (ISC). We combine unsupervised 1-D clustering algorithms with seismologically informed methods and machine learning techniques. We use epidemic type aftershock sequence (ETAS) generated catalogs as well as classically declustered catalogs to benchmark the method.

How to cite: Köhler, J., Li, W., Faber, J., Rümpker, G., Stöcker, H., and Srivastava, N.: A hybrid approach for declustering of earthquake catalogs, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11462, https://doi.org/10.5194/egusphere-egu23-11462, 2023.

The present paper focuses on the influence of Rayleigh and Love waves on the seismic structural performance of a simplified nonlinear beam structure representing a bridge column. The impact of surface waves in the structure is quantified directly by a coupled 3D SEM-FEM numerical wave
propagation simulation from the earthquake source to the structure using the Domain Reduction Method.
In the first step, ground motions, including basin-induced surface waves, are generated from a regional model containing the earthquake source and a simplified basin. Surface waves are extracted and characterized with the Normalized Inner Product (NIP) in terms of amplitude and
frequency content from ground motions at different locations inside the basin. In the second step, the seismic wavefield from the SE simulation is imposed in a FE model composed of a nonlinear structure placed over a portion of the basin sediments. The model considers soil-structure
interaction and structural non-linearity through a multifiber beam approach.
By placing the structure in different positions, the extracted surface waves and the structural damage can be linked to a specific location inside the basin. Therefore, the spatial variability of the structural damage and the surface wave characteristics can be quantified. Consequently, this work
evaluates if structural damage can be estimated only from typical ground motion intensity parameters or if other parameters associated with surface wave characteristics are necessary. The results show a correlation between obtained seismic damage with rotational components from
surface waves (torsion for Love waves and rocking for Rayleigh waves).

How to cite: Soto, V. and Lopez-Caballero, F.: Quantification of source- and basin-induced surface waves effects on the seismic performance of nonlinear structures, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11611, https://doi.org/10.5194/egusphere-egu23-11611, 2023.

Solving partial differential equations (PDEs) stably and accurately is essential in simulation analysis of a variety of geophysical phenomena. Designing appropriate discretization schemes for PDEs requires careful and rigorous mathematical treatment and has been a long-term research topic. The computational efficiency is additionally a long-standing challenge when what-if hazard scenario analysis is considered. The data-driven discretization is a hybrid approach to combine machine learning and physics-based simulations, which provides a methodology to derive better discretization schemes from reliable references obtained typically using known stable schemes with higher resolution grids. As the resultant schemes may inherit the physics described by the PDEs, surrogate models employing them are expected to be in good agreement with expensive simulations. It is also argued that the learnt schemes by neural network models can exhibit similar characteristics to known sophisticated algorithms and outperform them in terms of accuracy. However, the method has currently been assessed with only limited examples and the detailed mechanisms of the learnt schemes are not well understood. In this presentation, thorough assessment and investigation of learning discretization schemes are conducted by applying the methodology to several types of differential equations with different learning models for the schemes. Whether the methodology has the potential to derive new schemes is also discussed.

How to cite: Ishikawa, T.: On learning discretization schemes of partial differential equations in geoscience, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-11657, https://doi.org/10.5194/egusphere-egu23-11657, 2023.

Storm surges can have devastating effects on coastal communities. These events, often caused by tropical cyclones, are difficult to simulate due to the challenging nature of process-based modelling and the relative paucity of data covering extreme tropical cyclone conditions. In order to make optimal use of existing physical models, we build an emulator to actively learn the relationship between tropical cyclone characteristics and maximum storm surge height.

We used the ADCIRC physical storm surge model, a reliable but costly tool, to simulate a series of representative tropical cyclones that typically affect the coast near New Orleans. These initial storms were sampled using Latin hypercube design, varying tropical cyclone characteristics such as the landfall speed, central pressure, and others. By running the ADCIRC model for each of these events, we were able to determine the maximum sea surface height caused by each simulated storm. Next, we trained a Gaussian process to fit the maximum sea levels at each point along the coast given the tropical cyclones' characteristics as input. Through active learning, we iteratively selected additional tropical cyclones to further improve the emulator’s accuracy. Finally, we evaluated the model's performance using a held-out test set of idealised tropical cyclones.

Our emulator approach allowed us to efficiently create a high-quality, low-cost statistical model that can potentially be used to predict the probability of future storm surge heights. Additionally, it allowed us to separate uncertainties in the input distribution of tropical cyclone characteristics from uncertainties in the model itself. By better understanding these sources of uncertainties, we can work towards more accurately assessing the potential impacts of future storms on coastal communities.

How to cite: Thomas, S., Jones, D., Mayo, T., and Gopinathan, D.: Tropical cyclone storm surge emulation around New Orleans, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-13337, https://doi.org/10.5194/egusphere-egu23-13337, 2023.

Twitter is increasingly being used as a real-time human-sensor network during natural disasters, detecting, tracking and documenting events. Current wildfire models currently largely omit social media data, representing a shortcoming in current models, as valuable and timely information is transmitted via this channel. By including this data as a real-time data source, we aim to help disaster managers make more informed, socially driven decisions, by detecting and monitoring online social media sentiment over the course of a wildfire event. This monitoring model is coupled to a real-time forecasting of wildfire dynamics.

Real-time forecasting of wildfire dynamics, which has attracted increasing attention recently in fire safety science, is extremely challenging due to the complexities of the physical models and the geographical features. Running physics-based simulations for large-scale wildfires can be computationally difficult. We propose a novel algorithm scheme, which combines reduced-order modelling (ROM), recurrent neural networks (RNN), data assimilation (DA) and error covariance tuning for real-time forecasting/monitoring of the burned area. An operating cellular automata (CA) simulator is used to compute a data-driven surrogate model for forecasting fire diffusions. A long-short-term-memory (LSTM) neural network is used to build sequence-to-sequence predictions following the simulation results projected/encoded in a reduced-order latent space. 

We implement machine learning in a wildfire prediction model, using social media and geophysical data sources with sentiment analysis to predict wildfire instances and characteristics with high accuracy. The geophysical data is satellite data provided by the Global Fire Atlas, and social data is provided by Twitter. In doing this, we perform our own data collection and analysis, comparing regional differences in online social sentiment expression.

The performance of the proposed algorithm has been tested in recent massive wildfire events in California.

How to cite: Lever, J., Cheng, S., and Arcucci, R.: Social & Physics Based Data Driven Methods for Wildfire Prediction, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-15645, https://doi.org/10.5194/egusphere-egu23-15645, 2023.

The hybrid modelling approach combining machine learning and physics-based simulation has been used in a variety of ways to study tsunami and improve our understanding of this complex natural hazard. They are broadly applied for (1) Tsunami forecasting and early warning systems and (2) Tsunami hazard and risk assessment including sensitivity, analysis uncertainty studies and inverse modelling for estimating the source. 

Rigorous evaluation of such a hybrid approach is constrained by the limited size of available simulation datasets which is important to guide their usage by practitioners. This study investigates the application of a hybrid tsunami modelling technique (Ragu Ramalingam et al., 2022, Ragu Ramalingam et al., 2022) which offers a computationally efficient approach for hazard assessment where large events-sets must be modelled typical of probabilistic tsunami hazard and risk assessment (PTHA/PTRA). We use a large tsunami simulation dataset for a coastal region of eastern Sicily, Italy and try to address the following question:

• How to efficiently sample scenarios used to train the ML models?
• Where and when are such methods accurate? 
• How do they compare with other traditional modelling methods like Monte Carlo Sampling?

Additionally, the effort will deliver an open tsunami benchmarking dataset that can be utilised for further development, baseline comparison of various ML algorithms, and improved hyperparameter tuning.

References

Ragu Ramalingam, N., Johnson, K., Pagani, M., and Martina, M.: A hybrid ML-physical modelling approach for efficient approximation of tsunami waves at the coast for probabilistic tsunami hazard assessment, EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5642, https://doi.org/10.5194/egusphere-egu22-5642, 2022.

Ragu Ramalingam, N., Rao, A., Johnson, K., Pagani, M. and Martina, M. A hybrid ML-physical modelling approach for efficient probabilistic tsunami hazard and risk assessment, Proceedings of the 19th Annual Meeting of the Asia Oceania Geosciences Society (AOGS 2022), August 1-5, 2022, Virtual.

How to cite: Ragu Ramalingam, N., Abbate, A., Briseid Storrøsten, E., Johnson, K., Davies, G., Lorito, S., Pagani, M., and Martina, M.: Efficient Probabilistic Tsunami Hazard and Risk Assessment Using a Hybrid Modeling Approach: A Systematic Evaluation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-16906, https://doi.org/10.5194/egusphere-egu23-16906, 2023.

The terrestrial carbon cycle is one of the largest sources of uncertainty in climate projections. The terrestrial carbon sink which removes a quarter of anthropogenic CO2 emissions; is highly variable in time and space depending on climate. Previous studies have found that data-driven models such as random forest, artificial neural networks and long short-term memory networks can be used to accurately model Net Ecosystem Exchange (NEE) and Gross Primary Productivity (GPP) accurately, which are two important metrics to quantify the direction and magnitude of CO2 transfer between the land surface and the atmosphere. Recently, a new class of machine learning models called transformers have gained widespread attention in natural language processing tasks due to their ability to learn from large volumes of sequential data. In this work, we use Transformers to model NEE and GPP from 1996-2022 at 39 Flux stations in the ICOS Europe network using ERA5 reanalysis data. We can compare our results with traditional machine learning approaches to evaluate the generalisability and predictive performance of transformers for carbon flux modelling.

How to cite: Mazumder, S. and Prasad, A.: Modeling the Variability of Terrestrial Carbon Fluxes using Transformers, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-582, https://doi.org/10.5194/egusphere-egu23-582, 2023.

Weather forecasts today are typically issued in the form of ensemble simulations based on multiple runs of numerical weather prediction models with different perturbations in the initial states and the model physics. In light of the continuously increasing spatial resolutions of operational weather models, this results in large, high-dimensional datasets that nonetheless contain relevant spatial and temporal structure, as well as information about the predictive uncertainty. We propose invariant variational autoencoder (iVAE) models based on convolutional neural network architectures to learn low-dimensional representations of the spatial forecast fields. We specifically aim to account for the ensemble character of the input data and discuss methodological questions about the optimal design of suitable dimensionality reduction methods in this setting. Thereby, our iVAE models extend previous work where low-dimensional representations of single, deterministic forecast fields were learned and utilized for incorporating spatial information into localized ensemble post-processing methods based on neural networks [1], which were able to improve upon model utilizing location-specific inputs only [2]. By additionally incorporating the ensemble dimension and learning representation for probability distributions of spatial fields, we aim to enable a more flexible modeling of relevant predictive information contained in the full forecast ensemble. Additional potential applications include data compression and the generation of forecast ensembles of arbitrary size.

We illustrate our methodological developments based on a 10-year dataset of gridded ensemble forecasts from the European Centre for Medium-Range Weather Forecasts of several meteorological variables over Europe. Specifically, we investigate alternative model architectures and highlight the importance of tailoring the loss function to the specific problem at hand.

References:

[1] Lerch, S. & Polsterer, K.L. (2022). Convolutional autoencoders for spatially-informed ensemble post-processing. ICLR 2022 AI for Earth and Space Science Workshop, https://arxiv.org/abs/2204.05102.

[2] Rasp, S. & Lerch, S. (2018). Neural networks for post-processing ensemble weather forecasts. Monthly Weather Review, 146, 3885-3900.

How to cite: Chen, J., Höhlein, K., and Lerch, S.: Spatial representation learning for ensemble weather simulations using invariant variational autoencoders, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-1825, https://doi.org/10.5194/egusphere-egu23-1825, 2023.

The AtmoRep project asks if one can train one neural network that represents and describes all atmospheric dynamics. AtmoRep’s ambition is hence to demonstrate that the concept of large-scale representation learning, whose principle feasibility and potential was established by large language models such as GPT-3, is also applicable to scientific data and in particular to atmospheric dynamics. The project is enabled by the large amounts of atmospheric observations that have been made in the past as well as advances on neural network architectures and self-supervised learning that allow for effective training on petabytes of data. Eventually, we aim to train on all of the ERA5 reanalysis and, furthermore, fine tune on observational data such as satellite measurements to move beyond the limits of reanalyses.

We will present the theoretical formulation of AtmoRep as an approximate representation for the atmosphere as a stochastic dynamical system. We will also detail our transformer-based network architecture and the training protocol for self-supervised learning so that unlabelled data such as reanalyses, simulation outputs and observations can be employed for training and re-fining the network. Results will be presented for the performance of AtmoRep for downscaling, precipitation forecasting, the prediction of tropical convection initialization, and for model correction. Furthermore, we also demonstrate that AtmoRep has substantial zero-short skill, i.e., it is capable to perform well on tasks it was not trained for. Zero- and few-shot performance (or in context learning) is one of the hallmarks of large-scale representation learning and to our knowledge has never been demonstrated in the geosciences.

How to cite: Lessig, C., Luise, I., and Schultz, M.: AtmoRep: Large Scale Representation Learning for Atmospheric Data, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3117, https://doi.org/10.5194/egusphere-egu23-3117, 2023.

Numerical Earth system models (ESMs) are our primary tool for projecting future climate scenarios. Their simulation output is used by impact models that assess the effect of anthropogenic global warming, e.g., on flood events, vegetation changes or crop yields. Precipitation, an atmospheric variable with arguably one of the largest socio-economic impacts, involves various processes on a wide range of spatial-temporal scales. However, these cannot be completely resolved in ESMs due to the limited discretization of the numerical model. 
This can lead to biases in the ESM output that need to be corrected in a post-processing step prior to feeding ESM output into impact models, which are calibrated with observations [1]. While established post-processing methods successfully improve the modelled temporal statistics for each grid cell individually, unrealistic spatial features that require a larger spatial context are not addressed.
Here, we apply a cycle-consistent generative adversarial network (CycleGAN) [2] that is physically constrained to the precipitation output from Coupled Model Intercomparison Project phase 6 (CMIP6)  ESMs to correct both temporal distributions and spatial patterns. The CycleGAN can be naturally trained on daily ESM and reanalysis fields that are unpaired due to the deviating trajectories of the ESM and observation-based ground truth. 
We evaluate our method against a state-of-the-art bias adjustment framework (ISIMIP3BASD) [3] and find that it outperforms it in correcting spatial patterns and achieves comparable results on temporal distributions. We further discuss the representation of extreme events and suitable metrics for quantifying the realisticness of unpaired precipitation fields.

 [1] Cannon, A.J., et al. "Bias correction of GCM precipitation by quantile mapping: How well do methods preserve changes in quantiles and extremes?." Journal of Climate 28.17 (2015): 6938-6959.

[2] Zhu, J.-Y., et al. "Unpaired image-to-image translation using cycle-consistent adversarial networks." Proceedings of the IEEE international conference on computer vision. 2017.

[3] Lange, S. "Trend-preserving bias adjustment and statistical downscaling with ISIMIP3BASD (v1.0)." Geoscientific Model Development 12.7 (2019): 3055-3070.

How to cite: Hess, P., Lange, S., and Boers, N.: Improving global CMIP6 Earth system model precipitation output with generative adversarial networks for unpaired image-to-image translation, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3128, https://doi.org/10.5194/egusphere-egu23-3128, 2023.

Machine learning, and particularly neural networks, have been touted as a valuable accelerator for physical processes. By training on data generated from an existing algorithm a network may theoretically learn a more efficient representation and accelerate the computations via emulation. For many parameterized physical processes in weather and climate models this being actively pursued. Here, we examine the value of this approach for radiative transfer within the IFS, an operational numerical weather prediction model where both accuracy and speed are vital. By designing custom, physics-informed, neural networks we achieve outstanding offline accuracy for both longwave and shortwave processes. In coupled testing we find minimal changes to forecast scores at near operational resolutions. We carry out coupled inference on GPUs to maximise the speed benefits from the emulator approach.

How to cite: Chantry, M., Ukkonen, P., Hogan, R., and Dueben, P.: Emulating radiative transfer in a numerical weather prediction model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3256, https://doi.org/10.5194/egusphere-egu23-3256, 2023.

Dynamical models used in climate prediction often have systematic errors that can bias the predictions. In this study, we utilized machine learning to address this issue. Machine learning was applied to learn the error corrected by data assimilation and thus build a data-driven model to emulate the dynamical model error. A hybrid model was constructed by combining the dynamical and data-driven models. We tested the hybrid model using synthetic observations generated by a simplified high-resolution coupled ocean-atmosphere model (MAOOAM, De Cruz et al., 2016) and compared its performance to that of a low-resolution version of the same model used as a standalone dynamical model.

To evaluate the forecast skill of the hybrid model, we produced ensemble predictions based on initial conditions determined through data assimilation. The results show that the hybrid model significantly improves the forecast skill for both atmospheric and oceanic variables compared to the dynamical model alone. To explore what affects short-term forecast skills and long-term forecast skills, we built two other hybrid models by correcting errors either only atmospheric or only oceanic variables. For short-term atmospheric forecasts, the results show that correcting only oceanic errors has no effect on atmosphere variables forecasts but correcting only atmospheric variables shows similar forecast skill to correcting both atmospheric and oceanic errors. For the long-term forecast of oceanic variables, correcting the oceanic error can improve the forecast skill, but correcting both atmospheric and oceanic errors can obtain the best forecast skill. The results indicate that for the long-term forecast of oceanic variables, bias correction of both oceanic and atmospheric components can have a significant effect.

How to cite: He, Z., Brajard, J., Wang, Y., Wang, X., and Shen, Z.: Using machine learning to improve dynamical predictions in a coupled model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3321, https://doi.org/10.5194/egusphere-egu23-3321, 2023.

Data-driven models are becoming increasingly competent at tasks fundamental to weather and climate prediction. Relative to machine learning (ML) based atmospheric models, which have shown promise in short-term forecasting, ML-based ocean forecasting remains somewhat unexplored. In this work, we present a data-driven emulator of an ocean GCM and show that performance over a single predictive step is skilful across all variables under consideration. Iterating such data-driven models poses additional challenges, with many models suffering from over-smoothing of fields or instabilities in the predictions. We compare a variety of methods for iterating our data-driven emulator and assess them by looking at how well they agree with the underlying GCM in the very short term and how realistic the fields remain for longer-term forecasts. Due to the chaotic nature of the system being forecast, we would not expect any model to agree with the GCM accurately over long time periods, but instead we expect fields to continue to exhibit physically realistic behaviour at ever increasing lead times. Specifically, we expect well-represented fields to remain stable whilst also maintaining the presence and sharpness of features seen in both reality and in GCM predictions, with reduced emphasis on accurately representing the location and timing of these features. This nuanced and temporally changing definition of what constitutes a ‘good’ forecast at increasing lead times generates questions over both (1) how one defines suitable metrics for assessing data-driven models, and perhaps more importantly, (2) identifying the most promising loss functions to use to optimise these models.

How to cite: Furner, R., Haynes, P., Jones, D., Munday, D., Paige, B., and Shuckburgh, E.: An iterative data-driven emulator of an ocean general circulation model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-3340, https://doi.org/10.5194/egusphere-egu23-3340, 2023.

While efforts have been made to curb CO₂ emissions from aviation, the more uncertain non-CO₂ effects that contribute about two-thirds to the warming in terms of radiative forcing (RF), still require attention. The most important non-CO₂ effects include persistent line-shaped contrails, contrail-induced cirrus clouds and nitrogen oxide (NO_x) emissions that alter the ozone (O₃) and methane (CH₄) concentrations, both of which are greenhouse gases, and the emission of water vapour (H₂O). The climate impact of these non-CO₂ effects depends on emission location and prevailing weather situation; thus, it can potentially be reduced by advantageous re-routing of flights using Climate Change Functions (CCFs), which are a measure for the climate effect of a locally confined aviation emission. CCFs are calculated using a modelling chain starting from the instantaneous RF (iRF) measured at the tropopause that results from aviation emissions. However, the iRF is a product of computationally intensive chemistry-climate model (EMAC) simulations and is currently restricted to a limited number of days and only to the North Atlantic Flight Corridor. This makes it impossible to run EMAC on an operational basis for global flight planning. A step in this direction lead to a surrogate model called algorithmic Climate Change Functions (aCCFs), derived by regressing CCFs (training data) against 2 or 3 local atmospheric variables at the time of emission (features) with simple regression techniques and are applicable only in parts of the Northern hemisphere. It was found that in the specific case of O₃ aCCFs, which provide a reasonable first estimate for the short-term impact of aviation NO_x on O₃ warming using temperature and geopotential as features, can be vastly improved [1]. There is aleatoric uncertainty in the full-order model (EMAC), stemming from unknown sources (missing features) and randomness in the known features, which can introduce heteroscedasticity in the data. Deterministic surrogates (e.g. aCCFs) only predict point estimates of the conditional average, thereby providing an incomplete picture of the stochastic response. Thus, the goal of this research is to build a new surrogate model for iRF, which is achieved by :

1. Expanding the geographical coverage of iRF (training data) by running EMAC simulations in more regions (North & South America, Eurasia, Africa and Australasia) at multiple cruise flight altitudes,

2. Following an objective approach to selecting atmospheric variables (feature selection) and considering the importance of local as well as non-local effects,

3. Regressing the iRF against selected atmospheric variables using supervised machine learning techniques such as homoscedastic and heteroscedastic Gaussian process regression.

We present a new surrogate model that predicts iRF of aviation NO_x-O₃ effects on a regular basis with confidence levels, which not only improves our scientific understanding of NO_x-O₃ effects, but also increases the potential of global climate-optimised flight planning.

References

[1] Rao, P.; et al. Case Study for Testing the Validity of NO_x-Ozone Algorithmic Climate Change Functions for Optimising Flight Trajectories. Aerospace 2022, 9, 231. https://doi.org/10.3390/aerospace9050231

How to cite: Rao, P., Dwight, R., Singh, D., Maruhashi, J., Dedoussi, I., Grewe, V., and Frömming, C.: Towards a new surrogate model for predicting short-term NOx-O3 effects from aviation using Gaussian processes, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-4337, https://doi.org/10.5194/egusphere-egu23-4337, 2023.