Measurements in the remote unpolluted atmosphere have tremendous power to reveal processes that are happening on a global scale.   In the marine atmosphere where nitrogen oxide (NOx) levels are very low,  the photochemical loss rate of tropospheric ozone dominates over production, allowing loss processes to be sensitively explored.   We showed that bromine and iodine emitted from open-ocean marine sources initiate important global-scale catalytic ozone-destroying cycles and found that the deposition of ozone and subsequent reactions at the sea surface are a substantial pathway for production of volatile iodine.   Production of ozone in the remote atmosphere is predominantly regulated by the abundance of NOx, which also exerts substantial control over the hydroxyl radical (OH), the most important oxidant in the atmosphere.  It is now emerging that NOx regeneration pathways, namely the photolysis of particulate nitrate, could provide the dominant source of NOx to the marine atmosphere.  This has significant implications for our understanding of the chemistry of the remote troposphere.  This presentation discusses advances made in understanding these important, predominantly natural, cycles and their impacts on the atmosphere.

How to cite: Carpenter, L., Callaghan, A., Chance, R., Evans, M., Lee, J., Read, K., Rowlinson, M., Shaw, M., Sherwen, T., Andersen, S., Tinel, L., and Plane, J.: Discovering global-scale processes in the marine atmosphere, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11717, https://doi.org/10.5194/egusphere-egu24-11717, 2024.

Tropical cyclones (TCs), also referred to as hurricanes or typhoons, are amongst the deadliest and costliest natural hazards, affecting people, economies, and the environment in coastal areas around the globe when they make landfall. TCs are projected to become more intense in a warming climate, enhancing the risks associated with their wind speeds, precipitation and storm surges. It is therefore crucial to minimize future loss of life and by performing accurate TC risk assessments for coastal areas. Calculating TC risk at a global scale, however, has proven to be difficult, given the limited temporal and spatial information on landfalling TCs around much of the global coastline, and how this is going to change under climate change.

To overcome these limitations, we developed a novel approach to calculate TC risk under present and future climate conditions using the Synthetic Tropical cyclOne geneRation Model (STORM). STORM is a fully statistical model that can take any input dataset and statistically resamples this to an equivalent of 10,000 years of TC activity under the same climate condition. The resulting publicly available STORM dataset contains of enough TC activity in any coastal region of interest to adequately calculate TC probabilities and risk from. Furthermore, the STORM algorithm has been expanded with a future-climate module, enabling globally consistent local-scale assessments of (changes in) TC risk. This presentation will discuss the challenges and opportunities in using such synthetic datasets, particularly in the light of improving our understanding of TC risk. 

How to cite: Bloemendaal, N.: Weathering the STORM: Challenges and opportunities in tropical cyclone risk research , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21254, https://doi.org/10.5194/egusphere-egu24-21254, 2024.

Numerical weather prediction models are utilized to forecast extreme rainfall events at fine resolutions; however, these models possess inherent errors due to the parameterization and discretization of differential equations, which diminishes simulation accuracy. Recent advancements in machine learning methods indicate their potential capability to significantly enhance forecast results. In this study, multiple extreme rainfall events for the Pamba river basin during the Indian Summer Monsoon Period spanning 2-4 days were forecasted using the WRF model. Pamba, one of the flood-prone basins in southern states of India (Kerala), experiences severe flood events annually. While numerous studies have been conducted to simulate the Kerala flood of 2018, those demonstrating the application of high-resolution rainfall data for the Pamba river basin remain unexplored. Therefore, in this study, we simulated multiple storm events during the period from 2015 to 2018 using the WRF model at a high resolution (1 km * 1 km) and a temporal resolution of a one-hour interval. The WRF-simulated rainfall dataset was further post-processed using various machine learning algorithms, including Random Forest, Support Vector Machine, and extreme gradient boost (XGBoost), to reduce bias in the hourly forecast of extreme rainfall events. Several cross-validation training and testing procedures were carried out using various algorithms, and the forecasted and predicted results were compared with ERA5 hourly data of 10*10 km resolution. Results indicated that XGBoost, with hyperparameter tunings, substantially reduced the Root Mean Square Error (RMSE); it was able to reduce the RMSE by up to 50% across the river basin. This hybrid model will provide a more accurate forecast of hourly extreme rainfall during the Indian Summer Monsoon Period for Pamba river basin, with high resolution essential for flood forecasting and warning.

How to cite: Krishnankutty Nair, A. and Singh, S.: Improving extreme rainfall forecasting for a flood prone region: A hybrid modelling approach, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-857, https://doi.org/10.5194/egusphere-egu24-857, 2024.

The Atmospheric Boundary Layer Height (ABLH) is a fundamental parameter for many meteorological applications and climate change assessment and evaluation. A large number of methods for ABLH determination have been proposed; however, there is no sufficiently reliable and feasible method for this purpose. The rise of intelligence-based mathematical models for feature determination in data space has allowed their application to solve problems similar to ABLH determination. This article describes the development of a mathematical model based on artificial intelligence for ABLH determination, in which an introductory analysis of the data space from the point of view of machine learning, unsupervised, and supervised methods is presented. The methods explored are the mountain method, subtractive clustering, and the classic K-means and its soft counterpart, Fuzzy C-means. Furthermore, an analysis was conducted to determine which similarity function—whether Euclidean, Manhattan, Mahalanobis, or Cosine—best fits for ABLH estimation in each unsupervised method. For classification in a supervised fashion, the best suitable models, among others, are support vector machines and decision trees. Different internal metrics (Silhouette Index, Calinski-Harabasz score) and external metrics (root mean square error and adjusted Rand score), with a reference made by means of visual inspection by an expert, were used to evaluate the methods. The unsupervised mountain method with the Manhattan similarity function proved to be the most feasible, as it is a non-stochastic method, its computation time is reduced, and it does not require an ABLH reference. The data used was extracted from several sources: 83 days of quasi-continuous LIDAR data with 23,000 data points located at Brest, France, measured with a MiniMPL from the Meteo France LIDAR network, were used. An ABLH reference from a radiosonde adjacent to the site of the LIDAR system was used. The references range from October to December 2018. The root mean square error achieved for the whole dataset was 600 m for the mountain method. The presented method is shown to be effective for various atmospheric situations, regardless of their complexity.

How to cite: Estrada, S. and Quintero Montoya, O. L.: Development of a mathematical model for the determination of the atmospheric boundary layer height using artificial intelligence, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1043, https://doi.org/10.5194/egusphere-egu24-1043, 2024.

The U.S. Navy is building a coupled thermosphere-ionosphere prediction system.  As part of this project, we are developing a new dynamical core (DyCore) extending from the ground to the exobase (~500 km).  The DyCore must be able to handle large variations in both temperature and composition, which motivates a new Horizontally Explicit Vertically Implicit (HEVI) time integrator.  Unlike traditional linear Implicit-EXplicit (IMEX) methods commonly used in numerical weather prediction (NWP), HEVI does not require a fixed reference state.  Our DyCore combines HEVI with a Specific Internal Energy Equation (SIEE) and a Spectral Element Method (SEM) spatial discretization to form a robust, whole-atmosphere model for the neutral atmosphere.  We present results for two test cases using the proposed DyCore: an idealized heating/cooling test extending into the middle thermosphere and a perturbation experiment yielding nonhydrostatic baroclinic instability. The idealized heating/cooling test, which is compared to corresponding results from the hydrostatic Navy Global Environmental Model (NAVGEM), demonstrates that HEVI is more robust than traditional linear IMEX methods.  The baroclinic instability test shows that HEVI, when combined with a banded lower-upper (LU) direct solve, is efficient and allows a large timestep.  These numerical results suggest that our HEVI-enabled DyCore is a good candidate for the proposed thermosphere-ionosphere prediction system.

This work was funded by the Office of Naval Research Marine Meteorology and Space Weather program.

How to cite: Kelly, J. F., Giraldo, F. X., Reinecke, P. A., Alves, F., Barton, C. A., and Eckermann, S. D.: Horizontally Explicit Vertically Implicit (HEVI) Time-Integrators for a Non-Hydrostatic Whole Atmosphere Models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2375, https://doi.org/10.5194/egusphere-egu24-2375, 2024.

Existing weather models are known to have poor skill at forecasting rainfall over East Africa, where there are regular threats of drought and floods. Improved precipitation forecasts could reduce the effects of these extreme weather events and provide significant socioeconomic benefits to the region. We present a novel machine learning based method to improve precipitation forecasts in East Africa, using postprocessing based on a conditional generative adversarial network (cGAN). This addresses the challenge of realistically representing tropical rainfall in this region, where convection dominates and is poorly simulated in conventional global forecast models. We postprocess hourly forecasts made by the European Centre for Medium-Range Weather Forecasts Integrated Forecast System at 6-18h lead times, at 0.1^o resolution. We combine the cGAN predictions with a novel neighbourhood version of quantile mapping, to integrate the strengths of both machine learning and conventional postprocessing. Our results indicate that the cGAN substantially improves the diurnal cycle of rainfall, and improves rainfall predictions up to the 99.9^th percentile of rainfall. This improvement persists when evaluating against the 2018 March-May season, which had extremely high rainfall, indicating that the approach has some ability to generalise to more extreme conditions. We explore the potential for the cGAN to produce probabilistic forecasts and find that the spread of this ensemble broadly reflects the predictability of the observations, but is also characterised by a mixture of under- and over-dispersion. Overall our results demonstrate how the strengths of machine learning and conventional postprocessing methods can be combined, and illuminate what benefits machine learning approaches can bring to this region.

How to cite: Antonio, B., McRae, A., MacLeod, D., Cooper, F., Marsham, J., Aitchison, L., Palmer, T., and Watson, P.: Postprocessing East African rainfall forecasts using a generative machine learning model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2897, https://doi.org/10.5194/egusphere-egu24-2897, 2024.

The UFS-R2O Project, which began in July 2020 as a five-year plan with deliverables for the first three years funded, has made significant progress in developing the medium-range and sub-seasonal to seasonal (MRW/S2S) predictions, a regional, high-resolution hourly-updating and convection-allowing ensemble system for prediction of short range severe weather (CAM/SRW), and a Hurricane Application developing a very high-resolution Hurricane Analysis and Forecast System (HAFS) with storm following moving nests.  The Project is organized with Application Teams and Development Teams interacting with each other to reflect the cross-cutting nature of the UFS components and infrastructure. It  fostered successful collaborations between the National Centers for Environmental Prediction (NCEP) Environmental Modeling Center, several NOAA research labs, the National Center for Atmospheric Research (NCAR), the Naval Research Lab (NRL), and multiple universities and cooperative institutes.  Most sIgnificant outcomes of the project thus far are the implementation of the HAFSv1 ahead of the schedule, and the development of a six-way global coupled (atmosphere/ ocean/ land/ sea-ice/ wave/ aerosol) modeling system, both within the UFS framework, major accomplishments from the community modeling perspective.  

The UFS-R2O Project has entered into its second phase (2023-2024), albeit with reduced funding, to continue the momentum built during the first phase.  While the first three years of the project were focused on engineering and infrastructure, Phase II is primarily targeting systematic testing and evaluation of the prototype UFS configurations for selecting the candidates for potential transition to operations in the next few years.  In addition, Phase II of the project includes a new Seasonal Forecast System (SFS) Application Tean established to develop SFS v1 that will replace the legacy Climate Forecast System (CFSv2) currently in operations since 2011.

This presentation describes the outcomes of the UFS R2O Project for the first three years, and highlights the progress and plans for the Phase II.

How to cite: Tallapragada, V., Whitaker, J., and Kinter, J.: NOAA's Unified Forecast System Research to Operations (UFS R2O) Project Phase II - Accomplishments, Progress and Future Plans, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2930, https://doi.org/10.5194/egusphere-egu24-2930, 2024.

Satellite sea surface wind fields can compensate for the shortcomings of conventional observation data, thereby improving the forecasting skills of global medium-range numerical weather models. China successfully launched the HY2D satellite carrying a Ku band microwave scatterometer (HSCAT) on May 19, 2021. It can provide a large amount of high-quality sea surface wind field data for numerical forecasting models. In order to test the potential application of HY2D sea surface wind field data in the Global Assimilation Forecasting System of the China Meteorological Administration (CMA-GFS). A three-step study was conducted, with the first step being the timeliness evaluation of HY2D wind, followed by the quality evaluation of HY2D wind using ERA5 and buoy data, and finally assessment of impacts of the HY2D wind assimilation on the analyses and forecasts. Two sets of assimilation experiments were conducted: the control experiment without HY2D wind (CTRL) and sensitivity experiment with HY2D wind based on a new quality control scheme (HY2D-FlagQC). The experimental period is from March 1, 2023 to April 1, 2023. The results show that the timeliness of HY2D wind field obtained through National Satellite Ocean Application Service (NSOAS) has been improved by about 20% compared to Koninklijk Nederlands Meteorologisch Instituut (KNMI). But the timeliness fluctuation is relatively large in terms of time and space. The root mean square error of HY2D wind field is less than 2m/s. After assimilating the HY2D wind, the analysis errors of the wind fields in the lower-middle troposphere of the tropics and the southern hemisphere (SH) are significantly reduced. Furthermore, assimilating the HY2D wind data can improve the forecast skill of wind, geopotential height, and temperature in the troposphere of the tropics and SH. 

How to cite: Wei, C., Han, W., Liu, Y., Hu, H., Lu, H., Xiao, H., and Wang, D.: Assimilation of HY-2D scatterometer wind field data in CMA-GFS, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3462, https://doi.org/10.5194/egusphere-egu24-3462, 2024.

Aerosols and droplets are the main factors of visibility reduction by scattering and absorbing light. For visibility predictions in operational NWP models, hydrometeors are often considered to be the dominant factor in the total extinction, whereas aerosol effects are usually simplified or omitted in models developed for relatively clean regions. In China, also many NWP studies related to visibility forecasting during haze-fog processes have been conducted, primarily focusing on severely polluted periods before 2018. These studies often employed visibility parameterizations that considered either aerosol extinction alone or hydrometeor extinction alone. Therefore, the significance of incorporating both aerosol and hydrometeor extinction into visibility forecasting during haze-fog processes remains uncertain, particularly under recent rapid changes in aerosol concentration, composition, and hygroscopicity in China.

In this study, we first use the 3-D meteorology fields from the Shanghai Meteorological Service WRF-ADAS Real-Time Modeling System (WARMS) to drive the Community Multiscale Air Quality (CMAQ) model. In this version, CMAQ is used in an off-line mode and visibility is diagnosed by combining extinctions due to hydrometeors and aerosols. Satellited derived NOx emissions using the Daily Emissions Constrained by Satellite Observations (DECSO) algorithm have been incorporated to give more up-to-date emissions. We analyze the results of a one-month forecasting period during the winter of 2021-2022 to assess the model's performance and understand the impact of hydrometeor and aerosol extinction on operational visibility forecasting. We find that for the city of Shanghai, aerosol extinction has a minor impact on the model’s performance when forecasting visibility below 1 km but becomes crucial for predictions spanning 1-10 km. Comparison against observations shows that the model well captures the general contributions from various chemical constituents with nitrate as the most important factor in aerosol extinction (~60%). Furthermore, our assessment of the North China Plain (NCP) highlights that in highly polluted areas aerosols could be significant for visibility below 1 km. Finally, we conduct case studies with the fully coupled WRF-Chem model and compare results with the offline WARMS-CMAQ system. Aerosol effects on fog and visibility forecasting due to feedbacks between aerosols, radiation, and cloud physics will be discussed.

How to cite: Xie, Y., Wang, X., Gao, Y., Chen, B., van der A, R., Ding, J., Gu, W., Zhou, M., and Wang, H.: Improving Visibility Forecasting during Haze-fog Processes in Shanghai and Eastern China: the Significance of Aerosol and Hydrometeor, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3686, https://doi.org/10.5194/egusphere-egu24-3686, 2024.

This study explores the potential impact of global navigation satellite system radio occultation (RO) data assimilation on the tropical cyclone (TC) intensity forecast over the western North Pacific. The forecast experiments are performed through a regional model for six TCs occurring in 2020. RO data are mainly obtained from the Constellation Observing System for Meteorology, Ionosphere, and Climate Mission II. The forecasts with and without assimilation of RO data are compared, and their difference is regarded as the impact of RO data on TC forecasts. Overall, the forecasts tend to underestimate the TC intensity relative to the best track data. Compared to the forecasts assimilating without RO data, forecasts assimilating with RO data improve the initial conditions and reduce the underestimation of TC intensity forecast by 13 kt and 16 hPa in subsequent forecasts. This intensity improvement is more significant for TCs developing in drier environments than those in moister environments. The main period of intensity increase is 48-24 h prior to TCs developing to the maximum intensity. The assimilation of RO data increases the moisture around the TC centers, especially at mid-levels (700-300 hPa). It also increases the low-level vorticity but reduces the mid-level vorticity around the TC centers. These characteristics favor TCs with stronger surface wind speed and lower sea surface pressure. In summary, this study highlights the positive contribution of RO data to TC intensity forecast and explores the potential mechanisms.

How to cite: Teng, H.-F.: Impact of radio occultation data assimilation on tropical cyclone intensity forecast over the western North Pacific, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4760, https://doi.org/10.5194/egusphere-egu24-4760, 2024.

Data assimilation is a widely used method for estimating the state and associated uncertainties in numerical models. While ensemble-based approaches are common, their computational expense arises from necessary ensemble integrations. This study improves the Weather Research and Forecasting–Advanced Research WRF (WRF-ARW) model by integrating it with the Parallel Data Assimilation Framework (PDAF) in a fully online mode. Through minimal modifications to the WRF-ARW code, an efficient data assimilation system is developed, leveraging parallelization and in-memory data transfers to minimize file I/O and model restarts during assimilation. The clear separation of concerns between method development and model application, facilitated by PDAF's model-agnostic structure, is an advantage. Evaluating the assimilation system through a twin experiment simulating a tropical cyclone reveals improved accuracy in temperature, U and V fields. The assimilation process incurs minimal overhead in run time compared to the model without data assimilation, demonstrating excellent parallel performance. Consequently, the online WRF-PDAF system proves to be an efficient framework for high-resolution mesoscale forecasting and reanalysis.

How to cite: Shao, C.: Augmenting WRF with PDAF for an Online Localized Ensemble Data Assimilation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4851, https://doi.org/10.5194/egusphere-egu24-4851, 2024.

Data-driven machine learning methods for weather forecasting have experienced a steep progress over the last years, with recent studies demonstrating substantial improvements over physics-based numerical weather prediction models. Beyond improved forecasts, the major advantages of purely data-driven models are their substantially lower computational costs and faster generation of forecasts, once a model has been trained. However, in contrast to ensemble forecasts from physical weather models, most efforts in data-driven weather forecasting have been limited to deterministic, point-valued predictions only, making it impossible to quantify forecast uncertainties which is crucial for optimal decision making in applications.

Our overarching aim is to evaluate and compare methods for creating probabilistic forecasts from data-driven weather models. The uncertainty quantification (UQ) approaches we compare are either based on generating ensemble forecasts from data-driven weather models via perturbations to the initial conditions, or based on statistical post-hoc UQ methods. The perturbation-based methods either leverage initial conditions from the ECMWF IFS ensemble, add random Gaussian noise to the deterministic initial conditions, or add random field perturbations based on past observations (Magnusson et al., 2009). The post-hoc approaches operate on deterministic forecasts and quantify forecast uncertainty using established post-processing methods, namely distributional regression networks (Rasp and Lerch, 2018) and isotonic distributional regression (Walz et al., 2022; Henzi et al., 2021).

Using forecasts from Pangu-Weather (Bi et al., 2023), we evaluate these UQ methods over Europe for selected user-relevant weather variables, such as wind speed at 10 m, temperature at 2 m, and geopotential height at 500 hPa. We focus on daily initialised Pangu-Weather forecasts for 2022 with a forecast horizon of up to 7 days and compare their performance against ECMWF IFS ensemble forecasts. Our results suggest that Pangu-Weather predictions combined with UQ approaches yield improvements over the ECMWF ensemble forecasts for lead times of up to 5 days in terms of the Continuous Ranked Probability Score. However, it strongly depends on the variable of interest which of the UQ methods performs best, none of the different UQ methods performs best over all variables and lead times. Post-hoc UQ methods tend to perform better for shorter lead times, while initial condition perturbations are superior for longer lead times, with in particular the random field method showing promising results.

References:

• Bi, K., Xie, L., Zhang, H., Chen, X., Gu, X. and Tian, Q. (2023). Accurate medium-range global weather forecasting with 3D neural networks. Nature, 619, 533–538.
• Henzi, A., Ziegel, J. F. and Gneiting, T. (2021). Isotonic distributional regression. Journal of the Royal Statistical Society Series B: Statistical Methodology, 83, 963–993.
• Magnusson, L., Nycander, J. and Källén, E. (2009). Flow-dependent versus flow-independent initial perturbations for ensemble prediction. Tellus A: Dynamic Meteorology and Oceanography, 61, 194.
• Rasp, S. and Lerch, S. (2018). Neural networks for postprocessing ensemble weather forecasts. Monthly Weather Review, 146, 3885–3900.
• Walz, E.-M., Henzi, A., Ziegel, J. and Gneiting, T. (2022). Easy Uncertainty Quantification (EasyUQ): Generating Predictive Distributions from Single-valued Model Output. Preprint, available at https://arxiv.org/abs/2212.08376.

How to cite: Horat, N., Bülte, C., Quinting, J., and Lerch, S.: Uncertainty quantification for data-driven weather models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5395, https://doi.org/10.5194/egusphere-egu24-5395, 2024.

The Global Navigation Satellite System (GNSS) ground-based network in Europe is a comparatively dense network that provides valuable humidity information through Zenith Total Delays (ZTDs) and tropospheric gradients. ZTDs include information on column water vapor, while tropospheric gradients provide information on moisture distribution. Recently, we developed the tropospheric gradient operator (Zus et al., 2023) and implemented it in the Weather Research and Forecasting (WRF) model (Thundathil et al., 2023, under review).

We have conducted ZTD and tropospheric gradient assimilation experiments over a couple of periods, which lasted for two months. We will present our latest test period, the Benchmark Campaign organized within the European COST Action, in May and June 2013. Data from more than 250 GNSS stations in central Europe covering Germany, the Czech Republic, and part of Poland and Austria were assimilated during this period. The data assimilation (DA) system used a rapid update cycle of 3-dimensional variational DA with 6-hourly cycles for two months.

Our research methodology involved configuring a 0.1 x 0.1-degree mesh in the WRF model with 50 vertical levels up to 50 hPa for Europe. Model forcing was done with the European Centre for Medium-Range Weather Forecasts (ECMWF) operational analysis. We conducted three runs, which included the assimilation of conventional datasets from ECMWF (or control run), ZTD added on top of the control run, and ZTD and gradients on top of the control run. We observed a significant reduction of the root mean square errors; we observed a 42 % and 16 % reduction for ZTDs and gradients in the ZTD assimilation run, which further reduced to 43 % and 21 % for ZTDs and gradients in the ZTD and gradient assimilation. Validation with the atmospheric reanalysis ERA5 and radiosondes revealed improvements in the lower troposphere.

We conducted an additional sensitivity experiment using a sparsely distributed GNSS network. This process involved reducing the station density from roughly 0.5 degrees to 1 degree by replacing the original network with one consisting of 100 stations. We found that the improvement in the humidity field with the assimilation of ZTD and gradients from the sparse station network (1-degree resolution) is roughly the same as in the humidity field with the assimilation of ZTD only from the dense station network (0.5-degree resolution). Therefore, the assimilation of gradients in addition to ZTDs is particularly interesting in regions with a few GNSS stations. It may also be considered a cost-effective way to increase the density of networks.

After preliminary testing of the GNSS ZTD plus gradient assimilation with WRF, we are ready to move to convective-scale assimilation using an ensemble-based approach over different regions and seasons. We will be presenting initial results from our high-resolution simulations.

References

Zus, F., Thundathil R., Dick G., and Wickert J. "Fast Observation Operator for Global Navigation Satellite System Tropospheric Gradients." Remote Sensing 15, no. 21 (2023): 5114.

Thundathil, R. M., Zus, F., Dick, G., and Wickert, J. "Assimilation of GNSS Tropospheric Gradients into the Weather Research and Forecasting Model Version 4.4.1", Geoscientific Model Development Discussion [preprint], in review, 2023.

How to cite: Thundathil, R., Zus, F., Dick, G., and Wickert, J.: Impact of GNSS tropospheric gradient assimilation and sensitivity analysis, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5656, https://doi.org/10.5194/egusphere-egu24-5656, 2024.

Ensemble forecasts play a pivotal role in weather prediction, providing valuable insights into the inherent uncertainty of atmospheric processes. Strategies in ensemble construction involve generating multiple simulations by perturbing initial conditions, model parameters, or both. This diverse set of forecasts allows meteorologists to capture a range of possible future scenarios, acknowledging the inherent complexity of the atmosphere. Model resolution is a critical factor, influencing the representation of small-scale features and improving the overall accuracy of ensemble predictions. Additionally, forecast range-related issues address the challenge of extending predictions beyond a few days, where uncertainties tend to grow. Combining advanced statistical techniques with cutting-edge modeling technologies helps refine ensemble forecasts, enhancing our ability to anticipate and mitigate the impacts of weather-related events on society and the environment.

The investigation based operational global ensemble forecast system from NCEP, CMC, ECMWF and CMA to focus on the analyses of ensemble design that combined to the data assimilation for initial condition perturbation and various stochastic physical perturbations. The impact of model resolutions (both horizontal and vertical) will be addressed to the different atmospheric characteristics, such as forecast uncertainty, reliability and resolution. The forecast capability and predictability to the extreme events will be discussed from single model ensemble and multi-model ensemble. Finally, the 1st-moment and 2nd-moment ensemble forecast calibration will be demonstrated from traditional statistical method and machine learning based ensemble reforecasts

How to cite: Zhu, Y.: An assessment of prediction and predictability through the state-of-the-art global ensemble forecast systems, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6802, https://doi.org/10.5194/egusphere-egu24-6802, 2024.

One of the most hazardous windstorms was observed in Switzerland on February 15, 1925. The storm is categorized as a 'High-impact Foehn Storm' that affected all Foehn regions of Switzerland. All communities, stables, and houses were wholly or partially damaged in the canton of Glarus. In previous work, the Weather Research and Forecasting Model (WRF) was used for downscaling the storm from the Twentieth Century Reanalysis (20CRv2) down to a grid width of 3 km. While many storm features were realistically simulated, wind speeds in the Glarus Valley, where most damage occurred, remained well below the expected values. Here, we go one step further by using a Large-Eddy Simulation model (LES) to analyze whether high gust peaks would occur at the bottom of the valley. For this, the PArallelized Large-eddy simulation Model (PALMv6.0) is coupled to WRFv4.1.2. In the first stage, WRFv4.1.2 was downscaled to a resolution of 1x1 km² by using the "Twentieth Century Reanalysis" (20CRv3) as a boundary condition. Three nested domains with resolutions 25km, 5 km, and 1 km were set up for the simulation experiment. The second stage involves downscaling PALMv6.0 to a resolution of 20 m by using the output of WRFv4.1.2 as a boundary condition. The simulation shows strong winds between Netstal and Näfels on Earth's surface. Peak gusts of 40 m/s and more hit the valley floor south of Näfels. Strong turbulence fields reaching the ground at high velocities are observed in the central valley in the south-north direction. The simulation shows good agreement with the damage described, and the simulated peak gusts easily reach the measured maxima of extreme storms. Being able to realistically simulate the local characteristics of a Foehn storm that occurred a century back opens a new window to quantitative analyses of past extremes and their impacts.

How to cite: Shastri, R. P., Brönnimann, S., and Stucki, P.: Numerical Investigation of High Impact Foehn storm in February 1925 using WRF and PALM models., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6813, https://doi.org/10.5194/egusphere-egu24-6813, 2024.

The Cross-track Infrared Sounder (CrIS) observations (O) contributed greatly to numerical weather prediction. Further contribution depends on the success of all-sky data assimilation, which requires a method to produce realistic cloud/rain band structures from background fields (i. e., 6-h forecasts), and to remove large biases of all-sky simulation of brightness temperature in the presence of clouds. In this study, CrIS all-sky simulations of brightness temperatures at an arbitrarily selected window channel within Typhoon Hinnamnor (2022) are investigated. The 3-km Weather Research and Forecasting model with three microphysics schemes were used to produce 6-h background forecasts (B). The O−B statistic deviate greatly from Gaussian distribution with large biases in either water clouds, or thin ice clouds, or thick ice clouds within Typhoon Hinnamnor. By developing a linear regression function of three all-sky simulations of brightness temperature from 6-h forecasts with three microphysics schemes, the O−B statistics approximate a Gaussian normal distribution in water clouds, thin ice clouds and thick ice clouds. Taking the regression function that is established by a training dataset to combine 6-h background forecasts at later times, the cloud/rain band structures compared much more favorably with CrIS observations than those from an individual microphysics, and the O−B biases are significantly reduced. The work in this study to quantify and remove biases in background fields of brightness temperature and generating realistic typhoon cloud/rain band structures in background fields will allow a better description of center position, intensity and size to improve typhoon forecasts.

How to cite: Niu, Z.: Improving All-sky Simulations of Typhoon Cloud/Rain Band Structures of NOAA-20 CrIS Window Channel Observations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6917, https://doi.org/10.5194/egusphere-egu24-6917, 2024.

The Advanced Weather Research Forecast model (WRF-ARW) is used to investigate the potential impacts of assimilating the FengYun-4A (FY-4A) Geostationary Interferometric Infrared Sounder (GIIRS) long-wave temperature sounding channel data on prediction of Typhoon In-Fa (2021). In addition, a series of data assimilation experiments are conducted to demonstrate the added value of the FY-4A/GIIRS data assimilation for typhoon forecasts. It is shown that the higher spectral resolution and broader coverage of GIIRS radiance data can positively impact the model analysis and forecasts with larger temperature and moisture increments at the initial time of simulations, thus producing the better simulation for typhoon warm core aloft, vortex wind structure and spiral rainfall band. Moreover, the assimilation of the GIIRS data can also lead to better storm steering flows and consequently better typhoon track forecasts. Overall, the assimilation of FY-4A/GIIRS temperature sounding channel data shows some added values to improve the track and storm structure forecasts of Typhoon In-Fa.

How to cite: Zhang, L., Niu, Z., Weng, F., Dong, P., Huang, W., and Zhu, J.: Impacts of Direct Assimilation of the FY-4A/GIIRS Long-Wave Temperature Sounding Channel Data on Forecasting Typhoon In-Fa (2021), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6940, https://doi.org/10.5194/egusphere-egu24-6940, 2024.

MPAS-JEDI, a relatively-new data assimilation (DA) system for the Model for Prediction Across Scales – Atmosphere (MPAS-A) based upon the Joint Effort for Data assimilation Integration (JEDI), allows to assimilate cloud-/precipitation-affected satellite microwave and infrared radiance data to analysis microphysical parameters, e.g., mixing ratios of hydrometeors. Global cycling DA experiments were conducted in the context of MPAS-JEDI’s hybrid-3DEnVar configured at 30km resolution with 80-member ensemble input at 60km that is produced using MPAS-JEDI's ensemble of 3DEnVar. The benchmark experiment assimilates conventional observations plus clear-sky radiances from AMSU-A and MHS. All-sky experiments add the assimilation of all-sky microwave (MW) radiances from AMSU-A’s and/or ATMS’s window channels over water as well as infrared (IR) channels of two geostationary sensors GOES-ABI and Himawari-AHI. In addition to the impact assessment on dynamic and thermodynamic variables, we investigated more the impact on cloud forecasts in terms of fitting to ABI/AHI radiance data at different wavelengths. The community radiative transfer model (CRTM) is used as the observation operator in both all-sky radiance DA and evaluation. The substantial positive impact on cloud forecasts was obtained with all-sky microwave DA (individually or collectively from AMSU-A and ATMS) in terms of a better forecast fitting to observed ABI/AHI channel 13's radiances up to 7 days, especially over tropical regions, where the day-1 forecast root-mean-square error is reduced up to 10%. Cloud forecast impact from assimilating all-sky ABI/AHI 3 water vapor channels' radiances is more limited although a clear benefit is seen for middle/upper troposphere moisture field, which is consistent with ABI/AHI water vapor channels' sensitivity height. Future research direction for all-sky MW and IR radiance DA with MPAS-JEDI will also be discussed.

How to cite: Liu, Z., Ban, J., and Banos, I.: Improving cloud forecasts with assimilation of cloud-/precipitation-affected microwave and infrared radiances using MPAS-JEDI, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7021, https://doi.org/10.5194/egusphere-egu24-7021, 2024.

We developed a rice paddy model based on the Noah LSM considering the standing water layer during the irrigation period. In the model, we adopted a consistent subcanopy process from thin to thick canopy conditions and considered small scalar roughness length of water surface in rice paddy field. We evaluated the model’s performance against observations from three rice paddy sites with different leaf area index (LAI) and water depths during the growing season. Two simulations were performed in an offline mode: the fixed irrigation simulation of Noah LSM with saturation moisture in the top two soil layers during the irrigation period (IRRI) and the developed model simulation (RICE). The evaluation results showed that RICE outperformed IRRI in the simulating ground, sensible (H) and latent heat (LH) fluxes and topsoil temperature (T_soil) on hourly and diurnal time scales. Two sensitivity tests of RICE were performed in relation to the subcanopy resistance and standing water layer: RICE without consideration of small roughness length of water surface during the irritation period (BARE) and RICE with a constant standing water depth (FIX). The sensitivity tests showed that BARE calculated very low subcanopy resistance values when the sum of LAI and stem area index was less than 2 m² m^-2, which resulted in cold biases in the daily mean T_g and T_soil and also led to overestimation of daily mean LH. There was no significant difference in RICE and FIX with hourly and seasonal time scale statistics, suggesting that H, LH,  T_g and T_soil of the developed model are not sensitive to changes in water depth. The structure of the developed model was also discussed.

How to cite: Lim, H.-J. and Lee, Y.-H.: Development of Rice Paddy Model Based on Noah LSM: Consistent Parameterization of Subcanopy Resistance from the Ponded Water to Dense Rice Canopy , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7968, https://doi.org/10.5194/egusphere-egu24-7968, 2024.

An experiment reported in Mesinger and Veljovic (JMSJ 2020) and at the preceding EGU General Assembly, showed an advantage of the Eta over its driver ECMWF ensemble members in placing 250 hPa jet stream winds east of the Rockies.  Verifications subsequent to 2020 confirmed this advantage.  A byproduct of that experiment was that of the Eta ensemble switched to use sigma, Eta/sigma, also achieving 250 hPa wind speed scores better than their driver members, although to a lesser extent.  It follows that the Eta must include feature or features additional to the eta coordinate responsible for this advantage over the ECMWF.

An experiment we have done strongly suggests that the van Leer type finite-volume vertical advection of the Eta, implemented in 2007, may be a significant contributor to this advantage.  In that experiment, having replaced a centered finite-difference Lorenz-Arakawa scheme, this finite-volume scheme enabled a successful simulation of an intense downslope windstorm in the lee of the Andes.

Another likely and perhaps unique feature of the Eta contributing to that advantage is its sophisticated representation of topography, designed to arrive at the most realistic grid-cell values with no smoothing (Mesinger and Veljovic, MAAP 2017).

While apparently a widespread opinion is that it is a disadvantage of terrain intersecting coordinates that “vertical resolution in the boundary layer becomes reduced at mountain tops as model grids are typically vertically stretched at higher altitudes (Thuburn, 10.1007/978-3-642-11640-7 2011),” a comprehensive 2006 NCEP parallel test gave the opposite result.  With seemingly equal PBL schemes, the Eta showed a higher surface layer accuracy over high topography than the NMM, using a hybrid terrain-following system (Mesinger, BLM 2023).

Hundreds of thousands of the Eta forecasts and experiments performed demonstrate that the relaxation lateral boundary condition, almost universally used in regional climate models (RCMs), in addition to conflicting with the properties of the basic equations used, is unnecessary.  Similarly, so-called large scale or spectral nudging, frequently applied in RCMs, based on an ill-founded belief, should only be detrimental if possible numerical issues of the limited area model used are addressed.  Note that this is confirmed by the Eta vs ECMWF results we refer to above.

Even so, to have large scales of a nested model ensemble members most times more accurate than those of their driver members, surely requires not only the absence of detrimental techniques, but also the use of a lateral boundary condition (LBC) scheme that is not inducing major errors.  The scheme of the Eta is at the outflow points of the boundary prescribing one less condition than at the inflow points (e.g., Mesinger and Veljovic, MAAP 2013), and has for that reason been referred to by McDonald (MWR 2003) as one of “fairly well-posed” schemes.

How to cite: Mesinger, F., Veljovic, K., Chou, S. C., Gomes, J. L., Lyra, A. A., and Jovic, D.: Eta features, additional to the vertical coordinate, deserving attention, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8324, https://doi.org/10.5194/egusphere-egu24-8324, 2024.

Convective-scale ensembles are routinely used in operational centres around the world to produce probabilistic precipitation forecasts, but a lack of spread between members is providing forecasts that are frequently overconfident. This deficiency can be corrected by increasing spread, increasing forecast accuracy or both. A recent development in the Met Office forecasting system is the inclusion of Large-Scale Blending (LSB) in the convective-scale data assimilation scheme. This method aims to reduce the synoptic-scale forecast error in the analysis by reducing the influence of the convective-scale data assimilation at scales that are too large to be constrained by the limited domain. These scales are instead initialised using output from the global data assimilation scheme, which we expect to reduce the forecast error and, thus, improve the spread-skill relationship. In this study, we have quantified the impact of LSB on the spread-skill relationship of hourly precipitation accumulations by comparing forecast ensembles with and without LSB over a 17-day summer trial period. This trial found modest but significant improvements to the spread-skill relationship as calculated using metrics based on the Fractions Skill Score. Skill is improved for a lower precipitation centile by an average of 0.56% at the largest scales, but a corresponding degradation of spread limits the overall correction. The spread-skill disparity is reduced the most in the higher centiles due to a more muted spread response, with significant reductions of up to 0.40% obtained at larger scales. Case study analysis using a novel extension of the Localised Fractions Skill Score demonstrates how spread-skill improvements transfer to smaller-scale features, not just the scales that have been blended. There are promising signs that further spread-skill improvements can be made by implementing LSB more fully within the ensemble.

How to cite: Gainford, A.: Improvements in the spread-skill relationship of precipitation in a convective-scale ensemble through blending, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9034, https://doi.org/10.5194/egusphere-egu24-9034, 2024.

Persistence is a natural first approximation or a baseline to seasonal temperature forecasting.  In the present study, winter and spring persistence in mean montly temperatures in the circumpolar Arctic is explored in long time-series of monthly mean data for the winter and spring seasons.

Locally, very high temporal correlations, as well as significant negative correlations are detected

Physically, the persistence may be traced to snow cover and sea-ice extent.  The variability in these factors may contribute directly to seasonal variability in the radiation budget as well as in surface fluxes, but there are also indirect, but detectable impacts upon regional circulation patterns.

How to cite: Ólafsson, H. and Ekrami, N.: Arctic temperature persistence in winter and spring and seasonal forecasting, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9797, https://doi.org/10.5194/egusphere-egu24-9797, 2024.

In modern forecasting it is now a common technique to use an ensemble of forecasts generated by Numerical Weather Prediction (NWP) models. This necessitates a statistical approach be taken when using these weather predictions to inform decision-making and leveraging probabilities in the production of forecasts. It is often required to take the spread of predictions made by NWPs in the ensemble and reduce these to a single value, a pseudo-deterministic forecast, analogous to a forecast made be a traditional deterministic NWP, in order to allow end users to make binary decisions often defined at a definite threshold. These values may be representative of a single physical parameter modelled (e.g. road surface temperature) or may combine multiple parameters in a physically consistent manner (e.g. the road surface temperature coupled to the depth of water on the road for calculating road state), and are used by stakeholders in a number of sectors often to inform safety critical decision making. Therefore, it is important to ensure that the methodology used to reduce the ensemble of predictions to a pseudo-deterministic forecast is as accurate as possible and can retain information related to the ensemble spread , whilst ensuring consistency in parameters through the spatial and temporal domain.

The Surface Transport Forecast (STF) system produces forecasts for different transport surfaces in response to NWP outputs. The STF system is architected such that it runs simultaneously for each member of the NWP forecast ensemble, producing a corresponding ensemble of STF predictions. This enables the computation of a pseudo-deterministic forecast, which retains the maximum amount of information provided by the NWP ensemble.

To reduce the STF ensemble to a pseudo-deterministic forecast a Kernel Density Estimation (KDE) is utilised to build Probability Density Functions (PDFs), which can be readily interrogated using standard statistical techniques. It is found that pseudo-deterministic forecasts, which are consistent across a combination of physical modelled parameters, can be determined using covariant techniques, ensuring the ensemble is reduced as late as possible in the forecast production keeping the maximum benefit provided by the forecast spread. We will present the numerical and computational implementation of the described method in our STF system. Further, we will analyse the pseudo-deterministic forecasts produced and verify the validity of results at specific locations using multiple years of road observations.

How to cite: Wiggs, J., Eyles, J., and Lake, A.: Creating a Pseudo-Deterministic Forecast for Surface Transport from an NWP Ensemble with Consistency Across Multiple Variables using KDE, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11209, https://doi.org/10.5194/egusphere-egu24-11209, 2024.

The prediction of excessive rainfall using numerical weather prediction (NWP) models is unequivocally difficult owing to the myriad of complexities that must be resolved (e.g., parent storm dynamics, microphysics) in order to forecast the placement and intensity of rainfall correctly. However, machine learning (ML) has provided a new avenue by which we can generate predictions of excessive rainfall with sufficient lead time to inform decision makers and planners to the threat of inclement weather. ML techniques are able to decode known long-standing relationships between environmental predictors and convective hazards from long historical records, and they have demonstrated tremendous value in predicting weather hazards at longer lead times (e.g., Hill et al. 2023). Further, continued effort by the meteorological community to explain ML models and their forecasts is building trust between developers and end users. As a result, their use in meteorological hazard forecasting is expanding, particularly into the medium range (e.g., 4-8 days) when forecasters are reliant on relatively coarse NWP models to create forecasts.

In this work, we are using Random Forests (RFs) to generate daily probabilistic forecasts of excessive rainfall at 1-8 day lead times. The RFs are trained using output from the Global Ensemble Forecast System and historical observations of excessive rainfall. Environmental parameters like precipitable water and CAPE, as well as modeled precipitation, are spatiotemporally arranged so the RFs can learn spatial and diurnal patterns that associate with excessive rainfall. The RF models are evaluated against a spatio-temporally varying climatology and show skill out to 7 days, and routinely outperform human-based forecasts past a 1-day lead time. In this presentation, we will highlight performance characteristics of the RFs into the medium-range (e.g., out to 8 days) and discuss the implications of excessive rainfall definitions in RF model training. Additionally, we will present an ensemble prediction framework that provides estimates of uncertainty and ranges of forecast solutions that operational forecasters desire at extended lead times.

How to cite: Hill, A. and Schumacher, R.: Medium-Range Excessive Rainfall Prediction with Machine Learning, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11708, https://doi.org/10.5194/egusphere-egu24-11708, 2024.

It remains challenging to represent subgrid convection in weather and climate models at horizontal grid resolution across the gray zone, in which convective clouds are only partially resolved by the model dynamics and it is required for the representation of subgrid convection to have a generalized transitional behavior as the model’s horizontal resolution varies.  A practical approach for such a representation is to scale the eddy transport of physical properties from a conventional convection parameterization scheme by a quadradic function of the fractional area covered by convective updrafts in the grid cell (Arakawa and Wu, 2013).  Despite this approach’s popularity, its generalization is limited theoretically by the fact that the coarse-graining statistical analysis that gave rise to the approach involved only an idealized scenario of deep convection in quasi-equilibrium.  Additionally, when applying this approach, there is a theoretical ambiguity associated with the validity of conventional convection parameterizations for a fractional area covered by convective updrafts in the grid cell that is not close to zero.

An alternative approach for subgrid convection representation across the gray zone is to apply a unified plume scheme that treats subgrid convection as nonlocal asymmetric eddies due to unresolved convection relative to the grid-mean vertical flow (Park, 2014).  This unified plume scheme represents unresolved convection relative to the grid-mean vertical motion without relying on quasi-equilibrium assumptions in conventional convection parameterizations.  Its generalized transitional behavior across the gray zone is naturally controlled by the size of the plumes representing unresolved convection that varies with the model’s horizontal resolution.  It simulates all unresolved convective transport of atmospheric properties within a single steady framework, allowing multiple convective plumes.  It also includes the prognosis of unresolved cold pool and convection organization within the planetary boundary layer.  The unified plume scheme circumvents the theoretical limitation and ambiguity of the above approach based on conventional convection parameterization.  It also rectifies the lack of plume memory across the time step in conventional convection parameterizations.

This presentation will focus on an ongoing effort to experiment with the alternative unified approach for representing subgrid convection across the gray zone in NOAA’s Unified Forecast System.  Results from 1-D and 3-D case studies will be shown to highlight the advantage of the unified plume scheme.

References:

Arakawa, A., and C.-M. Wu, 2013: A unified representation of deep moist convection in numerical modeling of the atmosphere. Part I. J. Atmos. Sci., 70, 1977–1992.

Park, S., 2014: A unified convection scheme (UNICON). Part I: Formulation. J. Atmos. Sci., 71, 3902–3930.

How to cite: Bao, J.-W., Michelson, S., Li, H., and Park, S.: A Unified Representation of Subgrid Convection in NOAA’s Unified Forecast System, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11953, https://doi.org/10.5194/egusphere-egu24-11953, 2024.

National Oceanic and Atmospheric Administration’s (NOAA’s) Environmental Modeling Center (EMC) is a lead developer of operational Numerical Weather Prediction (NWP) systems at the National Weather Service (NWS), which are used for the protection of life and property and the enhancement of the economy. EMC transitions to operations and maintains more than 20 numerical prediction systems that are used by NWS, NOAA, other United States (U.S.) federal agencies, and various other stakeholders. These systems are developed through a close collaboration with academic, federal and commercial sector partners. EMC maintains, enhances and transitions-to-operations numerical forecast systems for weather, ocean, climate, land surface and hydrology, hurricanes, and air quality for the U.S. and global domains.

NOAA’s operational predictions are transitioning to the Unified Forecast System (UFS) framework in order to simplify the operational prediction suite of modeling systems. The UFS is being designed as a community-based, comprehensive atmosphere-ocean-sea-ice-wave-aerosol-land coupled Earth modeling system with coupled data assimilation and ensemble capabilities, organized around applications spanning from local to global domains and predictive time scales ranging from sub-hourly analyses to seasonal predictions.  Disparate legacy operational applications that have been developed and maintained by EMC in support of various stakeholder requirements are being transitioned to the UFS framework. The transition started several years ago and is planned to continue over the next few years. Fewer resulting applications will consolidate NCEP’s Production Suite that shares a set of common scientific components and technical infrastructure.  This streamlined suite is expected to accelerate the transition of research into operations and simplify maintenance of operational systems.

This talk describes major development and operational implementation projects at EMC over the last couple of years including for example a new UFS-based hurricane application, recent advances in the use of satellite data and a new verification system. We will present EMC plans for the next few years, within the overall NOAA strategy, and how planned efforts link with other modeling efforts within NOAA, in the broader U.S. and international community.

How to cite: Stajner, I., Gross, B., Tallapragada, V., Levit, J., Montuoro, R., Mehra, A., Kleist, D., and Yang, F.: NOAA’s Environmental Modeling Center Update: Transitioning to Unified Forecast System Applications for Operations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12029, https://doi.org/10.5194/egusphere-egu24-12029, 2024.

A new land surface package developed at Environment and Climate Change Canada (ECCC) has been evaluated in the context of the medium-range global deterministic numerical weather prediction (NWP) system. The evaluation is performed by comparison of NWP forecasts against near-surface and

atmospheric analyses. The new land surface package includes i) new databases to specify soils and vegetation characteristics, ii) improved initialization of land surface variables by the assimilation of space-based remote sensing observations, and iii) a more sophisticated land surface scheme.

Evaluation for the screen-level air temperature and humidity indicates that the new land surface package resulted in smaller STDEs and larger temporal correlation between forecasts and analyses comparing to the current operational configuration. The improvement is greater for humidity than for air temperature.

Upper-air evaluation indicates that the impact of the new land surface package on the Planetary boundary layer (PBL) is substantial but more mixed, with large spatial variability in terms of its effect.

This study also investigated the physical and statistical links between near-surface and upper-air forecast errors at the medium range.

How to cite: Alavi, N., Belair, S., Carrera, M., Abrahamowicz, M., Bilodeau, B., Simjanovski, D., Charpentier, D., Badawy, B., and Leroyer, S.: Impact of a new land surface package in Canadian  numerical weather prediction system on the medium range weather forecast in the lower and upper atmosphere, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12345, https://doi.org/10.5194/egusphere-egu24-12345, 2024.

The lower tropospheric enhanced gravity wave drag (GWD) parameterization has been operational in Global Forecast System (GFS) since late 1990s. The scheme is based on Kim and Arakawa and further revised with the addition of flow blocking (Kim and Doyle). For UFSR2O project, there have been collaborative efforts to improve the GWD parameterization by revising the mountain induced GWD. Revisions include the updates in GWD and flow blocking (Choi and Hong), and turbulent orography form drag of Beljaars et al. Sensitivity experiments are performed to investigate the importance of partitioning GWD and flow blocking in the skill of medium-range forecasts. Alternative approach for TOFD (Richter et al.) is tested. Importance of the representation of sub-grid orography statistics is also examined. 

How to cite: Hong, S., Bao, J.-W., Michelson, S., Grell, E., Toy, M., Olson, J., and Yang, F.: Sensitivity Experiments of a Mountain-Induced Gravity Wave Drag Parameterizations for Global Weather Forecasting, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12794, https://doi.org/10.5194/egusphere-egu24-12794, 2024.

The Geophysical Fluid Dynamics Laboratory (GFDL)’s System for High‐resolution prediction on Earth‐to‐Local Domains (SHiELD) model typically uses the National Centers for Environmental Prediction (NCEP) Global Forecast System (GFS) analyses to initialize its medium-range global forecasts. Both initial condition (IC) and forecast model have an impact on model prediction skills. The quality of the IC is partially determined by the model short-range forecast used as first guess in data assimilation. 

A data assimilation (DA) system has been developed for the global SHiELD to demonstrate the prediction skills of the model initialized from its own analysis. The DA system largely leverages the advanced DA techniques used in GFS and assimilates all the observations assimilated in GFS. Compared to the SHiELD forecasts initialized from GFS analysis, SHiELD forecast skill is significantly improved by using its own analysis. Tremendous improvement was found in the Southern Hemisphere with positive impact lasting up to 10 days. The DA system is also useful in identifying and understanding model errors. The most noticeable model error detected by the DA system originates from the TKE-EDMF boundary layer scheme. The model error leads to insufficient ensemble spread, which could not be fully addressed by the multiplicative inflation and stochastic physics schemes used in the system. Including two versions of the TKE EDMF scheme in the ensemble can alleviate the systematic model error, which further improves forecast skills. The use of the interchannel correlated observation errors for Infrared Atmospheric Sounding Interferometer (IASI) and Cross-track Infrared Sounder (CrIS) was also investigated, which improves the forecast skill up to day 5 and further reduces the impact of the model error in the marine stratocumulus region. Further understanding of the model error associated with the TKE-EDMF scheme will be presented. 

How to cite: Tong, M., Harris, L., Zhou, L., Gao, K., Kaltenbaugh, A., and Xiang, B.: Improved Weather Predictions Through Data Assimilation for GFDL SHiELD, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13187, https://doi.org/10.5194/egusphere-egu24-13187, 2024.

This presentation provides recent research highlights of the project PREVENIR, including radar quantitative precipitation estimates (QPE), ensemble nowcasting, data assimilation, numerical weather prediction (NWP), and hydrological model prediction. PREVENIR is an international cooperation project between Argentina and Japan since 2022 for five years under the Science and Technology Research Partnership for Sustainable Development (SATREPS) program jointly funded by the Japan International Cooperation Agency (JICA) and the Japan Science and Technology Agency (JST). The main goal is to develop an impact-based early warning system for heavy rains and urban floods in Argentina. PREVENIR takes advantage of leading research on Big Data Assimilation (BDA) with the Japan’s flagship supercomputer “Fugaku” and its predecessor “K” and develops a total package for disaster prevention, namely, monitoring, QPE, nowcasting, BDA and NWP, hydrological model prediction, warning communications, public education, and capacity building. The total package for disaster prevention will be the first of its kind in Argentina and will provide useful tools and recommendations for the implementation of similar systems in other parts of the world.

How to cite: Miyoshi, T., Skabar, Y. G., Otsuka, S., Amemiya, A., Ruiz, J., Ushio, T., Tomita, H., Ushiyama, T., and Konishi, M.: Second Year Progress of PREVENIR: Japan-Argentina Cooperation Project for Heavy Rain and Urban Flood Disaster Prevention, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13504, https://doi.org/10.5194/egusphere-egu24-13504, 2024.

Cyclones provides the majority of water supplies in the Mediterranean and are essential elements of the climate of the region. The most intense of them lead to natural disasters because of their violent winds and extreme rainfall. Identifying systematic errors in the predictability of Mediterranean cyclones is therefore essential to better anticipate and prevent their impact. The aim of this work is to understand what processes determine their predictability. 

We investigate the predictability of Mediterranean cyclones in a systematic framework using an ensemble prediction system. First, a reference dataset of 2853 cyclones is obtained by tracking lows in the ERA5 reanalysis, using an algorithm developed for the North Atlantic and adapted for the Mediterranean region. Then we investigate their predictability using IFS ensemble reforecasts in a homogeneous configuration over 22 years (2000-2021). The predictability in the reforecasts is quantified using probabilistic scores on cyclones trajectories and on intensity (mean sea level pressure) and then crossed with explanatory variables such as geographic area, cyclone velocity, season and intensity.

The evolution of location error with lead time shows a two phases growth, until and beyond 72 h, which will be discussed. When crossing the location and intensity errors with the explanatory variables, we can identify the conditions leading to a poorer (respectively better) predictability. In particular the velocity of cyclones appears to play an important role in the predictability of the location, the slower the cyclone the better the predictability, while the season is shown to play a greater role on the predictability of the intensity. These characteristics are also dependant on the sub-region considered and on the intensity of the low itself, the deeper the cyclone, the poorer the predictability in both the location and the intensity.

How to cite: Doiteau, B., Pantillon, F., Plu, M., Descamps, L., and Rieutord, T.: What determines the predictability of a Mediterranean cyclone?  , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13557, https://doi.org/10.5194/egusphere-egu24-13557, 2024.

How to cite: Nakazawa, T., Miyoshi, T., Sakajo, T., and Takatama, K.: An overview of Japan’s Moonshot Goal 8 R&D program for controlling and modifying the weather by 2050, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13655, https://doi.org/10.5194/egusphere-egu24-13655, 2024.

Cloud microphysics parameterizations are generally divided into two categories: bin models that explicitly calculate the evolution of the drop size distribution (DSD) and bulk models that represent the DSD with a specific function. The Weather Research and Forecasting (WRF) Double-Moment 6-class (WDM6) scheme is one of the bulk microphysics options in the WRF model and is widely utilized for both research and operational purposes. In WDM6 scheme, the gamma form with a single static shape parameter is applied for the DSD of rain. This study adopts a generalized double-moment normalization method for the rain DSD in WDM6 scheme. Previous study mentions that the advantage of the generalized double-moment normalization method lies in its ability to singnificantly reduce the observed DSD scatter. Therefore, it can concisely represent the DSD with appropriate shape parameters, c and μ. The modified WDM6 is evaluated through simulations of an idealized 2D squall line and a summer precipitation case over the Korean peninsula. Based on similar experimental results from the original WDM6 and the modified WDM6 schemes, we can confirm that the generalized double-moment normalization method in the WDM6 scheme is properly implemented. We further collected the observed shape parameters suitable for the generalized double-moment DSD of rain over a two-year summer period (2018, 2019). The modified WDM6, with the observed shape parameters, simulates a more comparable spatial distribution of acummulated precipitation that occurred on 6 August 2013 with the observation, compared to the original WDM6. More detailed simulation results will be presented at the conference.

* This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT). (grant no.RS-2023-00208394).

How to cite: Jo, J., Park, S.-Y., Lim, K.-S. S., Bang, W., and Lee, G.: Implementation of the Generalized Double-Moment Normalization Method in the Cloud Microphysics Scheme, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13807, https://doi.org/10.5194/egusphere-egu24-13807, 2024.

Planetary boundary layer (PBL) parameterizations using the eddy diffusivity - mass flux (EDMF) technique for turbulent mixing in the convective PBL have been popularly used in weather and climate models.  When including moist adjustment processes, some numerical implementations of the EDMF parameterization may result in unphysical solutions of cloud condensate, for example, negative cloud water quantities.  To solve this problem, a procedure to obtain a positive definite solution is proposed to solve the moist EDMF equations.  In this presentation, we will demonstrate the formulation of the solution procedure and show examples of its impact on the PBL mixing simulation using a single-column model.

How to cite: Grell, E. and Bao, J.-W.: A Positive-Definite Moist EDMF Parameterization Scheme for Turbulent Mixing in the PBL, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13978, https://doi.org/10.5194/egusphere-egu24-13978, 2024.

Korea Institute of Atmospheric Prediction Systems (KIAPS) has developed a global forecasting system, Korean Integrated Model (KIM) and the model now operates with 12-km horizontal resolution. With plans to develop the numerical model in horizontally and vertically higher resolution, smoothed hybrid sigma-pressure (SMH) coordinate has applied to KIM to cover the influence of the terrain structure. The SMH is proposed to alleviate artificial circulations that horizontal pressure gradients and advection can be appeared along complex surfaces by reducing small-scale components more rapidly with height (Choi and Klemp, 2021). 
In this research, we focus on the prediction with higher-resolution topography in the SMH coordinate and it is revealed that more realistic data can be utilized than the previous topography adapted in hybrid sigma coordinate. The SMH coordinate could well reflect the steepness and roughness of complex region such as terrains near mountains without stability issue. To investigate the sensitivity to the detailed topographic data, case studies such as heatwave, cold surge and rainfall are dealt with especially in the Korean peninsula consisted of complex terrain. By considering more complex topography, the SMH coordinate performs better in capturing precipitation peak and temperature bias. In addition, it will be discussed that vertical propagation to the upper atmosphere is appropriately controlled due to the SMH coordinate. This study can contribute to the future work on adjusting diffusion coefficient by optimizing the SMH coordinate in much higher resolution.

How to cite: Kong, H.-J., Park, J.-R., and Nam, H.: Response of the SMoothed Hybrid sigma-pressure (SMH) coordinate to higher-resolution topographic data in Korean Integrated Model (KIM), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14761, https://doi.org/10.5194/egusphere-egu24-14761, 2024.

The Korean Integrated Model (KIM) is a global numerical weather prediction (NWP) system developed by the first phase project of the Korea Institute of Atmospheric Prediction Systems (KIAPS) and has been used as the operational NWP system at the Korea Meteorological Administration (KMA) since April 2020. The second phase project of KIAPS aims at developing a next-generation NWP system to seamlessly predict from very short-range to extended medium-range. To improve the extended medium-range forecast, one of the main goals of KIAPS is to develop the ensemble prediction system with coupling to land, ocean, and sea ice. The production of extended medium-range reforecast data is necessary to understand the climatological characteristics and model biases of KIM. KIAPS developed an initial version of reforecasting system based on the KIM atmopheric model. The system has a spatial resolution of 50 km (NE090NP3) and consists of 91 vertical layers. We produce reforecast of the cold season cases for 20 years (from 2001 to 2020) and perform the diagnosis and verification of reforecast data. A suite of sensitivity experiments are also performed to investigate the impact of initial perturbations on the ensemble prediction system.

How to cite: Kim, S.-W., Shim, T., Hong, J.-Y., and Park, H.-J.: Development of extended medium-range reforecasting system based on the Korean Integrated Model (KIM), EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15364, https://doi.org/10.5194/egusphere-egu24-15364, 2024.

Weather Research and Forecasting (WRF) plays a crucial role in studying atmospheric dynamics and investigating the mesoscale weather prediction phenomena. However, WRF model offers lots of different configurations for physics, dynamics, and domain options that need to be investigated. From these configurations, domain options offer nesting techniques which may affect the fundamental structure and the performance of the simulations. Nesting options may impact the representation of fine-scale processes by increasing the resolution for the desired domain, compared to single-domain simulations. Existing studies on comparison of different nesting configurations in mesoscale domains are limited. This study presents a comparative analysis of three different nesting configurations in the WRF model over Türkiye. Accuracy of WRF-based short-term (24 to 48 hourly) temperature, wind, and precipitation forecasts over a 30-day period in November 2021 is investigated utilizing ground-station based observations. Three different model configurations are investigated: single domain, one-way feedback nested, and two-way feedback nested runs for the same time period and region. Root mean square error (RMSE), error standard deviation, and correlation coefficient were calculated for all three configurations. This study contributes to the optimization of nesting configurations in WRF mesoscale weather predictions, aiding decision-making processes reliant on accurate short-term forecasts in Türkiye.

How to cite: Çatal, A. C., Arık, A., Yılmaz, M. T., and Yücel, İ.: Impact of Nesting Techniques Over Short-Term WRF Forecast Accuracy, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15467, https://doi.org/10.5194/egusphere-egu24-15467, 2024.

Machine Learning (ML) became pervasive in every domain of the research, providing opportunities of modeling phenomena that were difficult to capture using known equations. From small models running on student computers, to giant LLMs trained on the whole Internet, ML models come in all shapes and sizes. To the meteorological community, one branch of this research stands out as revolutionary - ML-based global weather models.

ML-based global weather models lie on the opposite end of the spectrum compared to numerical weather prediction (NWP) models. Instead of representing the physics in a form of equations and solving these equations on the model grid, ML models are purely data-driven - even if they managed to represent physics internally, the inference of that physics would remain a black box.

Yet, these models underwent significant advancement in the past year - and three of them stand out - GraphCast (Google), ClimaX (Microsoft) and MetNet (Google). The former two, open-sourced for research purposes, are being tested currently at Belgingur. Having many years of experience with running and deploying NWP weather models, we notice how working with these models differs from working with the new class of ML-based (or data-driven) models.

This talk discusses essential differences between working with NWP and ML-based weather models. What we can, and what we cannot control? What does the process of working with such an ML model look like? What is the main advantage of an ML model run in production? What are the main obstacles in deploying an ML model and running it operationally?

With the current pace of the growth of Machine Learning models, we will be encountering them in our everyday work sooner or later. Knowing the challenges and opportunities of them will help us understand how to use them to our advantage.

How to cite: Stanisławska, K. and Rognvaldsson, O.: Running global Machine Learning weather models - challenges, observations and conclusions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15614, https://doi.org/10.5194/egusphere-egu24-15614, 2024.

Belgingur Ltd. has created a novel weather forecasting framework, called Weather On Demand – WOD, that is deployable in the cloud and on in-house hardware and which can be customised for any location world-wide at a very short notice.

The WOD framework is a distributed system for:

• Running the WRF weather model for data-assimilation and forecasts by either triggering scheduled or on-demand jobs.
• Gathering upstream weather forecasts and observations from a wide variety of sources.
• Processing data for long to medium-term storage.
• Making results available through APIs.
• Making data files available to custom post-processors.

Much effort is put into starting processing as soon as the required data becomes available and in parallel when possible.

Recent additions to the WOD system include the potential of:

• Optional use of the hybrid data assimilation techniques of the WRF Data Assimilation system [1, 2].
• Set up a multi-domain dispersion forecast of volcanic ash and gases.
• Use of the Verif [3] verification package to compare forecasts, both upstream and WOD, to observations.
• Using different sources of initial data to that of the boundary forcing data.

On-going developments focuses on the use of in-situ UAV profiles and radar data as input to the WOD data assimilation system.

We have further started experimenting with using global models, both conventional NWP models as well as novel ML models (cf. abstract no. EGU24-15614).

References:

[1] Xuguang Wang, Dale M. Barker, Chris Snyder, and Thomas M. Hamill, 2008: A hybrid ETKF–3DVAR data assimilation scheme for the WRF model. Part I: Observing system simulation experiment. Mon. Wea. Rev., 136, 5116–5131.

[2] Xuguang Wang, Dale M. Barker, Chris Snyder, and Thomas M. Hamill, 2008: A Hybrid ETKF–3DVAR Data Assimilation Scheme for the WRF Model. Part II: Real Observation Experiments. Mon. Wea. Rev., 136, 5132–5147.

[3] Nipen, T. N., R. B. Stull, C. Lussana, and I. A. Seierstad, 2023. Verif: A Weather-Prediction Verification Tool for Effective Product Development. Bulletin of the American Meteorological Society 104, 9; 10.1175/BAMS-D-22-0253.1.

How to cite: Rognvaldsson, O. and Stanislawska, K.: The Weather On Demand weather forecast framework - Recent developments and outlook, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16714, https://doi.org/10.5194/egusphere-egu24-16714, 2024.

Atmospheric predictability is intrinsically limited by the upscale growth of initial small-scale, small-amplitude errors. For practical predictability, model error and initial-condition uncertainty also contribute significantly. The accurate representation and interactions of these factors within numerical weather prediction systems determine the extent to which forecast uncertainty is correctly modeled. An improved understanding of upscale error-growth mechanisms and their flow dependence in numerical weather prediction models has several implications: it enables more focused model verification and development, aids in recognizing limitations in emerging forecasts systems like machine-learning-based approaches, and may indicate when the intrinsic limit of predictability has been reached.

Studying the flow dependence of error growth requires a local perspective, which is not provided by the traditional spectral perspective on upscale error growth. We here take a complementary approach and apply a feature-based perspective. We have developed an automated algorithm to identify error features in gridded data and track their spatial and temporal evolution. Errors are considered in terms of potential vorticity (PV) and near the tropopause, where they maximize. A previously derived PV-error tendency equation is evaluated to quantify the different contributions to error-growth experiments with the global prediction Model ICON from the German Weather Service. Errors in these experiments grow from differences in the seeding of a stochastic convection scheme. In a suite of experiments, this source of uncertainty competes with initial-condition uncertainty of varying magnitude. Evaluation of the process-specific error-growth rates allow the detailed quantification of the upscale-growth mechanisms. For this purpose, we integrate the growth rates over the respective area associated with an error feature. Examination of the combined growth rates of all features in an upscale-error-growth experiment reproduces a previously found three-stage multi-scale upscale-growth paradigm. Illustration the importance of flow dependence, the growth rates from a single feature can substantially differ from the overall average. Further highlighting this importance, intrinsic limits of predictivity can be identified for some features even in the presence of substantial initial-condition uncertainty. The presentation will conclude with a comparison of error evolution in conventional numerical weather prediction systems to a data-driven, machine-learned model.

How to cite: Schmidt, S., Riemer, M., and Selz, T.: A feature-based framework to investigate atmospheric predictability., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17438, https://doi.org/10.5194/egusphere-egu24-17438, 2024.

The CERISE (CopERnIcus climate change Service Evolution) project aims to develop land and coupled land-atmosphere data assimilation systems for the next generation of coupled reanalysis. This encompasses technical enhancements to the system architecture as well as scientific changes to improve the quality of the reanalyses.

Recent work has focussed on developing ensemble perturbation methods for the land-surface. The existing ensemble spread in model variables at and near the land-surface is known to be insufficient which can cause problems when assimilating interface observations in a coupled system. This is because the existing ensemble perturbations are mainly applied to upper air atmospheric variables. One way to increase the spread at the surface is to directly perturb land-surface parameters such as vegetation cover and leaf area index. Results from this approach are encouraging in offline and coupled experiments.

Another part of the project is to enhance the assimilation of passive microwave radiances over land. Currently the use of surface-sensitive passive microwave channels are largely limited to the ocean due to challenges in forward modelling of complex and heterogenous land surfaces. In CERISE, machine learning approaches are being explored to develop an observation operator to enable the use of these observations over land and snow surfaces.

Finally, developing quasi-strongly coupled land-atmosphere assimilation is a key objective of the project. Developments so far have focussed on technical changes to build a framework to allow stronger coupling than the current weakly coupled assimilation strategy. A summary of recent progress in the CERISE project will be presented.

How to cite: Weston, P., de Rosnay, P., Herbert, C., and Pinnington, E.: Enhanced coupled land-atmosphere data assimilation for reanalysis, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18054, https://doi.org/10.5194/egusphere-egu24-18054, 2024.

Skillfully forecasting the evolution of tropical cyclones (TC) is crucial for
the human populations in areas at risk, and an essential indicator of a storm’s
potential impact is the Maximum Sustained Wind Speed, often referred to as
the cyclone’s intensity. Predicting the future intensity of ongoing storms is
traditionally done using statistical-dynamical methods such as (D)SHIPS and
LGEM, or as a byproduct of fully dynamical models such as the HWRF model.
Previous works have shown that deep learning models based on convolutional
neural networks can achieve comparable performances using infrared and/or
passive microwave satellite imagery as input. Recently, multi-task models have
highlighted that jointly learning the future intensity and other indicators such
as the TC size with shared network weights can improve the performance in the
context of intensity estimation. This ongoing work aims to evaluate which tasks
and architectures can lead to the best improvement for intensity forecasting.

How to cite: Dauvilliers, C., Charantonis, A., and Monteleoni, C.: Evaluating multi-task learning strategies for tropical cyclones itnensity forecasting from satellite images, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18151, https://doi.org/10.5194/egusphere-egu24-18151, 2024.

The Korean Integrated Model (KIM) has been in operation at Korea Meteorological Administration (KMA) since April 2020 and its forecasting performance has been improved by updating model physical processes and data assimilation system. The model performance is comparable to that of the Unified Model run in parallel with the KIM at KMA during Boreal summer, but is relatively poor during the winter. One of the major biases in 5-day temperature forecasts for Norther Hemisphere winter is the low atmospheric cold bias over the Arctic region, and thus this study modifies the initialization of sea ice properties (sea ice thickness and temperature) to reduce the bias. First, the initial sea ice thickness data prescribed by climatology data produced using reanalysis data from the past 10 years (2000~2009) is replaced using the latest (2019~2021) reanalysis data. Second, the initial temperatures of the 1st and 2nd sea ice layers are set to the sea water freezing temperature instead of the currently applied first guess (background) sea ice temperatures. The effects of initialization modification on the medium-range forecasts of KIM are analyzed by performing two sets of experiments: cold start and warm cycle experiments without and with a data assimilation system in January 2022. The latest sea ice thickness initial data shows that sea ice thickness has decreased by about a factor of two. And its adoption by KIM increases surface and lower atmospheric temperatures in the Arctic sea ice region, alleviating cold biases in the region for both analysis and forecasts. In addition to sea ice thickness, sea ice temperature initialization modifications enhance Arctic warming and lead to greater improvement of cold bias. The warming effect in the lower Arctic is consistent in both cold start and warm cycle experiments. However, secondary effects induced by the Arctic warming occur significantly only in the warm cycle experiment and significantly affect forecasts fields not only in the polar region but also in the Southern Hemisphere and mid-latitude regions. Skill scores for medium-range forecasts in January 2022 are mostly improved (degraded) for the 12 UTC (00 UTC) initial conditions in the warm cycle experiment.

How to cite: Choi, H.-J., Kim, S. H., Kim, B.-M., Koo, M.-S., Cho, E.-H., and Kwon, Y. C.: Effects of initialization of sea ice properties on medium-range forecasts in the Korean Integrated Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18966, https://doi.org/10.5194/egusphere-egu24-18966, 2024.

How to cite: Wang, C., Žagar, N., and Vasylkevych, S.: Benefits of initializing equatorial waves on extratropical forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18982, https://doi.org/10.5194/egusphere-egu24-18982, 2024.

The “Year Without Summer” of 1816 was characterized by extraordinarily cold and wet periods in Central Europe, and it was associated with severe crop failures, famine, and socio-economic disruptions. From a modern perspective and beyond its tragic consequences, the summer of 1816 represents a rare occasion to analyze the adverse weather (and its impacts) after a major volcanic eruption. However, given the distant past, obtaining the high-resolution data needed for such studies is a challenge. In our approach, we use dynamical downscaling, in combination with 3D-variational data assimilation of early instrumental observations, for assessing a cold-air outbreak in early June 1816. 
Our downscaling simulations reproduce and explain meteorological processes well at regional to local scales, such as a foehn wind situation over the Alps with much lower temperatures on its northern side. Simulated weather variables, such as cloud cover or rainy days, are simulated in good agreement with (eye) observations and (independent) measurements, with small differences between the simulations with and without data assimilation. However, validations with partly independent station data show that simulations with assimilated pressure and temperature measurements are closer to the observations. In turn, data assimilation requires careful selection, preprocessing and bias-adjustment of the underlying observations. Our findings underline the great value of digitizing efforts of early instrumental data and provide novel opportunities to learn from extreme weather and climate events as far back as 200 years or more.

How to cite: Stucki, P., Pfister, L., Brönnimann, S., Brugnara, Y., Hari, C., and Varga, R.: Dynamical downscaling and data assimilation for a cold-air outbreak in the European Alps during the Year Without Summer 1816, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19593, https://doi.org/10.5194/egusphere-egu24-19593, 2024.

Irrigation is one agricultural practice that contributes to maintain an optimal soil water content for crop development. Currently, farmers find this practice as an essential method for adapting to climate change. The Earth science community identified some irrigation effects beyond soil moisture and plant growth impact, as multiple studies found an influence on atmospheric variables such as 2 m temperature, relative humidity and even precipitation. Moreover, the effect of irrigation on the Earth’s system has been studied on various temporal and geographical scales and with different climate and land surface models. However, there are few studies that simulated the effect of irrigation on higher resolutions on a regional scale. Therefore, the aim of this study is to include the representation of irrigation processes in the operational ICON-nwp in Limited Area Mode on the EURO-CORDEX domain. The implementation of the current irrigation parameterization in ICON-nwp coupled with TERRA is an adaptation of the CHANNEL scheme developed by Valmassoi et al. (2020). We found suitable to include this scheme in the land surface and atmosphere interface of the icon-nwp-2.6.6-nwp0 version. The present study consists of four sensitivity experiments with different irrigation water amounts, namely 2.6 mmd-1, 6.7 mmd-1 and two fixed soil water contents, field capacity and saturation. All experiments have the same irrigation frequency (1 day), length (24 hours), and simulation period (May to August). The model settings for the experiments are 3 km resolution, 75 vertical levels and ICON boundary and initial conditions. The results from the difference between experiments and the control run demonstrate that ICON captures the irrigation effect on land surface atmospheric variables. As expected, soil moisture content increased on different magnitudes in all experiments. Moreover, 2 m temperature values dropped on average -0.74 K in irrigated areas. Likewise, energy fluxes were sensible to the different irrigation amounts.

How to cite: Roque, J., Valmassoi, A., and Keller, J.: Irrigation parameterization in the Operational Numerical Weather Prediction model ICON-nwp, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20366, https://doi.org/10.5194/egusphere-egu24-20366, 2024.

In 2016, the U. S. Congress created the Joint Technology Transfer Initiative (JTTI) program in the Office of Oceanic and Atmospheric Research (OAR), the research wing of the National Oceanic Atmospheric Administration (NOAA).  Within OAR, the Weather Program Office (WPO) is responsible for managing the JTTI program.  The main mission of this program is to continuously develop and transition the mature weather technologies from the research community to the National Weather Service (NWS) operations.  

JTTI selects promising Research to Operations (R2O) transition projects through two types of competitions: one for the external community (non-NOAA) that includes private, academic sectors and non-profit organizations through Notices of Funding Opportunities; and the other for the NOAA scientific community.  Additionally, the JTTI program collaborates with the NWS Office of Science and Technology Integration (OSTI) and provides funding for the Unified Forecasting System - R2O project and testbed activities.  JTTI-funded R2O projects cover three main frameworks within the NWS forecasting operations: the observational, modeling, and products and services frameworks.  The topics covered include data assimilation; convective scale weather modeling; stochastic physics; ensemble model building; hydrologic modeling; post-processing of model output on time scales ranging from hourly to subseasonal; high impact weather forecasting tools; artificial intelligence/machine learning; and social behavioral and economic science. To date, the JTTI program has funded 155 R2O projects and transitioned 20 projects to the NWS operations.  In this paper, we present the JTTI implementation process in NOAA and share some of the successful R2O stories.

How to cite: Kondragunta, C., Pratt, A., Garrett, K., Kurkowski, N., Sellers, W., and Kunkel, V.: Crossing the Valley of Death : Transitioning Weather Research to Operations in NOAA, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20553, https://doi.org/10.5194/egusphere-egu24-20553, 2024.

Weather radars commonly suffer from the data-missing problem that limits their data quality and applications. Traditional methods for the completion of weather radar missing data, which are based on radar physics and statistics, have shown defects in various aspects. Several deep learning (DL) models have been designed and applied to weather radar completion tasks but have been limited by low accuracy. This study proposes a dilated and self-attentional UNet (DSA-UNet) model to improve the completion of weather radar missing data. The model is trained and evaluated on a radar dataset built with random sector masking from the Yizhuang radar observations during the warm seasons from 2017 to 2019, which is further analyzed with two cases from the dataset. The performance of the DSA-UNet model is compared to two traditional statistical methods and a DL model. The evaluation methods consist of three quantitative metrics and three diagrams. The results show that the DL models can produce less biased and more accurate radar reflectivity values for data-missing areas than traditional statistical methods. Compared to the other DL model, the DSA-UNet model can not only produce a completion closer to the observation, especially for extreme values, but also improve the detection and reconstruction of local-scale radar echo patterns. Our study provides an effective solution for improving the completion of weather radar missing data, which is indispensable in radar quantitative applications.

How to cite: Gong, A., Chen, H., and Ni, G.: Improving the Completion of Weather Radar Missing Data with Deep Learning, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3548, https://doi.org/10.5194/egusphere-egu24-3548, 2024.

How to cite: Feldmann, M., Poulain-Auzeau, L., Gomez, M., Beucler, T., and Martius, O.: Convective environments in AI-models - What have AI-models learned about atmospheric profiles?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5373, https://doi.org/10.5194/egusphere-egu24-5373, 2024.

Surface solar irradiance (SSI) is a pivotal component in addressing climate change. As an abundant and non-fossil energy source, it is harnessed through photovoltaic (PV) energy production. As the contribution of PV to total energy production grows, the stability of the power grid faces challenges due to the volatile nature of solar energy, predominantly influenced by stochastic cloud dynamics. To address this challenge, there is a need for accurate, uncertainty-aware, near real-time, and regional-scale SSI forecasts with forecast horizons ranging from minutes to a few hours.

Existing state-of-the-art SSI nowcasting methods only partially meet these requirements. In our study, we introduce SHADECast [1], a deep generative diffusion model designed for probabilistic nowcasting of cloudiness fields. SHADECast is uniquely structured, incorporating deterministic aspects of cloud evolution to guide the probabilistic ensemble forecast, relying only on previous satellite images. Our model showcases significant advancements in forecast quality, reliability, and accuracy across various weather scenarios.

Through comprehensive evaluations, SHADECast demonstrates superior performance, surpassing the state of the art by 15% in the continuous ranked probability score (CRPS) over diverse regions up to 512 km × 512 km, extending the state-of-the-art forecast horizon by 30 minutes. The conditioning of ensemble generation on deterministic forecasts further enhances reliability and performance by more than 7% on CRPS.

SHADECast forecasts equip grid operators and energy traders with essential insights for informed decision-making, thereby guaranteeing grid stability and facilitating the smooth integration of regionally distributed PV energy sources. Our research contributes to the advancement of sustainable energy practices and underscores the significance of accurate probabilistic nowcasting for effective solar power grid management.

References

[1] Carpentieri A. et al., 2023, Extending intraday solar forecast horizons with deep generative models. Preprint at ArXiv. https://arxiv.org/abs/2312.11966 

How to cite: Carpentieri, A., Folini, D., Leinonen, J., and Meyer, A.: SHADECast: Enhancing solar energy integration through probabilistic regional forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5571, https://doi.org/10.5194/egusphere-egu24-5571, 2024.

Flood early warning in fast responding catchments challenges our forecasting systems. It requires frequently updated, accurate and high-resolution rainfall forecasts to provide timely warning of rainfall amounts that will reach a catchment in the coming hours. The Netherlands is a typical example, with polder systems below sea level, a high level of urbanization and catchments with short response times. The need for better short-term rainfall forecasts is clearly present, but this is generally not feasible with numerical weather prediction (NWP) models alone. Hence, an alternative rainfall forecasting method is desirable for the first few hours into the future.

Rainfall nowcasting can provide this alternative but quickly loses skill after the first few hours. A promising way forward is a seamless forecasting system, which tries to optimally combine rainfall products from nowcasting and NWP. In this study, we applied the STEPS blending method to combine rainfall forecasts from ensemble radar nowcasts with those from the Harmonie-AROME configuration of the ACCORD NWP model in the Netherlands. This blending method is part of the open-source nowcasting initiative pysteps. To make blending possible in an operational setup, including the needs of involved water authorities, we made several adjustments to the blending implementation in pysteps, for instance:

• We reduced the computational time by using a faster preprocessing and advection scheme.
• We improved the noise initialization (needed for generating ensemble members) to allow for stable forecasts, also when one or both product(s) contain(s) no rain.
• We enabled a dynamic disaggregation of the 1-hour resolution NWP forecasts to match the temporal resolution of the radar nowcast.

We operationalized the updated blending framework in the flood forecasting platforms of the involved water authorities. Given a forecast duration of 12 hours for the blended forecast and a 10-minute time step, average computation times are 3.4 minutes for a deterministic run and 12.3 minutes for an ensemble forecast with 10 members on a 4-core machine. Preprocessing takes approximately 10 minutes and only needs to occur when a new NWP forecast is issued. We tested the implementation for an entire, rainy summer month (July 15 to August 15, 2023) and analyzed the results over the entire domain. The results demonstrate that the blending method effectively combines radar nowcasts with NWP forecasts. Depending on the statistical score considered (such as RMSE and critical success index), the blending method performs either better or on par with the best-performing individual product (radar nowcast or NWP). A consistent finding is that the blending closely tracks the nowcast quality during the initial 1 to 2 hours of the forecast (in this study, the nowcast had lower errors than NWP during the first 2 – 2.5 hours), after which it gradually transitions into the NWP forecast. At longer lead times, the seamless product retains local precipitation structures and extremes better than the NWP product. It does this by leveraging information from the radar nowcast and the stochastic perturbations. Based on these results, a seamless forecasting approach can be regarded as an improvement for the involved water authorities.

How to cite: Imhoff, R., Van Ginderachter, M., van Heeringen, K.-J., Radema, M., De Kock, S., Reinoso-Rondinel, R., and De Cruz, L.: Towards seamless rainfall and flood forecasting in the Netherlands: improvements to and validation of blending in pysteps, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5849, https://doi.org/10.5194/egusphere-egu24-5849, 2024.

Implementing the 3-Dimensional Variational (3DVar) data assimilation technique using high-density automatic weather station (AWS) observations substantially improves the precipitation simulation and forecast capabilities in the Weather Research and Forecasting (WRF) model. Given the impact of spatial distribution and quantity of observation data on assimilation effectiveness, there is a growing need to assimilate the most efficient amount of observation data to improve the precipitation forecast accuracy, especially in the context of the proliferation of data from diverse sources. This study investigates the impacts of spatial density of assimilated data on enhancing model predictions, focusing on a squall line event in Beijing on 2 August 2017 which has approximately 2400 AWSs in the simulation domain. Seven experiment groups assimilating varying proportions of AWS data (3.125, 6.25, 12.5, 25, 50, 75, and 100 percent of total AWSs) were conducted, comprising 10 experiments per group. The results were then compared with the experiment without data assimilation (CTRL) and the observations. Results show that while the WRF model roughly captured the evolution of this event, it overestimated the precipitation amount with significant deviations in precipitation locations. A general positive correlation was observed between the spatial density of assimilated data and the enhancement in model performance. However, there is a notable threshold beyond which additional data ceases to enhance forecast accuracy. The model performs best when the ratio of the number of assimilated AWSs to the model simulated area reaches 1/40 km^-2. Moreover, significant variations in improvement effects across experiments within the same group indicate the substantial impact of spatial distribution of assimilated AWSs on forecast outcomes. This study provides a reference for devising more efficient and cost-effective data assimilation strategies in numerical weather prediction.

How to cite: Qiao, Z., Li, B., Gong, A., and Ni, G.: Impact of Spatial Density of Automatic Weather Station Data on Assimilation Effectiveness in WRF-3DVar Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5909, https://doi.org/10.5194/egusphere-egu24-5909, 2024.

Weather forecasting is a vital topic in meteorological analysis, agriculture planning, disaster management, etc. The accuracy of forecasts varies with the prediction horizon, spanning from nowcasting to long-range forecasts. The extended range forecast, which predicts weather conditions beyond two weeks to months ahead, is particularly challenging. This difficulty arises from the inherent variability in weather systems, where minor disturbances in the initial condition can lead to significantly divergent future trajectories.

Numerical Weather Prediction (NWP) has been the predominant approach in this field. Recently, deep learning (DL) techniques have emerged as a promising alternative, achieving performance comparable to NWP [1, 2]. However, their lack of embedded physical knowledge often limits their acceptance within the research community. To enhance the trustworthiness of DL-based weather forecasts, we explore a transformer-based framework which considers complex geospatial-temporal (4D) processes and interactions. Specifically, we select the Pangu model [3] with a 24-hour lead time as the initial framework. To extend the prediction horizon to two weeks ahead, we employ a low-rank adaptation for model finetuning, which saves computation resources by reducing the number of parameters to only 1.1% of the original model. Besides, we incorporate multiple oceanic and atmospheric indices to capture a broad spectrum of global teleconnections, aiding in the selection of important features.

Our contributions are threefold: first, we provide an operational framework for foundation models, improving their applicability in versatile tasks by enabling training rather than limiting them to inference stages. Second, we demonstrate how to leverage these models with limited resources effectively and contribute to the development of green AI. Last, our method improves performance in extended-range weather forecasting, offering enhanced prediction skills, physical consistency, and finer spatial granularity. Our methodology achieved reduced RMSE on T2M, Z500, and T850 for 0.13, 139.2, and 0.52, respectively, compared to IFS. In the future, we plan to explore other settings, such as predicting precipitation and extreme temperatures.

REFERENCES
[1] Nguyen, Tung, et al. "ClimaX: A foundation model for weather and climate." arXiv preprint arXiv:2301.10343 (2023).
[2] Lam, Remi, et al. "Learning skillful medium-range global weather forecasting." Science (2023): eadi2336.
[3] Bi, Kaifeng, et al. "Accurate medium-range global weather forecasting with 3D neural networks." Nature 619.7970 (2023): 533-538.

How to cite: Zhao, S., Xiong, Z., and Zhu, X. X.: Enhanced Foundation Model through Efficient Finetuning for Extended-Range Weather Prediction, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6155, https://doi.org/10.5194/egusphere-egu24-6155, 2024.

How to cite: Whan, K., Cambier van Nooten, C., Schmeits, M., Wijnands, J., Schreurs, K., and Shapovalova, Y.: Improving precipitation nowcasting using deep generative models: a case-study and experiences in R2O , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6545, https://doi.org/10.5194/egusphere-egu24-6545, 2024.

Korea Meteorological Administration (KMA) has been providing the public with an hourly precipitation forecast updated every 10 minutes for the next 6 hours since 2015. This forecasts, named as the Very Short-Range Forecast (VSRF), differs from other longer forecasts ? such as short-range and medium-range forecasts issued by forecasters. The VSRF is automatically generated by a system based on two different models: MAPLE (McGill Algorithm for Precipitation nowcasting by Lagrangian Extrapolation) and KLAPS (Korea Local Analysis and Prediction System). 

MAPLE, based on Variational Echo Tracking (VET) from radar observations, has an intrinsic disadvantage: its performance decreases rapidly. On the other hand, numerical weather prediction systems like KLAPS are not initially as effective as MAPLE due to model balancing factors such as spin-up, but they maintain initial skill for a slightly longer period. Therefore, to provide the best predictions to the public, it is necessary to merge the two models properly. KMA conducted tests to determine the optimal way to utilize both models and established weights for each model based on their performance and precipitation tendencies. According to a 4-year evaluation, MAPLE outperforms for up to 2 hours, while KLAPS performs better after 4 hours. Consequently, the two models were merged with a hyperbolic tangent weight applied between 2 and 4 hours, and we named it as the best guidance. 

The best guidance was verified against precipitation observed by 720 raingauges over South Korea during the summer seasons from 2020 to 2023. It demonstrated better skill compared to both MAPLE and KLAPS. The average threat scores, with a rain intensity threshold of 0.5 mm/h throughout the forecast period, were 0.40 for the best guidance, 0.38 for MAPLE, and 0.35 for KLAPS.

The best guidance depends on both MAPLE and KLAPS. Therefore, KMA is actively working to improve the performance of each model. Additionally, a very short-range model based on AI is currently under development and running in semi-operations.

How to cite: lee, H. Y., Kim, J., Son, J., and Kim, S.-J.: Very Short-Range Precipitation Forecast in Korea Meteorological Administration, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6856, https://doi.org/10.5194/egusphere-egu24-6856, 2024.

The Very Short-Range Forecast (VSRF) for precipitation from the Korea Meteorological Administration (KMA) is released every 10 minutes, providing forecasts for the next 6 hours at 10-minute intervals. However, when the forecast is provided to the public, it is updated at 10-minute interval, but only provides up to 6 hours at every hour. Consequently, from the public's perspective, forecasts for specific times may change every 10 minutes. While this allows users to access the latest updates, it also poses a challenge in terms of reduced reliability due to constantly changing predictions.

This study aims to assess the prediction performance and variability between forecasts released at 10-minute intervals and those at 1-hour intervals. We evaluated with the Very Short-Range Forecast numerical model KLAPS in VSRF and seek to determine which approach offers more valuable information from the public's standpoint. The assessment focuses on two distinct types of precipitation. The first involves convective showers, which sporadically appear over short durations, driven by atmospheric instability during the Korean Peninsula's summer. The second relates to systematic precipitation associated with a frontal boundary accompanying a medium-scale low-pressure system. For convective showers, the 1-hour interval exhibits better performance and continuity, particularly as the forecast time extends. In the case of systematic precipitation, the 1-hour interval remains superior, though the skill is not as prominent as with convective showers. This highlights that an abundance of information doesn't always equate to high-quality information.

How to cite: Son, J., Kim, J., and Kim, S.: Does more frequent Very Short-Range Forecast provide more useful information?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6873, https://doi.org/10.5194/egusphere-egu24-6873, 2024.

The 2024 Gangwon Winter Youth Olympic Games (GANGWON 2024) will be held in the province of Gangwon in the Republic of Korea from January 19 to February 1, 2024, which already hosted the Olympic Winter Games PyeongChang 2018. In order to successfully host these first Winter YOG to be held in Asia, which will be held for the first time in Asia, it is necessary to provide customized weather information for decision-making in game operation and support in establishing game strategies for athletes and their teams. Accordingly, the Korea Meteorological Administration develops point-specific numerical forecast guidance for major stadiums and provides it to the field to support successful hosting of YOG and improvement of performance. Numerical forecast guidance is the final data delivered to consumers or forecasters as post-processed numerical model data that has been corrected by applying altitude correction and statistical methods to produce highly accurate forecasts. For a total of 13 forecast elements (temperature, minimum/maximum temperature, humidity, wind direction/speed, precipitation, new snow cover, sky conditions, precipitation probability, precipitation type), we developed user-customized numerical forecast guidance specialized for competition points  (Gangneung Olympic Park, Pyeongchang Alpensia Venue, Biathlon Center, Olympic Sliding Center departure/arrival, Wellyhilli departure/arrival, High1 departure/arrival). Through the process of Perfect Prognostic Method (PPM), Model Output Statistics (MOS), optimization, and optimal merging, the systematic errors inherent in the numerical model are removed, and the optimal data (BEST) with improved forecasting performance is provided as customized numerical forecast guidance specific to stadium locations.  In the prediction performance evaluation for the period of December 2023, the accuracy (improvement rate) compared to the average of available models was temperature 1.49℃ (18%), humidity 12% (25%), wind speed 1.87m/s (33%), and visibility 12.8km (17%).

How to cite: Kim, Y.-H., Cho, E., Jang, S., Kim, J., Bok, H., and Kim, S.: Development of stadium-specific numerical forecast guidance for weather forecast for the 2024 Gangwon Winter Youth Olympic Games, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7086, https://doi.org/10.5194/egusphere-egu24-7086, 2024.

This study was conducted to develop a modified precipitation model for its amount and existence by combining machine learning method, Extreme Gradient Boosting(XGBoost), with ECMWF IFS(Integrated forecasting system) and, finally, estimate the related performance.

According to the analysis of regional precipitation characteristic, prior to its development, the ratio of precipitation existence was various on a basis of a forecast’s district and its season. These different patterns on each district makes it necessary to develop the regional and seasonal model respectively.

And, the first attempt at the machine learning showed the importance of each feature as input-variables, as a result of which cloud physics-related features, for example large-area precipitation, total precipitation, visibility and what not, proved so significant. However, the insufficient amount of these feature’s data seemed to result in overfitting. And therefore, the feasible features, except for cloud physics-related things, of IFS data were used. In addition, auxiliary features and their gradient for every lead-time were calculated and added: relative vorticity, divergence, equivalent potential temperature, main 6 patterns for Korean summer and so on. The number of features amounted to around 144 with which for the 9-year training set, 2013~2021, based learning to be conducted regionally, followed by using validation-set of 2022.

As a result of validation for precipitation existence and its amount up to 135 hours ahead on the 10 regions at 00UTC in summer of 2022, Critical Success Index(CSI) was more improved by 10.3% than before. Accuracy(ACC) for each lead-time rose by 6% and its fluctuation also decreased. And the correction by this machine learning alleviated the overfitting trend of precipitation forecast amount produced by the original model, and improved correlation and linearity between observation and forecast. In particular, while the machine learning prevailed over the original model up to 100 hours ahead, from then on, both of them showed similar performance or that of the former was downward slightly. If the above-mentioned cloud physics features are used to further sharpen machine learning technique, its performance should be enhanced more and more.

How to cite: Cho, E., Kim, Y.-H., Kim, S., and Kwon, Y. C.: The Development of precipitation model modifed with ECMWF IFS and XGBoost and its performance verification, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7091, https://doi.org/10.5194/egusphere-egu24-7091, 2024.

Since fog is an important weather phenomenon affecting the traffic safety, accurate fog forecasting should be attained to minimize meteorological disasters. Most fog forecasts determine only the presence or absence of fog based on less visibility than 1 km, which is known as the visibility diagnostic method. During this process, fog could be predicted by the visibility calculated in the numerical weather prediction (NWP) model using the cloud liquid water content (LWC) near the surface. In this study, we investigated to increase the accuracy of fog forecast by optimizing the reconstruction of moisture distribution method, which can simulate the intensity of fog as well as the presence or absence of fog. The performances of the fog simulations were examined by modifying the relative humidity threshold at a height of 2 m and the stability parameters which affect turbulence and also one of the important criteria for fog occurrence. When we applied the optimize parameters to fog prediction in the winter seasons, the probability of detection (POD) has been increased significantly from 0.21 to 0.54. These improvements were attributed to the corrected relative humidity threshold and the stability parameters. Although the false alarm rate (FAR) remained almost unchanged, the critical success index (CSI) has been improved slightly lesser than those of the POD. When we analyzed the life cycle of fog, it takes time for the NWP model to simulate water droplets in the fog-developing stage. Therefore, the accuracy of the fog simulation is intimately related to the reconstruction of moisture distribution. The NWP model, however, showed a better performance in the process of fog dissipation than the reconstruction of moisture distribution method that was sensitive to temperature and turbulence. In conclusion, the reconstruction of moisture distribution led to a considerable improvement of the fog prediction in the generation and development stage since we used the optimized humidity threshold. It is also expected that accurate fog prediction could be achieved in the future by considering the aerosol effects, which is another importance factor for the fog generation.

How to cite: Kim, E., Park, S.-Y., You, J.-W., and Lee, S.-H.: Improvements in fog predictions via a modified reconstruction of moisture distribution using the Weather Research and Forecasting(WRF) model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7291, https://doi.org/10.5194/egusphere-egu24-7291, 2024.

Rapid advancements in data-driven weather prediction have shown notable success, particularly in nowcasting, where forecast lead times span just a few hours. Transformer-based models, in particular, have proven effective in learning spatiotemporal connections of varying scales by leveraging the attention mechanism with efficient space-time patching of data. This offers potential improvements over traditional nowcasting techniques, enabling early detection of convective activity and reducing computational costs. 

In this presentation, we demonstrate the effectiveness of a modified Earthformer model, a space-time Transformer framework, in addressing two specific nowcasting challenges. First, we introduce a nowcasting model that predicts ground-based 2D radar mosaics up to 2-hour lead time with 5-minute temporal resolution, using geostationary satellite data from the preceding two hours. Trained on a benchmark dataset sampled across the United States, our model exhibits robust performance against various impactful weather events with distinctive features. Through permutation tests, we interpret the model to understand the effects of input channels and input data length. We found that the infrared channel centered at 10.3 µm contains skillful information for all weather conditions, while, interestingly, satellite-based lightning data is the most skilled at predicting severe weather events in short lead times. Both findings align with existing literature, enhancing confidence in our model and guiding better usage of satellite data for nowcasting. Moreover, we found the model is sensitive to input data length in predicting severe weather events, suggesting early detection of convective activity by the model in rapidly growing fields. 

Second, we present the initial attempts to develop a multi-source precipitation nowcasting model for Austria, tailored to predict impactful events with convective activities. This model integrates satellite- and ground-based observations with analysis and numerical weather prediction data to predict precipitation up to 2-hour lead time with 5-minute temporal resolution.  

We conclude by discussing the broad spectrum of applications for such models, ranging from enhancing operational nowcasting systems to providing synthetic data to data-scarce regions, and the challenges therein.

How to cite: Küçük, Ç., Giannakos, A., Schneider, S., and Jann, A.: Nowcasting with Transformer-based Models using Multi-Source Data , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7536, https://doi.org/10.5194/egusphere-egu24-7536, 2024.

The run-to-run variability of numerical weather prediction systems is at the heart of forecasters' concerns, especially in the decision-making process when high-stakes events are considered. Indeed, forecasts that are brutally changing from one run to another may be difficult to handle and can lose credibility. This is all the more true nowadays, as many meteorological centres have adopted the strategy of increasing runs frequency, some reaching hourly frequencies. However, this aspect has received little attention in the literature, and the link with predictability has barely been explored.

In this study, run-to-run variability is investigated through 24h-accumulated precipitations forecasted by AROME-EPS, Météo-France's high resolution ensemble, which is refreshed 4 times a day. Focusing on the probability of some (warning) thresholds being exceeded, results suggest that how forecasts evolve over successive runs can be used to improve their skill, especially reliability. Various possible aspects of run sequence have been studied, from trends to rapid increases or decreases in event probability at short lags, also called "sneaks" or "phantoms", as well as the persistence of a non-zero probability through successive runs. The added value provided by blending successive runs, known as lagging, is also discussed.

How to cite: Marchal, H., Bouttier, F., and Nuissier, O.: On the usefulness of considering the run-to-run variability for an ensemble prediction system, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7753, https://doi.org/10.5194/egusphere-egu24-7753, 2024.

The research project “FOrecasting radiation foG by combining station and satellite data using Machine Learning (FOG-ML)” represents a comprehensive effort to advance radiation fog prediction using machine learning (ML) techniques, with focus on the XGBoost algorithm. The nowcasting period is up to four hours into the future.

The initial phase of the project involved developing a robust classification-based model that could accurately forecast the occurrence of radiation fog, a challenging meteorological phenomenon. Radiation fog is particularly difficult to predict because it depends on a complex interplay of factors such as ground cooling, humidity, and minimal cloud cover. It often forms rapidly and in local areas. This required careful analysis of the chronological order of the data and consideration of autocorrelation to increase the effectiveness of model training.

Building upon this foundation, the next two phases concentrated on improving the model’s forecasting performance for visibility classes (step 2) and for absolute visibility values (step 3). The main focus was then on a nowcasting period of up to two hours. This nowcasting period is critical in fog prediction as it directly impacts transportation planning and safety. The use of ground-level observations in step 2 and integration of satellite data in step 3 provided a rich dataset that allowed for more nuanced model training and validation.

In the latest phase of research, satellite data has been incorporated to further refine the prediction model, especially regarding the fog formation and dissipation. Satellite imagery provides additional variables of atmospheric data that are not readily available from ground-based observations. This integration aims to address one of the inherent limitations in fog forecasting methods, particularly in areas where ground-based observations are sparse.

Throughout the different stages, the project emphasized the need for thorough data processing and validation. This included the implementation of cross-validation techniques to assess the generalizability of the models and the use of various metrics to gauge their predictive power. This has also included the incorporation of trend information, which has proven to be crucial for forecasting with XGBoost. Our research has also shown that not only the overall performance, but also the performance of the transitions (fog formation and resolution) should be analyzed to get a complete picture of the model performance. This finding was consistent throughout the entire study, regardless of classification-based forecast or regression-based forecast.

We have been able to significantly improve the performance of our nowcasting model with each step. We will be presenting the key findings and latest results from this research at EGU24.

All results from step 1 can be found in “Current Training and Validation Weaknesses in Classification-Based Radiation Fog Nowcast Using Machine Learning Algorithms” from Vorndran et al. 2022. All results from step 2 can be found in “Improving classification-based nowcasting of radiation fog with machine learning based on filtered and preprocessed temporal data” from Schütz et al. 2023.

How to cite: Schütz, M., Bendix, J., and Thies, B.: Radiation fog nowcasting with XGBoost using station and satellite data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8449, https://doi.org/10.5194/egusphere-egu24-8449, 2024.

Ensemble weather forecast post-processing can generate more reliable probabilistic weather forecasts compared to the raw ensemble. Often, the post-processing method models the future weather probability distribution in terms of a pre-specified distribution family, which can limit their expressive power. To combat these issues, we propose a novel, neural network-based approach, which produces forecasts for multiple lead times jointly, using a single model to post-process forecasts at each station of interest. We use normalizing flows as parametric models to relax the distributional assumption, offering additional modeling flexibility.We evaluate our method for the task of temperature post-processing on the EUPPBench benchmark dataset. We show that our approach exhibits state-of-the-art performance on the benchmark, improving upon other well-performing entries. Additionally, we analyze the performance of different parametric distribution models in conjunction with our parameter regression neural network, to better understand the contribution of normalizing flows in the post-processing context. Finally, we provide a possible explanation as to why our method performs well, exploring per-lead time input importance.

How to cite: Mlakar, P., Merše, J., and Faganeli Pucer, J.: Ensemble forecast post-processing based on neural networks and normalizing flows, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9528, https://doi.org/10.5194/egusphere-egu24-9528, 2024.

According to the demand of the Winter Olympic Organizing Committee for snow depth prediction, the application of multi-source new data in snow depth was studied based on densely artificial snow-depth measurement, microscopic snowflake shape observation and PARSIVEL data. The specific conclusions are as follows: (1) Most of the Snow-Liquid-Ratio(SLR) in Beijing competition zone was between 0.69 and 1.43 (unit: cm/mm, the same below), while that in Yanqing zone was between 0.53 and 1.17. But 7.5% of the SLRs in Yanqing zone exceeded 3.5, which all occurred in the same period of the key service time of 2022 Beijing Winter Olympics, making it more difficult to predict new snow depth. (2) The higher the SLR, the lower the daily minimum surface temperature and lowest air temperature.  Plate or column ice crystals, rimed snowflakes, and dendritic snowflakes were observed, whose corresponding SLRs increased. The average falling speed of particles falling below 2m/s can be used as an indicator of phase transfer. (3) The vertical distributions of temperature and humidity with SLR <1 or >2 were summarized. It was found that when the cloud area coincided with the dendritic growth zone with height close to Yanqing zone, the SLR would be more than 2, higher than that of Beijing zone. (4) A weather concept model generating large SLR was extracted. Snow in Beijing is often accompanied by easterly winds in boundary layer, which is easy to form a wet and ascending layer in the lower troposphere due to the blocking of western mountain. In the late winter season, helped by the temperature’s profile, it tends to produce unrimed dendritic snowflakes, leading to a great SLR.

How to cite: Du, J., Yu, B., Dai, Y., Li, S., Xu, L., Fu, J., Li, L., and Jing, H.: Application Research of Multi-source New Detection Data in Snow Depth Prediction for Beijing Winter Olympics, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9659, https://doi.org/10.5194/egusphere-egu24-9659, 2024.

The rapid development of artificial intelligence algorithms has generated considerable interest in the scientific community. The number of scientific articles relating to applying these algorithms for weather forecasting has increased dramatically in the last few years. In addition, the recent operational launch of products such as GraphCast has put this area of research even more in the spotlight. This work uses different Machine Learning and Deep Learning algorithms, namely ANN (Artificial Neural Network), RF (Random Forest) and GNN (Graph Neural Network), with the aim to improve the short-term (1-day lead time) forecasts provided by a physically-based forecasting system. Specifically, the CeSMMA laboratory, since January 2020, has been producing daily forecasts accessible via the https://cesmma.unical.it/cwfv2/ webpage related to a large portion of southern Italy. The NWP (Numerical Weather Prediction) system is based on the WRF (Weather Research and Forecasting) model, with boundary and initial conditions provided by the GFS (Global Forecasting System) model. The AI algorithms post-process the NWP output, applying correction factors achieved by a two-year training considering the observations of the dense regional monitoring network composed of ca. 150 rain gauges.

The results show that the AI is able to improve daily rainfall forecasts compared to ground-based observations. Specifically, the ANN reduces the average MSE (Mean Square Error) by approximately 29% and the RF by 21% with respect to the WRF forecast for the whole study area (about 15’000 km²). Moreover, the GNN applied to a smaller area (considering only 22 rain gauges) further reduces the MSE by 35% during the heaviest rainfall months.

In addition to improving the performance of the forecast, the AI-based post-processing provides reasonable precipitation spatial patterns, reproducing the main physical phenomena such as the orographic enhancement since it is not a surrogate model and benefits from the original physically-based forecasts.

Acknowledgements. This work was funded by the Next Generation EU - Italian NRRP, Mission 4, Component 2, Investment 1.5, call for the creation and strengthening of ‘Innovation Ecosystems’, building ‘Territorial R&D Leaders’ (Directorial Decree n. 2021/3277) - project Tech4You - Technologies for climate change adaptation and quality of life improvement, n. ECS0000009. This work reflects only the authors’ views and opinions, neither the Ministry for University and Research nor the European Commission can be considered responsible for them.

How to cite: Furnari, L., Yousuf, U., De Rango, A., D'Ambrosio, D., Mendicino, G., and Senatore, A.: Machine and Deep Learning algorithms to improve weather forecasts over a complex orography Mediterranean region, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9935, https://doi.org/10.5194/egusphere-egu24-9935, 2024.

On May 17, 2019, a rare severe convective weather occurred in Beijing, accompanied by local heavy rainstorm, hail, thunderstorm and gale. This severe convective weather occurred significantly earlier than normal years, bringing great challenge to the forecast. Using multiple observation data and radar four-dimensional variational assimilation products to analyze the triggering and development evolution of this severe convection. Under the conditions of no obvious weather scale system and local high potential unstable energy, the eastward advancement of the sea breeze front was the main factor triggering strong convection. As the northwest wind in the air increasing, the environmental conditions became stronger vertical wind shear, which was beneficial for the storm to maintain for a longer period of time. The supercell was the main cause of the convective weather. During the development of storms, they split into two parts and moved counterclockwise. The southern echo gradually weakened as it moved northward, while the northern echo moved southward, strengthening and developing into a super cell accompanied by a mesocyclone. The significant fluctuations in the height of the 0 ℃ layer within a small range resulted in different melting rates of hail during its descent, leading to the formation of spiky hail.

How to cite: Yu, B.: Analysis of a rare severe convective weather event in spring in Beijing of China, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10809, https://doi.org/10.5194/egusphere-egu24-10809, 2024.

Forecasting precipitation over Africa, the largest landmass in the tropics, has been a long standing problem. The unique conditions of the West African monsoon result in large and long lasting mesoscale convective systems. Global numerical weather prediction (NWP) models have gridsizes in the 10s of kilometers, particular when run in ensemble mode, leaving convection to be parameterized. This often results in precipitation being forecast on too large scales, in the wrong places, and with too weak intensity, ultimately leading to little to no skill in tropical Africa.

It has been argued that convection permitting (CP) NWP forecasts would cure some of the problems described above but those have only recently become feasible in an operational setting, although ensembles are still deemed to be too expensive. Here, we systematically compare regional deterministic CP and global ensemble forecasts in the region over multiple rainy seasons for the first time. We analyze CP forecasts from AROME and Met Office Tropical African Model, and seven global ensemble forecasts from the TIGGE archive, both individually and as a multi-model ensemble. In order to create an uncertainty estimate, we create neighborhood ensembles from CP forecasts at surrounding grid points, which allows for a fair comparison to the ensembles and a probabilistic climatology. Considering both precipitation occurrence and amount, we use the Brier score (BS) and the continuous ranked probability score (CRPS), along with their decompositions in discrimination, miscalibration and uncertainty, for evaluation.

Using neighborhood methods, deterministic forecasts are turned into probabilistic forecasts, allowing a fair comparison with ensembles. All numerical forecasts benefit from Neighborhoods, improving their BS and CRPS in terms of both miscalibration and discrimination. We find all individual forecasts to have skill over most of tropical Africa, with some ensemble models lacking skill in some regions and the multi model showing the most overall skill. The CP forecasts TAM and AROME outperform non-CP forecasts mainly in the region of the little dry Season and the Soud. However, large areas of low skill in terms of CRPS remain and even with high resolution, numerical models still struggle to predict precipitation in tropical Africa. 

How to cite: Rein, F., Fink, A. H., Gneiting, T., Peyrille, P., Warner, J., and Knippertz, P.: Can convection permitting forecasts solve the tropical African precipitation forecasting problem?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12420, https://doi.org/10.5194/egusphere-egu24-12420, 2024.

In recent years, several national meteorological services (NMSs) have invested considerable resources in the development of a seamless prediction system: rapidly updating forecasts that integrate the latest observations, covering timescales from minutes to days or longer ahead (e.g. DWD's SINFONY; FMI's ULJAS, MetOffice's IMPROVER and Geosphere Austria's SAPHIR) [1]. This move was motivated mainly by rising expectations from end users such as hydrological services, local authorities, the renewable energy sector and the general public. The development of seamless prediction systems was made possible thanks to the increasing availability of high-resolution observations, continuing advances in numerical weather prediction (NWP) models, nowcasting algorithms, and improved strategies to combine multiple information sources optimally. Moreover, the rise of AI/ML techniques in forecasting and nowcasting can further reduce the computational cost to generate frequently updating seamless operational forecast products.

We present the journey of building the Belgian seamless prediction system at the Royal Meteorological Institute of Belgium, with the working title "Project IMA". IMA uses both the deterministic INCA-BE and the probabilistic pysteps-BE systems to combine nowcasts with the ALARO and AROME configurations of the ACCORD NWP model. In the lessons learned along the way, we focus on what is often omitted, moving from research to operations, and integrating what we learn from operations back into research. We discuss the benefits of integrating new developments within the free and open-source software (FOSS) pysteps [2]. Our experience shows that using and contributing to FOSS not only leads to more transparency and reproducible, open science; it also enhances international collaboration and can benefit other users, including developing countries, bringing us a step closer to the ambitious goal of Early Warnings for All by 2027 [3].

References

[1] Bojinski, Stephan, et al. "Towards nowcasting in Europe in 2030." Meteorological Applications 30.4 (2023): e2124.

[2] Imhoff, Ruben O., et al. "Scale‐dependent blending of ensemble rainfall nowcasts and numerical weather prediction in the open‐source pysteps library." Quarterly Journal of the Royal Meteorological Society 149.753 (2023): 1335-1364.

[3] WMO, "Early warnings for all: Executive action plan 2023-2027", 8 Nov 2022,  https://www.preventionweb.net/quick/75125.

How to cite: De Cruz, L., Van Ginderachter, M., Reyniers, M., Deckmyn, A., Dehmous, I., De Kock, S., Dewettinck, W., Imhoff, R., Montandon, E., and Reinoso-Rondinel, R.: Project IMA: Lessons Learned from Building the Belgian Operational Seamless Ensemble Prediction System, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12855, https://doi.org/10.5194/egusphere-egu24-12855, 2024.

Moderate to heavy rain produced by slantwise ascent of moist air above the cold high is prevalent in cold season in East China. The slantwise ascent is usually characterized by a southwest moist flow aroused by the so-called southern branch trough of 500hPa level to the south of the Qinghai-Tibet Plateau, while the cold high is usually formed by cold air damming, which is familiar to weather forecasters due to topographic feature of East China. The routine short-range forecast skill for this kind of precipitation of weather forecasters is usually limited by model performance. Through large sample model verification, our study indicates that, for the rainfall produced by southwesterly moist flow ascending above the cold high, the ECMWF model always underestimates the rainfall amount on the northeastern part of the rainfall belt, which could be taken as a systematic bias of the state-of-the-art global model. Our case studies indicate that the underestimation of rainfall amount is related to the weaker slant ascent of moist southwest flow forecast by ECMWF model than observation or reanalysis. The southwest flow above the northeastern flow induced by the cold high forms strong wind shear and warm-moist advection, which favors the occurrence of conditional symmetric instability producing strong slantwise ascent not well reflected by global model.

How to cite: Hu, N. and Fu, J.: Investigating Model Forecast Bias for Rainfall Produced by Slantwise Ascent above Cold High, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13809, https://doi.org/10.5194/egusphere-egu24-13809, 2024.

Over the past few years purely AI-driven global weather forecasting models have emerged that show increasingly impressive skill, raising the question whether AI models might soon compete with NWP models for selected forecasting tasks. At this point these AI-based models are still in the proof-of-concept stage and not ready to be used for operational forecasting, but entirely new AI-models emerge every 2-3 months, with rapidly increasing abilities. Furthermore, many of these models are orders of magnitude faster than NWP models and can run on modest computational resources enabling repeatable on-demand forecasts competitive with NWP. The low computational cost enables the creation of very large ensembles, which better represent the tails of the forecast distribution, which, if an ensemble is well calibrated, allows for better forecasting of rare and extreme events.

However, these AI-based weather forecasting models have not yet been rigorously tested by the meteorological community, and their utility to operational forecasters is unknown. In this presentation we propose several studies to address the above issues, grouped into two central foci:

(1) Nature of AI models: AI-based models have very different characteristics from NWP models. Thus, in addition to applying evaluation procedures developed for NWP models, we need to develop procedures that test for AI-specific weaknesses. For example, NWP models and their physics backbone guarantee certain properties - such as dynamic coupling between fields - that AI-based models are not required to uphold. Developing suitable tests is based on a fundamental understanding of the AI-based models.

(2) Forecaster Perspective: Evaluation of weather forecasting models should be performed with respect to particular applications of weather forecasts, and it is critical to have research meteorologists and operational forecasters involved in the evaluation process. Our initial evaluation of AI-based models in CIRA weather briefings revealed that these models have characteristics that make interpretation of their forecasts fundamentally different from the physics-based NWP model predictions meteorologists are familiar with. For example, the increasing “blurriness” of AI-based predictions with longer lead times is not a reflection of weaker atmospheric circulations, but rather a reflection of uncertainty. Evaluations aimed at specific meteorological phenomena and atmospheric processes will allow the community to make informed decisions in the future regarding in what environments and for which applications AI-based weather forecasting models may be safe and beneficial to use.

In summary, AI-based weather forecasts have different characteristics from familiar dynamically-based forecasts, and it is thus important to have a robust research plan to evaluate many different characteristics of the models in order to provide guidelines to operational forecasters and feedback to model developers. In this abstract we propose a number of characteristics to evaluate, present results we already obtained, and suggest a research plan for future work.

How to cite: Ebert-Uphoff, I., Stewart, J. Q., and Radford, J. T. and the CIRA-NOAA team: A Research Agenda for the Evaluation of AI-Based Weather Forecasting Models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13853, https://doi.org/10.5194/egusphere-egu24-13853, 2024.

Large-eddy simulations of an idealized tropical cyclone (TC) were conducted as benchmarks to provide statistical information about subgrid convective clouds at a convection-permitting resolution over a TC convection system in different stages. The focus was on the vertical and spatial distributions of the subgrid cloud and associated mass flux that need to be parameterized in convection-permitting models. Results showed that the characteristics of the subgrid clouds varied significantly in various parts of the TC convection system. Statistical analysis revealed that the subgrid clouds were mainly located in the lower troposphere and exhibited shallow vertical extents of less than 4 km. The subgrid clouds were also classified into various cloud regimes according to the maximum mass flux height. Local subgrid clouds differed in mass-flux profile shape and magnitude at various regimes in the TC convection system.

How to cite: Zhang, X. and Bao, J.-W.: Statistics of the Subgrid Cloud of an Idealized Tropical Cyclone at Convection-Permitting Resolution, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14232, https://doi.org/10.5194/egusphere-egu24-14232, 2024.

The Korea Meteorological Administration (KMA) has established and operated a standard verification system of the operational NWP models to evaluate the predictive performance of NWP model and compare them with other NWP models operated by domestic and foreign organization. This secures the objectivity of the verification results by applying the verification standards (WMO-No.485) presented by World Meteorological Organization (WMO), and being able to compare the performance with the numerical forecasting models of other institutions under the same conditions. The NWP models to be verified is a global, a regional, very short-range, and an ensemble prediction system and verification against analyses and observations are performed twice a day (00 UTC, 12 UTC). In addition to standard verification, precipitation, typhoon and various verification indexes (CBS index, KMA index, jumpiness index) are verified and used to evaluate the utilization of NWP models. The Korea Integrated Model (KIM), which is developed for Korea’s own NWP model, has been in operation since April 2020. Since the start of operation, the RMSE of 500hPa geopotential height (in Northern Hemisphere) has decreased every year, showing that forecast performance is improving. In addition, it can be seen that the 72-hour prediction accuracy for 12-hour accumulated precipitation (1.0 mm or more) in the Korean Peninsula area (75 ASOS stations) is also improving. As such, this study intends to discuss the predictive performance of the numerical forecast model based on the standard verification system and plans to improve the verification system in the future. 

How to cite: Park, S., Park, H., Lee, H., Song, S., Ha, J.-C., and Kwon, Y. C.: Status and Plan of Standard Verification System for the NWP model in Korea Meteorological Administration, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14541, https://doi.org/10.5194/egusphere-egu24-14541, 2024.

Radar-based precipitation nowcasting is one of the most prominent applications of deep learning (DL) in weather forecasting. The accurate forecast of extreme precipitation events remains a significant challenge for deep learning models, primarily due to their complex dynamics and the scarcity of data on such events. In this work we present the application of the latest state-of-the-art generative architectures for radar-based nowcasting, focusing on extreme event forecasting performance. We analyze a declination for the nowcasting task of all the three main current architectural approaches for generative modeling, namely: Generative Adversarial Networks (DGMRs), Latent Diffusion (LDCast), and our novel proposed Transformer architecture (GPTCast). These models are trained on a comprehensive 1-km scale, 5-minute timestep radar precipitation dataset that integrates multiple radar data sources from the US, Germany, the UK, and France. To ensure a robust evaluation and to test the generalization ability of the models, we concentrate on a collection of out-of-domain extreme precipitation events over the Italian peninsula extracted from the last 5 years. This focus allows us to assess the improvements these techniques offer compared to extrapolation methods, evaluating continuous (MSE, MAE) and categorical scores (CSI, POD, FAR), ensemble reliability, uncertainty quantification, and warning lead time. Finally, we analyze the computational requirements of these new techniques and highlight the caveats that must be considered when operational usage of these methods is envisaged. 

How to cite: Franch, G., Tomasi, E., Wanjari, R. U., and Cristoforetti, M.: Nowcasting of extreme precipitation events: performance assessment of Generative Deep Learning methods, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15431, https://doi.org/10.5194/egusphere-egu24-15431, 2024.

The crossing-point forecast (CPF) is a new type of ensemble-based forecast developed at the European Centre for Medium-Range Weather Forecasts. The crossing point refers to the intersection between the cumulative probability distribution of a forecast and the cumulative probability distribution of a model climatology. Originally, the CPF has emerged as a consistent forecast with the diagonal score, a weighted version of the continuous ranked probability score targeting high-impact events. Ranging between 0 and 1, the CPF can serve as an index for high-impact weather and thus directly be compared with the well-established extreme forecast index. The CPF is also interpretable in terms of a return period and conveys a sense of a “probabilistic worst-case scenario”.  Using a recent example of an extreme event affecting Europe, we illustrate and discuss the performance and specificities of this new type of forecast for extreme weather forecasting.

Ben Bouallegue, Z (2023).  Seamless prediction of high-impact weather events: a comparison of actionable forecasts. arXiv:2312.01673

How to cite: Ben Bouallegue, Z.: Forecasting extreme events with the crossing-point forecast , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16617, https://doi.org/10.5194/egusphere-egu24-16617, 2024.

In just two years, the idea of an operational data-driven system for medium-range weather forecasting has been transformed from dream to very real possibility. This has occurred through a series of publications from innovators, which have rapidly improved deterministic forecast skill. Our own evaluation confirms that these forecasts have comparable deterministic skill to NWP models across a range of variables. However, on medium-range timescales probabilistic forecasting, typically achieved through ensembles, is key for providing actionable insights to users. ECMWF is building on top of these recent works to develop a probabilistic forecasting system, AIFS. We will showcase results from our progress towards this system and outline our roadmap to operationalisation.

How to cite: Ben Bouallegue, Z., Alexe, M., Chantry, M., Clare, M., Dramsch, J., Lang, S., Lessig, C., Magnusson, L., Prieto Nemesio, A., Pinault, F., Raoult, B., and Tietsche, S.: AIFS – ECMWF’s Data-Driven Probabilistic Forecasting , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17158, https://doi.org/10.5194/egusphere-egu24-17158, 2024.

The increasing integration of renewable energy resources to the national grids necessitates
accurate prediction of power generation from those sources in terms of secure operation of
electricity grid system and energy trading. Electricity generation of renewable energy power
plants such as wind and solar are inherently affected by weather conditions. The wind condition
particularly is affected by surface characteristics such as orography and vegetation, therefore it is
the one of the near surface atmospheric variables having the strongest local variability. The high-
resolution Numerical Weather Prediction (NWP) models are utilized to take the local conditions
into account. WRF model is the one of the most common NWP models having been widely
investigated by various researchers. On the other hand, The Model for Prediction Across Scales
(MPAS) is a relatively new NWP model utilizing non-uniform mesh structures, developed by the
National Center for Environmental Predictions (NCEP). However, there are limited studies in the
literature which compare the prediction performance of WRF and MPAS model in terms of
surface wind speed. This study evaluates the prediction accuracy of near surface wind of two
downscaled NWP models namely, WRF-ARW and MPAS. Both models are configured with
almost identical physics suites and initialized with 3 hourly 00-UTC initialization of Global
Forecast System (GFS) data. The model outputs are obtained at 10 minutes interval for 48 hours
horizon. Hourly averaged model results are compared with observations from 104 on-site
meteorological stations located in Turkiye having different complexity in terms of correlation
coefficient and RMSE.

How to cite: Yalcin, R. D., Yilmaz, M. T., and Yucel, İ.: Evaluation of the Impact of Uniform and Non-Uniform Resolution Implementations in Numerical Weather Prediction Models over the Accuracy of Short-Term Wind Prediction, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17339, https://doi.org/10.5194/egusphere-egu24-17339, 2024.

Compared to alluvial floods, flash or pluvial floods are difficult to predict because they result from intense and brief periods of extreme precipitation. Luxembourg has a history of being impacted by floods, with notable occurrences in January 2011, May 2016, December 2017, January 2018, February 2019, and February 2020. However, July 2021 stands out as the most severe flood year on record in the region. In this study we are aiming to develop, a high-resolution numerical weather prediction (NWP) model for effective local heavy rainfall prediction in a nowcasting scenario and provide real time for flood simulation. The modeling relies on the Weather Research and Forecasting (WRF) model, which incorporates local Global Navigation Satellite System (GNSS) data assimilation and local precipitation observations to simulate small-scale, high-intensity convective precipitation.

As part of this, we will also test run the LISFlood flood model in an operational inundation forecast mode, meaning that the flood model will be run with the WRF precipitation forecasts as inputs.

The WRF model was configured for the Greater Region, utilizing a horizontal grid resolution of 12 km and incorporating high-resolution static datasets. Meteorological data i.e. July 13 -14 2021, from the Global Forecast System (GFS) were employed in the model setup as initial boundary condition. Zenith Total Delay (ZTD) data collected from various GNSS stations (112) across Germany and Luxembourg were assimilated into the model. Additionally, observational datasets including Surface Synoptic Observations (SYNOP), Upper Air Data, Radiosonde measurements (TEMP), and Tropospheric Airborne Meteorological Data Reporting (TAMDAR) were assimilated. Following this integration, an sensitivity analysis of various meteorological parameters such as precipitation, surface temperature (T2), and relative humidity was performed.

Keywords: NWP, WRF, Flash flood, LISFlood, Weather forecast, High-Resolution, GNSS, ZTD

How to cite: Rehman, H. U., Teferle, F. N., Hunegnaw, A., Schumann, G., Zus, F., and Muraleedharan Thundathil, R.: Enhancing Regional NWP Model with GNSS Zenith Total Delay Assimilation: A WRF and WRFDA 3D-Var Approach in the Greater Region of Luxembourg, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18548, https://doi.org/10.5194/egusphere-egu24-18548, 2024.

Recent work within the WCSSP FORSEA project and its successor FORWARDS has demonstrated that a hybrid statistical-dynamical forecasting technique combining model ensemble forecasts of equatorial waves with climatological rainfall statistics conditioned on wave phase and amplitude can provide additional skill in predicting high impact weather. The underlying rationale for the technique is twofold. Firstly that high impact rainfall events in the tropics are commonly associated with presence of equatorial waves; and secondly that while global models can adequately predict the evolution of dynamical structure of equatorial waves on time-scales of several days they do not predict the relationship between waves and rainfall well. In tests using the Met Office Global and Regional Forecasting System (MOGREPS) the hybrid forecast is found to outperform model rainfall forecasts from both the global and regional convection permitting versions of MOGREPS, however a weighted blend of the MOGREPS forecasts and the hybrid forecast was found to have the highest skill and further improvements in the method may be obtained by taking into consideration the effects of wave-superposition and interaction. To ascertain whether forecasts can be further improved by better predictions of wave amplitude and phase we compare to hypothetical best-case hybrid forecast computed using wave amplitudes and phases taken from reanalysis. This best-case scenario indicates that errors in forecasting all wave types diminish the hybrid forecast's skill, with the most significant reduction observed for Kelvin waves, suggesting that a significant improvement in the prediction of the propagation of equatorial waves would have a significant impact on rainfall prediction in the tropics. 

How to cite: Ferrett, S., Wolf, G., Methven, J., Frame, T., Holloway, C., Martinez-Alvarado, O., and Woolnough, S.: Forecasting tropical high-impact rainfall events using a hybrid statistical dynamical technique based on equatorial waves, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18938, https://doi.org/10.5194/egusphere-egu24-18938, 2024.

In the field of precipitation nowcasting recent deep learning models now outperform traditional approaches such as optical flow [1,2]. Despite their principled effectiveness, these models and their respective training setups suffer from particular shortcomings.  For instance, they often rely on pixel-wise losses, which lead to blurred predictions by which the model expresses its uncertainty [2]. Additionally, these losses can negatively impact training dynamics by overly penalizing small spatial or temporal discrepancies between predictions and actual observations, i.e., the double penalty problem [3]. Generative methods such as discriminative losses or diffusion models do not suffer from the blurring effect as much [1, 4]. However, training these methods is complicated because training success is highly sensitive to the network architecture as well as to the learning setup and its parameterization [5].

Previous research has shown that spatial verification methods such as the fractions skill score offer an easy-to-implement alternative to solve the problem of pixel-wise losses [6, 7]. However, the fact that each pixel within the neighborhood of a spatial kernel is weighted equally poses a limiting factor to their performance and potential. Inspired by theories of cognitive modeling and in relation to the fractions skill score loss, we introduce a dynamic locally binned density (DLBD) loss: Forecasting target is not the actual precipitation in a grid cell but a target distribution, which encodes the density of binned precipitation values in a locally weighted area of grid cells. The loss is then determined via the cross-entropy of the predicted and the target distribution. We show that our novel prediction loss avoids the double penalty problem.  It thus diminishes the negative impact of small spatial offsets. Moreover, it enables the learning model to gradually shift focus towards progressively more accurate predictions.

We achieve best performance by simultaneously training on multiple concurrent forecasting targets that cover different local extents. We schedule the weighting of the loss terms such that the focus shifts from larger to smaller neighborhoods over the course of training. This way, the DL model first learns density dynamics and basic precipitation shifts. Later, it focuses on minimizing small spatial deviations, tuning into the local dynamics towards the end of training.  Our DLBD loss is easy-to-implement and shows great performance improvements.  We thus believe that DLBD losses can also be used by other forecasting architectures where the current forecasting loss precludes smooth loss landscapes.

1: Leinonen et al. 2023: Latent diffusion models for generative precipitation nowcasting with accurate uncertainty quantification
2: Espeholt et al. 2022: Deep learning for twelve hour precipitation forecasts
3: Grilleland et al. 2009: Intercomparison of spatial forecast verification methods.
4: Ravuri et al. 2021: Skilful precipitation nowcasting using deep generative models of radar
5: Mescheder et al. 2018: Which training methods for GANs do actually converge?
6: Roberts et al. 2008: Scale-selective verification of rainfall accumulations from high resolution forecasts of convective events.
7: Lagerquist et al. 2022: Can we integrate spatial verification methods into neural-network loss functions for atmospheric science?

How to cite: Prosi, J., Otte, S., and Butz, M. V.: Dynamic Locally Binned Density Loss, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19257, https://doi.org/10.5194/egusphere-egu24-19257, 2024.

Photovoltaic (PV) power production is increasingly becoming a central pillar in the shift to renewable power sources. The use of solar irradiance has great potential, as it is practically limitless and globally provides magnitudes more energy to the Earth than currently or foreseeable required. Solar irradiance as a power source does, however come with certain downsides. Besides the effects of seasonality and day-night-cycles on its usable potential, it´s broad use suffers mostly from uncertainty through its volatility. The actual extent of solar irradiance at the surface of the Earth is strongly influenced by a variety of atmospheric phenomena, most prominently clouds and atmospheric turbidity. The forecasting of near-future solar irradiance can thereby be beneficial in the estimation of PV power production in itself and with the goal of maintaining a stable equilibrium in electrical grids.

To achieve nowcasts on a larger grid scope, forecasting of solar irradiance from satellite data can substitute forecasting of power output for individual sites. Satellite data, in contrast to ground-based data sources or NWP model estimates, is less reliant on the proper workings of a wide range of externalities. General-purpose spatiotemporal neural networks can be adapted to this task and provide predictions within a very short timeframe, with no requirement of HPC-infrastructure. A sparse model relying on a single satellite-based data source has less points of failure that could affect its forecasting performance and can be very efficient, but this sparsity could also reduce the achievable predictive accuracy. Benefits of smaller and simpler forecasting pipelines therefore may need to be balanced with requirements in terms of accuracy.

To gather more meaningful and reliable results, a variety of spatiotemporal neural networks is implemented and tested to provide a more meaningful foundation. The models were selected and evaluated with respect to their different architectural patterns and designs, to get a notion of architectures beneficial to this task and achieve a more generalizable argument concerning the use satellite data as the sole basis of solar irradiance nowcasting.

In an attempt of improving the viability of satellite-based nowcasting a commonly occurring flaw in near-real-time satellite data sources, missing or skipped frames, solutions to mitigate issues in operational nowcasting are considered. In place of ad-hoc preprocessing such as interpolation of missing data frames, an attempt to condition the models to missing frames is made.

How to cite: Gfäller, P., Schicker, I., and Papazek, P.: Viability of satellite derived irradiance data for ML-based nowcasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19321, https://doi.org/10.5194/egusphere-egu24-19321, 2024.

For weather forecasters and hydrologists, predicting rainfall in the short term – minutes to a few hours – is crucial for a range of applications. While traditional nowcasting methods excel in operational settings, they face limitations in predicting convective storm formation and high-intensity events. Enter deep learning, a powerful tool transforming numerous fields. Convolutional neural networks, in particular, have shown promise in improving nowcasting accuracy. These networks can learn complex patterns and relationships within data, like the intricate tapestry of rainfall variations observed in historical radar sequences. However, capturing long-term dependencies in this data remains a challenge, resulting in fuzzy nowcasts and underestimating high-intensity events. This study proposes a novel deep learning model that goes beyond simple extrapolation, effectively capturing both the spatial correlations and temporal dependencies within rainfall data. Our hybrid convolutional neural network architecture tackles this challenge through three key components: Decoder & Encoder: These modules focus on unraveling the intricate spatial patterns of rainfall and a temporal Module to learn the subtle long-term evolutions and interactions between rain cells over time. By capturing these temporal dependencies, the model can produce more accurate forecasts. To evaluate the model performance, it is compared against both deep learning and optical flow baselines. This presentation will introduce the model and provide a summary of its performance in spatiotemporal rainfall nowcasting.

Keywords: deep learning; spatiotemporal encoding, rainfall nowcasting; radar; optical flow

How to cite: Abdelhalim, A., Rico-Ramirez, M., Liu, W., and Han, D.: Advancing Spatiotemporal Rainfall Nowcasting through Deep Learning Techniques, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19377, https://doi.org/10.5194/egusphere-egu24-19377, 2024.

At MeteoSwiss, NWP and ML-based models are run operationally on a daily basis to provide weather forecasts and weather warnings for the general public. These forecasts come from various models that differ in lead times, initialization frequency, spatial resolution, and extents. We aim at combining those sources into a probabilistic, gridded weather forecast that is seamless in space and time. Creating a seamless forecast needs careful post-processing so as not to introduce cut-offs or unphysical behavior at the seams between the model runs. This includes using multiple forecast sources and forecast initializations (called lagged ensembles) and combining these using comprehensive blending methods. 

The first minimal viable product of a seamless forecast is currently being produced at MeteoSwiss, and will soon be available to the forecasters in real time. 

We evaluate the merit of these forecasts in terms of warning thresholds for rain and wind gusts. To do so, we compare reforecasts and observations from ground stations as well as rain radar observations from a set of past severe weather events over Switzerland. We benchmark the seamless forecast with individual forecast sources and post-processed products to evaluate the added value of seamlessly combining different forecast sources into one blended product. We furthermore plan to compare different methods for blending between sources soon.

How to cite: Bessenbacher, V., Bhend, J., Beusch, L., Nerini, D., Siegenthaler, C., Spirig, C., and Moret, L.: Evaluation of seamless forecasts for severe weather warnings , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19699, https://doi.org/10.5194/egusphere-egu24-19699, 2024.

Integrating the hybrid and multiscale analyses and the parallel computation is necessary for current data assimilation schemes. A local data assimilation method, Local DA, is designed to fulfill these needs. This algorithm follows the grid-independent framework of the local ensemble transform Kalman filter (LETKF) and is more flexible in hybrid analysis than the LETKF. Local DA employs an explicitly computed background error correlation matrix of model variables mapped to observed grid points/columns. This matrix allows Local DA to calculate static covariance with a preset correlation function. It also allows using the conjugate gradient (CG) method to solve the cost function and allows performing localization in model space, observation space, or both spaces (double-space localization). The Local DA performance is evaluated with a simulated multiscale observation network that includes sounding, wind profiler, precipitable water vapor, and radar observations. In the presence of a small-size time-lagged ensemble, Local DA can produce a small analysis error by combining multiscale hybrid covariance and double-space localization. The multiscale covariance is computed using error samples decomposed into several scales and independently assigning the localization radius for each scale. Multiscale covariance is conducive to error reduction, especially at a small scale. The results further indicate that applying the CG method for each local analysis does not result in a discontinuity issue. The wall clock time of Local DA implemented in parallel is halved as the number of cores doubles, indicating a reasonable parallel computational efficiency of Local DA.

How to cite: Wang, S. and Qiao, X.: A Local Data Assimilation Method (Local DA v1.0) and its Application in a Simulated Typhoon Case, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21770, https://doi.org/10.5194/egusphere-egu24-21770, 2024.

Calibration of convective-scale hourly precipitation based on the frequency-matching method was carried on using CMPASS observation and CMA-MESO 3km forecast data. The character of hourly precipitation bias was studied.The effect of frequency-matching method (FMM) on the bias correction of CMA-MESO 3km hourly precipitation forecasts was analyzed. In the bias characteristic analysis, the differences in precipitation intensity in different regions of the country and the differences in precipitation in different months were considered. The whole country was divided into 7 sub-regions for monthly analysis. In the bias correction based on the frequency-matching method, the daily variations of precipitation bias and the impact of increasing and decreasing precipitation values on the corrected precipitation scores were analyzed. The results show that CMA-MESO 3km forecasts have a wet bias in light rainfall in the cold season, while a dry bias dominates in moderate to heavy rainfall. In the warm season, except for the Tibet region, the hourly precipitation forecast bias of CMA-MESO 3km shows significant daily variations, with more precipitation in the afternoon and less at night and in the morning, especially for heavy rainfall. Therefore, whether to consider the daily variations of precipitation bias in the use of FMM correction mainly reflects in the summer, especially at night and in the morning. Considering the daily variations of precipitation bias is beneficial to improving the forecast skills (TS scores) for nighttime and morning in the summer. Further analysis shows that the positive contribution of FMM correction to forecast scores mainly comes from the increase in frequency adjustment, especially for heavy rainfall. However, for light rainfall with wet bias, FMM often results in negative contribution. Therefore, FMM has a significant improvement effect on heavy rainfall in winter and nighttime rainfall in summer. The reason for this result is that the hit rate of CMA-MESO hourly precipitation forecast is low, and the false alarm rate is generally high, especially for heavy rainfall. In this case, the increased precipitation significantly increases the hit rate, while the false alarm rate increases to a lesser extent, thereby improving the precipitation scores.

How to cite: Qiao, X., Wang, S., and Zeng, M.: Calibration of Convective-scale Hourly Precipitation Based on the Frequency-Matching Method, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-21772, https://doi.org/10.5194/egusphere-egu24-21772, 2024.

The Intraseasonal Oscillation (ISO) is commonly divided into two oscillations: the Madden-Julian Oscillation (MJO), which commonly occurs from November to April in winter, and the Boreal Summer Intraseasonal Oscillation (BSISO), which occurs from May to October. Recent studies have classified these two modes into different types using cluster analysis. Here, we analyze the oceanic and atmospheric variables from the reanalysis ERA5 to determine the influence of MJO and BSISO over the Tropical Americas during the period 1980-2018. We also evaluate how the models of the S2S represent the diverse types of MJO and BSISO by using the Pearson correlation, the root mean square error, and the Brier skill score.

The analysis shows that the four MJO types (slow, fast, stationary, and jumping) exhibit no convective signal over the Tropical Americas and the three BSISO types (canonical, north dipole, and east-expansion) have a strong signal on OLR, winds at 850 and 200 mb over the Tropical Americas. Considering the MJO types, the jumping and slow MJO reveal a small warm pool area, areas where the sea surface temperatures (SSTs) are higher than 28.5°C, over the Mexican Pacific, while the stationary and fast MJOs do not reach such high temperatures. Slow (fast) MJO has strong negative (positive) anomalies in SSTs over the central and Eastern Pacific Ocean. Considering the BSISO types, the canonical BSISO has the strongest westerly burst signal before the initiation of the BSISO events over the Maritime Continent, followed by easterly winds later. In contrast, the east-expansion BSISO shows weaker winds and negative OLR anomalies over Mexico. The northward dipole produces a small warm pool area over the Eastern Pacific Ocean when compared to the canonic and east expansion BSISO.

We conclude that the MJO and BSISO types have different physical mechanisms for modulating the intraseasonal changes in the atmospheric and oceanic variables over the Tropical Americas. We also find that the ECMWF model has the best correlation skill when compared to other models from the S2S project.

How to cite: Lazcano, L. and Dominguez, C.: Subseasonal forecast of the MJO over Tropical America, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-375, https://doi.org/10.5194/egusphere-egu24-375, 2024.

This study focuses on the soil moisture characteristics and its role in supporting the continental tropical convergence zone (CTCZ) during the active phase of the monsoon. Like rainfall, land surface parameters (soil moisture and evaporation) also show intraseasonal oscillation. Furthermore, the sub-seasonal and seasonal features of soil moisture are different from each other. During the summer monsoon season, the maximum soil moisture is found over western coastal regions, central parts of India, and the northeastern Indian subcontinent. However, during active phases of the monsoon (i.e., on sub-seasonal timescales), the maximum positive soil moisture anomaly was found in northern India. Land surface characteristics (soil moisture) also play a pre-conditioning role during active phases of the monsoon over the monsoon core zone of India. When it is further divided into two boxes, the north monsoon core zone and the south monsoon core zone, it is found that the preconditioning depends on that region's soil type and climate classification. Also, we calculate the moist static energy (MSE) budget during the monsoon phases to show how soil moisture feedback affects the boundary layer MSE and rainfall. A similar analysis is applied to the model run, but it cannot show the realistic preconditioning role of soil moisture and its feedback on the rainfall as in observations. We conclude that to get proper feedback between soil moisture and precipitation during the active phase of the monsoon in the model, the pre-conditioning of soil moisture should be realistic.

How to cite: Gautam, P., Chattopadhyay, R., Martin, G., Joseph, S., and Sahai, A. K.: Intraseasonal Oscillation of Land Surface Moisture and  its role in the maintenance of land CTCZ during the active  phases of the Indian Summer Monsoon, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-572, https://doi.org/10.5194/egusphere-egu24-572, 2024.

Drought, an extreme meteorological phenomenon, has significant impacts on a country's social, economic, and environmental stability. Early prediction of drought is crucial to provide warning and preparedness measures. Sub-seasonal prediction, which encompasses a few weeks to a few months ahead, is a critical timescale with limited memory of initial conditions, and not significantly controlled by boundary conditions. Presently, dynamical models have drawn much attention in the sub-seasonal precipitation forecast, however, the accuracy in drought prediction remains low. Currently, various dynamical models such as North American Multi-Model Ensemble (NMME) provide sub-seasonal prediction of hydro-meteorological variables for the entire globe. The efficacy of NMME model output for sub-seasonal drought prediction has not been explored in India. Also, a comprehensive study regarding the inclusion of climate indices as potential predictors for S2S drought prediction is lacking in the literature. We have investigated the potentiality of NMME precipitation output for sub-seasonal drought prediction over India and found out that the NMME model output doesn’t show a reasonable S2S forecast for 3-months standardized precipitation index (SPI3). Further, the study utilized data-driven models such as auto-regression, support vector regression (SVR), XGboost, and recurrent neural network (RNN) with climatic indices and previous month lagged value as predictors to improve the prediction skill. The results show that statistical models are superior to dynamic models. Although the previous monthly data is adequate for lead 1 drought prediction for most of the grids over India, the inclusion of climatic oscillation information was found to be the potential predictor and necessary for higher lead predictions. For example, the western disturbance index helped predict droughts at 2-months lead for the Northwest region of India. Moreover, the wavelet-based post-processing technique has shown the potential to enhance drought predictions significantly. The outcomes of this study will provide an outlook for the sub-seasonal to seasonal drought prediction over India and aid in the improvement of decision-making.

How to cite: Singh, S. and Valiya Veetil, S.: Sub-seasonal to seasonal (S2S) prediction of droughts over India using different data-driven models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-975, https://doi.org/10.5194/egusphere-egu24-975, 2024.

Recent studies suggest that La Niña events can be classified into two categories: mega La Niña and equatorial La Niña. The understanding of the variations in boreal summer intraseasonal oscillation (BSISO) behaviors between such two conditions remains uncertain. Results in this work show during equatorial La Niña summers, in conjunction with the more adequate intraseasonal column-integrated moisture anomalies, the weaker intraseasonal outgoing longwave radiation anomalies are observed over the western North Pacific (WNP) at 3 pentads lag of the peak phase for the Maritime Continent (MC) BSISO events than during mega conditions. Such changes are closely linked with the different propagation features, specifically northwestward and northeastward propagations under mega and equatorial conditions respectively. The distinct propagations under these two conditions could be partly explained by the background column-integrated moisture anomalies. Under equatorial conditions, the less sufficient background moisture anomalies over the tropical western Pacific (WP), in comparison to mega conditions, suppress the activities of the BSISO and its northwestward propagation here. Meanwhile, the enhanced moisture anomalies over the northwestern MC and its surrounding area (NWMC) facilitate the northeastward propagation. Under mega conditions, the background moisture anomalies over the tropical WP are not significant. The southward moisture anomaly gradient over the NWMC hinders the meridional northward propagation and makes some BSISO activities move to the tropical WP region, performing the zonal westward propagation as a whole. The moisture budget and multi-scale interaction diagnoses also emphasize the significant role of the propagation change in the moisture tendency difference averaged over the WNP. Moreover, the extratropical circulation anomalies associated with the MC BSISO events are also discussed. These findings provide new insights into BSISO activity and offer potential improvements for subseasonal forecast.

How to cite: Cao, C. and Wu, Z.: Distinct changes in boreal summer intraseasonal oscillation over the western North Pacific under mega and equatorial La Niña conditions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1244, https://doi.org/10.5194/egusphere-egu24-1244, 2024.

Subseasonal prediction of extremes has emerged as a top forecast priority but remains a great challenge. In this work, we explored two physical modes controlling the subseasonal variation and prediction of land cold extremes over Eurasia: the so-called North Atlantic Oscillation (NAO) and the Eurasian Meridional Dipole mode (EMD). The ECMWF model has shown its skill in predicting the Eurasian land cold extremes 2-4 weeks in advance mainly because of the skillful prediction of NAO and EMD. Further, we separated these observed events into the good prediction and poor prediction groups for those two modes to reveal the potential factors influencing the subseasonal prediction of land cold extremes. It is found that the good prediction group has a stronger initial amplitude and longer persistence, while the poor prediction group has a relatively weaker initial amplitude but rapid intensification. For EMD, the predictability is mainly due to the skillful prediction of the Ural blocking which is further traced back to the stratospheric variations.  

How to cite: Xiang, B.: The window of opportunity for subseasonal land cold extreme prediction over Eurasia  , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1376, https://doi.org/10.5194/egusphere-egu24-1376, 2024.

The hypothesis that predictability depends on the atmospheric state in the planetary-scale low-frequency variability in boreal winter was examined.We first computed six typical weather patterns from 500-hPa geopotential height anomalies in the Northern Hemisphere using self-organizing map (SOM) and k-clustering analysis. Next, using 11 models from the subseasonal-to-seasonal (S2S) operational and reforecast archive, we computed each model’s climatology as a function of lead time to evaluate model bias. Although the forecast bias depends on the model, it is consistently the largest when the forecast begins from the atmospheric state with a blocking-like pattern in the eastern North Pacific. Moreover, the ensemble-forecast spread based on S2S multimodel forecast data was compared with empirically estimated Fokker– Planck equation (FPE) parameters based on reanalysis data. The multimodel mean ensemble-forecast spread was correlated with the diffusion tensor norm; they are large for the cases when the atmospheric state started from a cluster with a blocking-like pattern. As the multimodel mean is expected to substantially reduce model biases and may approximate the predictability inherent in nature, we can summarize that the atmospheric state corresponding to the cluster was less predictable than others.

How to cite: Inatsu, M., Matsueda, M., Nakano, N., and Kawazoe, S.: Prediction Skill and Practical Predictability Depending on the Initial Atmospheric States in S2S Forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1548, https://doi.org/10.5194/egusphere-egu24-1548, 2024.

The diurnal cycle of SST (dSST) is influenced by the development of diurnal warm layers in the upper ocean. Observations show that the dSST rectifies intraseasonal SSTs, potentially leading to changes in intraseasonal weather patterns such as the Madden-Julian Oscillation (MJO). Here we analyze 15-day forecast composites of the coupled ocean-atmosphere and the atmosphere-only configurations of the Numerical Weather Prediction (NWP) models of the UK Met Office to show that a strong dSST in the coupled model leads to a faster MJO propagation compared with the atmosphere-only version of the model. A set of experiments using the coupled model was designed to reduce the strength of the dSST by imposing instant vertical mixing in the top 5 and 10 m of the ocean model. On a 15 lead-day time scale, weakening the dSST slows the MJO phase speed in the coupled model. On a 7 lead-day time scale, all coupled model runs display an underlying 5% increase in the MJO phase speed compared to the atmosphere-only model due to the presence of thermodynamic coupling unrelated to the dSST. The MJO phase speed increase due to the dSST is linearly related to the mean tropical dSST at lead day 1 in the coupled model. An additional 4% of the MJO phase speed increase between the control coupled model and the atmosphere-only model on a 7 lead-day timescale can be attributed to the presence of the dSST in the coupled model. Over 15 lead days, the coupled model produces a two-way feedback between the MJO and the dSST. The MJO conditions set the strength of the dSST in the coupled model. Consistent with observations, the dSST in the coupled model rectifies intraseasonal anomalies of SSTs such that stronger dSST leads to positive intraseasonal SST anomalies. The MJO convection response to these SST anomalies peaks 7 days later, and subsequently feeds back onto SST anomalies. The phase relationship between MJO convection, dSST and intraseasonal SST anomalies is consistent with the relationship between dSST and MJO propagation speed. Overall, our experiments demonstrate the importance of high vertical resolution of the upper ocean in predicting the eastward propagation of the MJO in an NWP setting, potentially creating repercussions for seasonal predictions and climate projections should this feedback be unrepresented in the models.

How to cite: Karlowska, E., Matthews, A., Webber, B., Graham, T., and Xavier, P.: Process-based analysis of the MJO phase speed error in the coupled NWP model of the UK Met Office: a two-way feedback between the MJO and the diurnal warm layers, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1591, https://doi.org/10.5194/egusphere-egu24-1591, 2024.

Cold wave (CW) events over India are usually observed during the boreal winter months, November to February. This study proposes an objective criterion using the actual, departure from normal and the percentile values of the daily gridded minimum temperature (Tmin) data for the monitoring of the CW events over the Indian region and also checks its usefulness in a multi-model ensemble extended range prediction system. The large-scale features associated with these CW events are also discussed.

The CW-prone region has been identified by utilizing this proposed criterion and considering the number of average CW days/year for the entire study period and recent decades. By calculating the standardized area-averaged (over the CW-prone region) Tmin anomalies time series, the CW events are identified from 1951 to 2022. Analyzing the temporal variability of these events, it is seen that there is no compromise in the occurrences of the CW events, even under the general warming scenarios. It is found that the long CW events (>7 days) are favoured by the La-Nina condition, and short CW events (≤7 days) are favoured by the neutral condition in the Pacific. Also, the blocking high to the northwest of Indian longitude with the very slow movement of the westerly trough to the east is found to be associated with the long CW events. In contrast, in the case of short events, the blocking high is not so significant. The multi-model ensemble prediction system is found to be reasonably skilful in predicting the CW events over the CW-prone region up to 2-3 weeks in advance with decreasing confidence in longer leads. Based on the forecast verifications, it is noticed that this forecasting system has a remarkable strength to provide an overall indication about the forthcoming CW events with sufficient lead time despite its uncertainties in space and time. 

How to cite: Mandal, R., Joseph, S., Sahai, A. K., Dey, A., Krishna, P. M., Pattanaik, D., Kaur, M., and Karmakar, N.: Real-time subseasonal prediction of cold waves over India, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1706, https://doi.org/10.5194/egusphere-egu24-1706, 2024.

Subseasonal predictions with a time scale of 2-4 weeks, which fills the gap between the weather and seasonal forecasts, are limited by the uncertainties arising from the initial conditions as well as the model physics. Therefore, to develop an efficient subseasonal prediction system, both these uncertainties need to be addressed. With this background, a multi-physics multi-ensemble approach has been adopted to develop a competent second-generation subseasonal prediction system at the Indian Institute of Tropical Meteorology (IITM), Pune, India. The first-generation prediction system developed at IITM is run operationally at the India Meteorological Department and has useful skills for up to two weeks.

A combination of physics perturbations and initial condition perturbations with a total of 18 ensemble members is present in the system. This system has been experimentally run since May 2022. The hindcast runs during 2003-2018 are also made on-the-fly. The initial results indicate a considerable improvement in the forecast skill compared to its predecessor and have reasonable deterministic prediction skill for up to three weeks. The system could provide skilful prediction of the subseasonal variations during the two contrasting monsoon seasons 2022 (above normal) and 2023 (below normal) 2-3 weeks in advance.

How to cite: Joseph, S., Dey, A., Mandal, R., Kalshetti, M., Phani, R., Waje, S., and Sahai, A.: Development of a Multi-physics Multi-ensemble Subseasonal Prediction System and its Real-time Performance during Contrasting Indian monsoons, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2747, https://doi.org/10.5194/egusphere-egu24-2747, 2024.

This study investigates the influence of the boreal summer intraseasonal oscillation (BSISO) on 10-30-day summer rainfall anomalies in Southwestern China (SWC) under the effects of Qinghai-Tibetan Plateau monsoon (QTPM) based on ERA5 reanalysis data and CN05.1 precipitation in 1981-2018. The results show that the 10-30-day rainfall anomalies in SWC have significant and joint feedback to variation of the second component of BSISO (BSISO2) and QTPM at lagging strong (weak) BSISO events by 0-12 days. Their lagged causal linkage and corresponding physical processes have been revealed by causal effect networks and composite analyses, which are most significant at 4-day and 12-day lag. Simultaneously, BSISO2 can induce wetter 10-30-day rainfall over southern SWC by motivating water vapor transport from the Bay of Bengal towards Yunnan province. More importantly, BSISO2 can modulate a northwest-propagating wave train from the western north Pacific towards SWC at the upper troposphere by vertical wave energy transport, which blocks the wave train propagating from the Lake Balkhash to east China–Japan most significantly at a 4-day lag and leads to drier eastern SWC. The process can be influenced by QTPM significantly which leads to the response of 10-30-day rainfall over SWC with lags of 0-12 days. Specifically, same-phase QTPM can trigger more active wave train propagation from high-latitude while opposite-phase QTPM enhances the low-latitude wave energy transport. The interference then facilitates baroclinic structure over eastern SWC at lagging 12 days with positive precipitation anomalies for same-phase events and negative precipitation for opposite-phase events.

How to cite: Yang, L., Chen, H., and Wang, S.: The joint effects of the boreal summer intraseasonal oscillation and Qinghai-Tibetan Plateau monsoon on the precipitation over Southwestern China , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2784, https://doi.org/10.5194/egusphere-egu24-2784, 2024.

The Madden-Julian Oscillation (MJO) is the dominant intraseasonal variability of eastward propagating atmospheric disturbances in the tropics. From its vast impacts on the sub-seasonal extreme events and predictability, the mean states controlling the MJO activity have been investigated. For example, the robust relationship between the Quasi-Biennial Oscillation (QBO) and the MJO has been suggested in the past several years. In the easterly QBO winters, the MJO exhibits stronger activity than the westerly QBO winters. 
Our study suggests another crucial factor that affects the MJO: a meridional humidity gradient of the atmospheric column in the vicinity of the Maritime Continent. With the change in the shape of the column humidity distribution, MJO variance shows a robust interannual modulation regardless of the QBO. The northward (southward) extension of the moisture increases (decreases) the mean state meridional humidity gradient, which leads to MJO development (decay) over the MC with increasing (decreasing) horizontal moisture advection. This robust relationship between mean state humidity and MJO activity is investigated in the CMIP6 models as two aspects: i) interannual variation of MJO and ii) future change in MJO. Both simulated MJO activities are largely affected by the mean state MHG, supporting the robust role of mean state moisture on the MJO shown in the observations. The results of this study provide a further understanding of seasonal MJO activity and sub-seasonal predictability.MJO activity and sub-seasonal predictability.

How to cite: Kang, D., Kim, D., and Kang, S.-Y.: Robust Relationship between Mean State Moisture and Interannual MJO Activity in Observations and CMIP6 Models, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2862, https://doi.org/10.5194/egusphere-egu24-2862, 2024.

Utilizing the Community Atmospheric Model version 4, the influence of Arctic sea ice concentration (SIC) on the predictability of the Ural Blocking (UB) in subseasonal timescale is investigated. Taking the zonal flows as the reference states, the optimal Arctic SIC perturbations that trigger zonal flows into UB events on subseasonal timescale are obtained with the conditional nonlinear optimal perturbation (CNOP) approach. The numerical results show that the Arctic SIC decline in the Greenland, Barents and Okhotsk Seas can trigger zonal flows into UB events on a timescale of four pentads (20 days). Further diagnosis shows that the SIC decline in these regions locally warms the low troposphere via diabatic processes in the first pentad. Then, dynamic processes, such as temperature advection, modulate the temperature in the middle troposphere and weaken the meridional temperature gradient between the Arctic and mid-latitudes upstream of the Ural sector. The weakened meridional temperature gradient further decelerates the background zonal flow near the Ural sector and triggers UB formation in four pentads. After that, the optimal Arctic SIC perturbations that have great influences on subseasonal UB predictions are also obtained with CNOP approach. It is found that SIC increase in the Greenland Sea, Barents Sea, and Okhotsk Sea would weaken the UB intensity while SIC decline in these regions would strengthen it. Further diagnoses show that the physical mechanisms are similar to those triggering UB formation. Moreover, utilizing the observing system simulation experiments, it is shown that targeted observations in the Barents Sea, Greenland Sea, and Okhotsk Sea can remarkably improve the prediction skills of UB in the fourth pentad. Numerical results show that targeted observations have a positive effect on 75% of 160 experiment members, reduce 35% forecast errors of the fourth pentad mean blocking index, and perform even better when the original forecast errors are greater. Further diagnosis shows that the improvement is related to the well-described westerly winds in the Ural region and its adjacent regions, corresponding to the more skillful predictions of blocking circulations. The above results supply a theoretical base for the design of Arctic SIC observations and more skillful subseasonal predictions for mid-latitude extreme weather.

How to cite: Dai, G., Mu, M., Ma, X., and Gao, Y.: Influence of Arctic sea ice concentration on extreme Ural blocking predictability in subseasonal timescales, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2890, https://doi.org/10.5194/egusphere-egu24-2890, 2024.

The precipitation over the eastern Tibetan Plateau (ETP, here defined as 29°–38°N, 91°–103°E) usually exhibits significant subseasonal variation during boreal summer. As the hot spot of land-air interaction, the influences of ETP surface soil temperature (T_soil) on the local precipitation through subseasonal land-air interaction are still unclear but urgently needed for improving subseasonal prediction. Based on station and reanalysis datasets of 1979–2018, this study identifies the evident quasi-biweekly (QBW) (9–30 days) periodic signal of ETP surface T_soilvariation during the early summer (May–June), which results from the anomalies of southeastward propagating mid-latitude QBW waves in the mid-to-upper troposphere. The observational results further show that the maximum positive anomaly of precipitation over the ETP lags the warmest surface T_soil by one phase at the QBW timescale, indicating that the warming surface T_soil could enhance the subseasonal precipitation. The numerical experiments using the WRF model further demonstrate the effect of warming surface T_soil  on enhancing the local cyclonic and precipitation anomaly through increasing upward sensible heat flux, the ascending motion, and water vapor convergence at the QBW timescale. In contrast, the effect of soil moisture over the ETP is much weaker than T_soil  at the subseasonal timescale. This study confirms the importance of surface T_soil over the ETP in regulating the precipitation intensity, which suggests better simulating the land thermal feedback is crucial for improving the subseasonal prediction.

How to cite: Qi, X., Yang, J., Xue, Y., Bao, Q., Wu, G., and Ji, D.: Subseasonal Warming of Surface Soil Enhances Precipitation Over the Eastern Tibetan Plateau in Early Summer, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3148, https://doi.org/10.5194/egusphere-egu24-3148, 2024.

Global warming is accelerating drought onset, causing more frequent flash drought events. These events occur at the subseasonal timescale in which rapid decreases in root-zone soil moisture (RZSM) increase risks of crop failure, wildfire, and heat stress globally. However, forecasting soil moisture and flash droughts at lead times beyond 2 weeks remains a significant challenge. Recently, machine learning methods with historical reanalysis data have shown improved forecast accuracy compared to state-of-the-art numerical weather prediction methods, but they can only produce skillful forecast within 10 days. Here we show that a convergence forecast model combining a deep learning approach with subseasonal retrospective forecasts (reforecast) from numerical models produces skillful subseasonal soil moisture and flash drought forecasts at lead times beyond 2 weeks. We train a deep learning architecture on combinations of reanalysis and reforecast from 2000 to 2015 and validate results during the testing period from 2018 to 2019. The subseasonal forecast skill of soil moisture of the convergence forecast model is much higher than those of current state-of-the-art numerical forecast models, deep learning bias corrected numerical forecast models, or the reanalysis-based deep learning models, which showed no skill after 2 weeks lead time. The convergence model also showed significantly improved performance for predicting flash droughts compared to the original or deep learning bias corrected numerical forecast models or reanalysis-based deep learning models.  A permutation analysis indicates that reanalysis precursors and soil moisture reforecast at lead times within 2 weeks both contribute significantly to the forecast skill at longer lead times. The convergence forecast model provides accurate and efficient subseasonal soil moisture and flash drought forecasting and is promising for accurately forecasting key variables and extreme events at the subseasonal timescale.

How to cite: Lesinger, K. and Tian, D.: Converging Deep Learning and Numerical Prediction for Skillful Subseasonal Soil Moisture and Flash Drought Forecasting, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3194, https://doi.org/10.5194/egusphere-egu24-3194, 2024.

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 in a changing climate. On the subseasonal-to-seasonal (S2S) forecasting timescale, skillful forecasts of hydro-meteorological hazards combined with targeted actions can prevent various vegetation damage and large-scale impacts (e.g. agriculture and food security, wildfire risk management, forest management,  biodiversity and flora protection,etc.).

We here focus on forest damage in Europe, defined as negative anomalies of the normalized difference vegetation index (NDVI). Compound drought and heat wave events are known to trigger low NDVI events in summer. A dry summer combined with warm and moist conditions during the previous winter can also have a negative impact. However, to our knowledge, there exists no comprehensive study of hydro-meteorological drivers triggering forest damage in Europe. Hence, the goal of our study is a) finding the optimal variables to predict summer forest damage in Europe, and b) assessing the S2S forecast skill of these variables. We develop an automated procedure to systematically identify hydro-meteorological conditions leading to forest damage, up to 18 months prior to occurrence. We train a model using AVHRR remote sensing observation of NDVI for the impact data, and ERA5 and ERA5-Land reanalysis datasets for the explicative variables. These variables include temperature, precipitation, dew point temperature, surface latent heat flux, soil moisture, and soil temperature. To bridge the research gap between the S2S forecasts of hydrometeorological variables and vegetation damage, we assess the forecast skill of variables from the S2S hindcast database of ECMWF identified as responsible for low NDVI events. The idea is to determine to what extent S2S models can predict conditions triggering forest damage, by identifying the sources of predictability or potential need for improvement.

How to cite: Rivoire, P., Dupuis, S., Guisan, A., and Vittoz, P.: Predicting Forest Damage in Europe: A Subseasonal-to-Seasonal Forecasting Approach for Hydro-meteorological Drivers, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3242, https://doi.org/10.5194/egusphere-egu24-3242, 2024.

Despite current global warming due to increasing greenhouse gases, severe cold winters have devastated the East Asia in recent decades. Efforts are being made to predict cold events using dynamic models and physically-based statistical models. In this study, we explore the potential predictability of the East Asian winter surface temperature by establishing a multiple linear regression model based on three precursors of time-evolved preconditions: 1) autumn Arctic sea-ice loss, 2) northern Eurasian sea level pressure pattern, and 3) the El Niño-Southern Oscillation (ENSO). Reduced autumn Arctic sea-ice was favorable for extreme cold events in the East Asia. Furthermore, the autumn Arctic sea-ice loss was accompanied by cyclonic circulations over northern Eurasia in November, which could have led to cold anomalies over the East Asia in the late winter. The preconditioning deep convection in La Niña events is a well-known indicator of exerted atmospheric wave propagation, resulting in cold winters over the East Asia. We suggested here that by combining Arctic sea-ice, atmospheric circulations, and ENSO, the predictability of East Asian winter surface temperature variability could be improved.

How to cite: Jang, Y.-S., Lim, H.-G., Jun, S.-Y., and Kug, J.-S.: Arctic sea ice loss and La Niña as precursors of extreme East Asian cold winters, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3737, https://doi.org/10.5194/egusphere-egu24-3737, 2024.

The predictability of certain extreme weather events can exceed the traditional two weeks by considering the boundary conditions. Targeted observations in sensitive areas on Arctic sea ice concentration (SIC) can improve the extended-range (4 pentads) forecast skills of long-lasting and strong Ural blocking (UB). The sensitive areas are determined based on the SIC optimally growing boundary errors, obtained by the conditional nonlinear optimal perturbation method. The sensitive areas are mainly located in the Barents Sea, Greenland Sea, and Okhotsk Sea. The results of observing system simulation experiments for 8 UB cases indicate that the targeted observations can remarkably improve the prediction skills of UB in the 4th pentad. Targeted observations have a positive effect on 75% of 160 experiment members, reduce 35% forecast errors of the 4th pentad mean blocking index, and perform even better when the original forecast errors are greater. Further diagnosis shows that targeted observations contribute to more accurate SIC boundary conditions in the Barents Sea, Greenland Sea, and Okhotsk Sea and reduce temperature errors in the lower and middle troposphere. It further results in well-described westerly winds in the Ural region and its adjacent regions, corresponding to the more skillful predictions of blocking circulations. The above results supply a theoretical base for the design of Arctic SIC observations and more skillful extended-range predictions for mid-latitude extreme weather.

How to cite: Gao, Y., Mu, M., and Dai, G.: Targeted Observations on Arctic Sea Ice Concentration for Improving Extended-range Prediction of Ural Blocking, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3796, https://doi.org/10.5194/egusphere-egu24-3796, 2024.

The summer northern annular mode (NAM) variability plays a crucial role in the summer climate variability and extremes of the Northern Hemisphere. In this study, we report a significant negative correlation between the March NAM and summer NAM during 1979–2022 and reveal the role of the spring stratosphere in this seasonal linkage. Particularly, it is found that the negative phase of March NAM features a strong meridional shear in the extended-North-Atlantic jet, which tends to generate planetary scale Rossby waves that propagate upward and poleward into the stratosphere. This increased stratospheric planetary wave activity in March transitions to weakened wave activity in May, leading to positive zonal wind anomalies in the polar stratosphere in May, extending downward to the troposphere in June and promoting the formation and persistence of positive summer NAM. The results provide both statistical and dynamical evidence for the role of the spring stratosphere in connecting the spring and summer circulation. 

How to cite: Xu, X., Wang, L., Wang, T., and Chen, G.: The role of stratospheric processes in the trans-seasonal connection between spring and summer northern annular modes, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3798, https://doi.org/10.5194/egusphere-egu24-3798, 2024.

The main objective of this study is to assess typhoon precipitation forecast skill on the subseasonal timescale. The 20-year reforecasts from the ECMWF 46-day ensemble (ENS) are utilized to compare with gridded surface observations in Taiwan. The analysis focuses on the dates when typhoons affect Taiwan (117-129°E and 19-28°N). 15 ENS grids around Taiwan area are used with the grid size of 0.8 x 0.8 degree. Historical rainfall observations are provided by the Central Weather Administration (CWA), which the observations from the surface stations are interpolated into a resolution of 1km x 1km grid box. A comparison between the ENS forecast data and gridded CWA rainfall observations is performed by searching the optimal percentile rank (PR) of gridded CWA rainfall that has the smallest mean difference against the ENS data. The result reveals that the ENS can somewhat capture the rainfall contrast between the mountainous area and plain area, despite its relatively lower horizontal resolution. However, the difference between ENS rainfall forecasts and surface observations significantly increases for the forecasts beyond 72 hours, due to the model's coarser resolution and typhoon track forecast errors.

The ENS typhoon track forecast errors in weeks 1-4 are analyzed by comparing the ensemble vortex tracks with the JTWC best tracks. The track forecast error is decomposed into the along-track (AT) and cross-track (CT) components. The analysis result shows negative mean AT errors, indicating slower translation speed biases in the model. The mean AT errors could reach up to 400 km for the 168 h forecasts after TC formations.

Given the significant typhoon track forecast errors, using the raw ENS rainfall forecasts for the operational TC forecasting/outlook become challenging. In response, we have developed a statistical Quantitative Precipitation Forecast (QPF) model to predict typhoon rainfall, considering the track biases in the ENS forecasts. The forecast tools developed in this study will be integrated into CWA’s subseasonal typhoon forecast system to support water resources management and disaster risk reduction.

How to cite: Hsu, H.-Y. and Tsai, H.-C.: Subseasonal Typhoon Precipitation Forecast in Taiwan Area Using the ECMWF Reforecasts: Forecast Verification and Application, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4208, https://doi.org/10.5194/egusphere-egu24-4208, 2024.

Land surface processes are strongly associated with heat waves (HWs). However, how the uncertainties in land surface processes owing to inaccurate physical parameters influence subseasonal HW predictions has rarely been explored. To examine the impact of parameter errors of land surface processes on the uncertainty of subseasonal HW predictions, five strong and long-lasting HW events over the middle and lower reaches of the Yangtze River (MLYR) are investigated. Based on the Weather Research and Forecasting (WRF) model, the conditional nonlinear optimal perturbation related to parameters (CNOP-P) approach is employed to address the aforementioned issues.

Numerical results demonstrate that the CNOP-P type errors of physical parameters cause large prediction errors for five HW event onsets. Two types of CNOP-Ps are obtained for HW events, called the type-1 CNOP-P and the type-2 CNOP-P. The type-1 (type-2) CNOP-P causes an approximately 3 °C (2 °C) warm (cold) bias during the HW period. Surface sensible and latent heat flux errors, especially flux exchange between vegetation canopy and canopy air, provide considerable uncertainty in subseasonal HW predictions. The type-1 (type-2) CNOP-P exhibits an underestimation (overestimation) of transpiration. Furthermore, it should be noted that the type-1 CNOP-P results in a substantial difference in soil moisture, a phenomenon that is demonstrated to be challenging to observe in the type-2 CNOP-P. The results indicate that understanding vegetation-atmosphere dynamics is crucial for improving subseasonal HW predictions. Jointly lowering soil-atmosphere and vegetation-atmosphere uncertainty can notably improve subseasonal HW prediction skills.

How to cite: Zhang, Q., Mu, M., Sun, G., and Dai, G.: Impact of Uncertainties in Land Surface Processes on Subseasonal Predictability of Heat Waves Onset Over the Yangtze River Valley, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4245, https://doi.org/10.5194/egusphere-egu24-4245, 2024.

This study uses the ECMWF 46-day ensemble to evaluate the subseasonal forecasts of tropical cyclones (TCs) in the western North Pacific, including TC formations, tracks, intensity, and precipitation forecasts. TC formations and the subsequent tracks are objectively detected in both real-time forecasts and also the 20-year ECMWF reforecasts. Additionally, a spatial-temporal track clustering technique is utilized to group similar vortex tracks in the 101-member real-time forecasts for operational application. The forecast verification focuses on evaluating the influence of large-scale environmental factors on TC forecast skills during weeks 1-4, such as the Western North Pacific Summer Monsoon (WNPSM), Madden Julian Oscillation (MJO), and Boreal Summer Intraseasonal Oscillation (BSISO). The Precision-Recall (PR) curve is used to represent the imbalanced TC data instead of the Receiver Operating Characteristic (ROC) curve. Better TC forecast skills are observed if model initialized on MJO Phases 6 and 7 for the week-1 forecasts, and on MJO Phases 4 and 5 for the weeks 2 and 3 forecasts. Also, TC forecast skills are better if the cumulative percentage of the WNPSM index (Wang et al. 2001) is larger than 60%. This study also investigats the TC precipitation forecast skill around Taiwan area.

The evaluation results obtained from this study has been integrated into the TC Tracker 2.0 system developed by Central Weather Administration (CWA). The system can generate a "Subseasonal TC Threat Potential Forecast" product to assist in disaster mitigation and water resources management for the Water Resources Agency. More details about the subseasonal TC forecast verifications and applications will be presented in the meeting

How to cite: Tsai, H.-C., Hsu, H.-Y., Lo, T.-T., and Chen, M.-S.: Verifications of Week-1 to Week-4 Tropical Cyclone Forecasts in the Western North Pacific from the ECMWF 46-Day Ensemble, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4272, https://doi.org/10.5194/egusphere-egu24-4272, 2024.

The real-time WWRP/WCRP Subseasonal to Seasonal (S2S) Prediction Project Phase 2 database was used to evaluate the prediction skill of tropical cyclone from eleven forecasting systems for the North Western Pacific. The variable introduced to evaluate S2S tropical cyclone prediction is daily tropical cyclone probability, which is the occurrence probability of tropical cyclone within 500 km in one day. Using such a definition, the occurrence of tropical cyclone is a dichotomous event. The skill of S2S tropical cyclone prediction can be evaluated using debiased Brier Skill Score, which is the traditional Brier Skill Score with impact of forecast ensemble size removed. Sensitivity tests were conducted to analyze the influence of difference in temporal window and radius in the definition of daily tropical cyclone probability. It is demonstrated that though the daily tropical cyclone probability would vary with a changed radius and temporal window, the debiased Briere Skill Score does not change much since it is related with the ratio of mean error of model forecast and the mean error of a reference climatological forecast. The robustness of the prediction skill indicates the suitability of using the daily tropical cyclone probability and debiased Brier Skill Score to measure tropical cyclone prediction skill at S2S timescale. Compared with the prediction skill of the S2S Prediction Project Phase 1, the real-time S2S tropical cyclone prediction is improved for some forecast systems. Some early results by combining multi-model tropical cyclone forecasts to improve tropical cyclone prediction will also be presented.

How to cite: Wang, X. and Vitart, F.: Evaluating Real-time Subseasonal to Seasonal Tropical Cyclone Prediction, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4665, https://doi.org/10.5194/egusphere-egu24-4665, 2024.

Considering the significant differences in the rainfall characteristics over East Asia between the early [May–June (MJ)] and late [July–August (JA)] summer, this study investigates the subseasonal predictability of the rainfall over East Asia in early and late summer, respectively. Distinctions are obvious for both the spatial distribution of the prediction skill and the most predictable patterns, that is, the leading pattern of the average predictable time (APT1) between the MJ and JA rainfall. Further analysis found that the distinct APT1s of MJ and JA rainfall are attributable to their different predictability sources. The predictability of the MJ rainfall APT1 is mainly from the boreal intraseasonal oscillation signal, whereas that of the JA rainfall APT1 is provided by the Pacific–Japan teleconnection pattern. This study sheds light on the temporal variation of predictability sources of summer precipitation over East Asia, offering a possibility to improve the summer precipitation prediction skill over East Asia through separate predictions for early and late summer, respectively.

How to cite: Li, X.: Subseasonal Predictability of Early and Late Summer Rainfall Over East Asia, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4955, https://doi.org/10.5194/egusphere-egu24-4955, 2024.

Summer monsoon precipitation over the Bay of Bengal (BoB) has pronounced intraseasonal variability (ISV), which has a close relationship to the local intraseasonal sea surface temperature (SST). Before heavy precipitation, intraseasonal SST in the BoB often has a warm anomaly and propagates northward, which drives the atmosphere and tends to trigger the convection. Besides the local air-sea interaction, the ISV of SST in the Arabian Sea (AS) also has an effect on the precipitation over the BoB. Results show that a prominent heavy precipitation usually occurs when the warm intraseasonal SST anomaly appears early in the AS and moves northward prior to that emerges in the BoB. The warm SST anomaly in the AS affects the sea level pressure and then trigger a southwestly wind anomaly in the center of AS. This wind anomaly promotes the wind convergence moving northward from the southern tip of Indian peninsula to the north India and northern BoB, which directly influence the vertical moisture advection and finally the precipitation. Understanding this process will be helpful to improve the predictive skill of the ISVs during the Indian Summer Monsoon.

How to cite: Xi, J.: Influence of Intraseasonal Variability of Sea Surface Temperature in the Arabian Sea on the Summer Monsoon Precipitation Over the Bay of Bengal, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5422, https://doi.org/10.5194/egusphere-egu24-5422, 2024.

The extratropical response to the Madden-Julian Oscillation (MJO) is modulated by two prominent modes of low-frequency sea surface temperature (SST) variability: the Atlantic Multidecadal Variability (AMV) and the Pacific Decadal Oscillation (PDO). Utilizing the UK Earth System Model (UKESM) 1100 year pre-industrial control simulation from CMIP6, this study offers a unique opportunity to explore decadal variability with an extensive dataset, surpassing the limitations of previous studies which focussed on reanalysis products.

The results underscore a statistically significant influence of both AMV and PDO on the extratropical response across all MJO phases. Non-linear interactions between the MJO teleconnection and SST forcing are observed prominently in the modification of the response to MJO phase 6 (enhanced convection over the western Pacific), with AMV+ and PDO+ background states amplifying distinct teleconnection patterns, notably the negative North Atlantic Oscillation (NAO-) and the deepened Aleutian Low responses, respectively. These changes are greater in magnitude than would be expected from the linear superposition of the individual atmospheric responses to the SST mode and the MJO. The amplification of the MJO phase 6 teleconnection to the North Atlantic aligns with prior research based on ERA5 reanalysis data.

While modulation of the response to MJO phase 3 (enhanced convection over the eastern Indian Ocean) is evident, it is less pronounced compared to phase 6, and the mechanisms via which it acts are less clear. Intriguingly, alterations in the teleconnection, such as a weaker Aleutian Low during PDO+, contradict the anticipated modulation. Since MJO phase 3 and PDO+ tend to weaken and strengthen the Aleutian Low, respectively, it would be reasonable to expect that these effects would cancel. Instead, the weakening of the Low after MJO phase 3 is increased during PDO+.

A possible mechanism for the modulation of the teleconnections is a linear superposition of Rossby wave modes excited by the MJO, contingent upon the SST state. In the case of MJO phase 6, this corresponds to an amplification of the existing modes, and hence of the expected response. For MJO phase 3, however, there is an indication that other Rossby wave modes may also be excited in certain SST states, leading to interference which is out of phase with the primary response.

Acknowledging the limitations of observational and reanalysis datasets, this study underscores the pivotal role of climate models in the effective study of decadal and multi-decadal variability. Importantly, the study has significant implications for extratropical forecasting over the coming decades. The modulation of the MJO teleconnection by AMV and PDO suggests modifications in predictability, crucial for refining forecasting techniques. Furthermore, these results provide a contextual foundation for studies examining MJO teleconnections in future climates, enabling a more accurate dissection of responses influenced by internal and anthropogenically forced variability.

How to cite: Matthews, A., Skinner, D., and Stevens, D.: Decadal variability of the extratopical response to the MJO: AMV and PDO modulation in the UKESM climate model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6229, https://doi.org/10.5194/egusphere-egu24-6229, 2024.

The Navy Earth System Prediction Capability (ESPC) is the Navy’s coupled ocean-atmosphere-sea ice model.  The current version of the Navy ESPC has 16 ensemble members and been operational since August 2020. The Navy ESPC has known biases in Madden-Julian Oscillation (MJO), which has a too strong amplitude and too fast propagation speed. During boreal winter, the MJO in the Navy ESPC is too strong due to biases in the vertical motion, which supports larger vertical moisture advection.  The MJO is too strong in this season due to excessive evaporation in the western Pacific supporting moistening to the east of the MJO convective center.  In this study, we examine the boreal winter MJO in the operational Navy ESPC ensemble.  We use process oriented diagnostics to explore the local and remote sources of biases that drive good and poor MJO forecasts. 

MJO forecasts are split into those that are well predicted and those that are poorly predicted.  Individual MJO events are tracked following Chikira (2014), using Hovmöllers of MJO filtered OLR averaged between 10N and 10S.  The MJO forecast performance is determined by comparing the forecasted MJO to the observed MJO based on the magnitude of the maximum amplitude of the MJO, the phase speed, duration of the event, and the location of the MJO convection.  Using the moisture mode framework, we examine the maintenance and propagation of moisture anomalies to identify how the local and remote sources of error affect MJO skill.  We use a moisture budget analysis to diagnose and understand the difference between the forecasts that performed well and those that performed poorly.  Additionally, we examine the effects that these forecast errors in the MJO have on extratropical cyclones, surface winds, and clouds in the Navy ESPC and how biases in the extratropics affect the skill of MJO-teleconnections.

How to cite: Rushley, S., Janiga, M., and Reynolds, C.: Local and remote sources of error inMJO forecasts in the Navy ESPC , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6452, https://doi.org/10.5194/egusphere-egu24-6452, 2024.

One main justification for subseasonal-to-seasonal (S2S) prediction is its identified sources of predictability. These sources include slowly varying phenomena, such as the MJO, stratospheric conditions, upper-ocean heat content, soil moisture, and sea ice. In practice, however, these presumed sources of S2S predictability have become the main targets of S2S prediction. For example, predicting the MJO, especially its propagation over the Indo-Pacific Maritime Continent, has been challenging. This raises a fundamental question: What are the predictability sources of the MJO? For global coupled prediction models, the primary sources of predictability are initial conditions and the governing laws. It is unclear, however, what elements in the initial conditions are more important to MJO prediction than others. It can be argued that the current practice of initializing forecasts using a single state of the system may not be optimal. Embedded initial conditions may provide an additional source of predictability that has yet to be fully explored.

How to cite: Zhang, C.: Sources of S2S and MJO predictability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6688, https://doi.org/10.5194/egusphere-egu24-6688, 2024.

Recently, it has been noticed that weather and climate changes over the Arctic and mid-latitude regions may have influenced the particulate matter concentrations and haze over East Asia. Among the various weather and climate conditions and climate indices could be an important factor in affecting variation of particulate matter (PM) concentrations. In this study, we examined the long-term changes in the sea ice cover, soil moisture, near-surface temperature and its link with the lower atmospheric circulation over Arctic and mid-latitude from 1950 to 2022, using modern reanalysis datasets. Long-term analyses show negative trends in sea ice cover over the Arctic and positive trends in near-surface temperature and SST, implying atmospheric stagnant and variation of PM concentration. Additionally, climate indices, related to teleconnection between the Arctic region and mid-latitude, co-related with understanding air quality. Based on climate indices, we have developed the air quality prediction model for reflecting variations in weather and climate conditions. Therefore, the findings in this study can likely be used for actual prediction systems based on long-term weather measurement datasets over the Arctic region.

Acknowledgment: This research was supported by a National Research Foundation of Korea Grant from the Korean Government (MSIT ; the Ministry of Science and ICT) (NRF- 2023M1A5A1090715).

How to cite: Park, J.-M., Lee, D., Kim, K., Kim, S., Lee, G., and Lee, K. H.: Weather and Climate conditions over the Arctic and mid-latitude regions affecting air quality, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7386, https://doi.org/10.5194/egusphere-egu24-7386, 2024.

Accurate soil moisture initial conditions in dynamical subseasonal forecast systems are known to improve the temperature forecast skill regionally, through more realistic water and energy fluxes at the land-atmosphere interface. Recently, results from the GEWEX-GASS LS4P (Impact of initialized land temperature and snowpack on sub-seasonal to seasonal prediction) multi-model coordinated experiment have provided evidence of the primal contribution of the initial surface and subsurface soil temperature over the Tibetan Plateau for capturing a hemispheric scale atmopsheric teleconnection leading to improved subseasonal forecasts. Yet, both the soil temperature and water content are key components of the soil enthalpy and we hypothesize that properly initializing one of them without modifying the other in a consistent manner can alter the soil thermal equilibrium, thereby potentially reducing the benefit of land initial conditions on subsequent atmospheric forecasts. This study builds on the protocol of the above-mentioned multi-model experiment, by testing different land initialization strategies in an Earth system model. Results of this pilot study suggest that a better mass and energy balance in land initial conditions of the Tibetan Plateau triggers a wave train which propagates through the northern hemisphere mid-latitudes, resulting in an improved large scale circulation and temperature anomalies over multiple regions of the globe. While this study is based on a single case, it strongly advocates for enhanced attention towards preserving the soil energy equilibrium at initialization to make the most of land as a driver of atmospheric extended-range predictability.

How to cite: Ardilouze, C. and Boone, A.: Soil enthalpy: an unheeded source of subseasonal predictability?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7705, https://doi.org/10.5194/egusphere-egu24-7705, 2024.

Subseasonal prediction fills the gap between weather forecasts and seasonal outlooks. There is evidence that predictability on subseasonal timescales comes from a combination of atmosphere, land, and ocean initial conditions. Predictability from the land is often attributed to slowly varying changes in soil moisture and snowpack, while predictability from the ocean is attributed to sources such as the El Niño Southern Oscillation. Here we use a unique set of subseasonal reforecast experiments with CESM2 to quantify the respective roles of atmosphere, land, and ocean initial conditions on subseasonal prediction skill over land. These reveal that the majority of prediction skill for global surface temperature in weeks 3-4 comes from the atmosphere, while ocean initial conditions become important after week 4, especially in the Tropics. In the CESM2 subseasonal prediction system, the land initial state does not contribute to surface temperature prediction skill in weeks 3-6 and climatological land conditions lead to higher skill, disagreeing with our current understanding. However, land-atmosphere coupling is important in week 1. Subseasonal precipitation prediction skill also comes primarily from the atmospheric initial condition, except for the Tropics, where after week 4 the ocean state is more important.

How to cite: Richter, J., Glanville, A., King, T., Kumar, S., Yeager, S., Duan, Y., Fowler, M., Jaye, A., Edwards, J., Caron, J., Dirmeyer, P., Danabasoglu, G., and Oleson, K.: Quantifying sources of subseasonal prediction skill in CESM2, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8357, https://doi.org/10.5194/egusphere-egu24-8357, 2024.

For many practical applications, e.g. agricultural planning, skillful weather predictions on the subseasonal timescale (2-4 weeks in advance) are key for making sensible decisions. Since traditional numerical weather prediction (NWP) models are often not capable of delivering such forecasts, we use an alternative forecasting approach combining both, physical knowledge and statistical models. Selected meteorological variables from ERA-5 reanalysis data are used as predictors for wintertime Central European mean 2-meter temperature and the occurrence of cold wave days at lead times of 14, 21 and 28 days. The forecasts are created by Quantile Regression Forests in case of continuous temperature values and Random Forest Classifiers in case of binary occurrence of cold wave days. Both model types are evaluated for the winters 2000/2001 to 2019/2020 using the Continuous Ranked Probability Skill Score for the continuous forecasts and the Brier Skill Score for the binary forecasts. As a benchmark model, a climatological ensemble obtained from E-OBS observational data is considered. We find that the used machine learning models are able to produce skillful weather forecasts on all tested lead times. As expected, the skill depends on the exact winter to be forecasted and generally decreases for longer lead times but is still achieved for individual winters and in the 20-winter mean at 28 days lead time. Since machine learning models are often subject to a lack of interpretability and thus considered to be less trustworthy, we apply Shapley Additive Explanations to gain insight into the most relevant predictors of the models’ predictions. The results suggest that both Random-Forest based models are capable of learning physically known relationships in the data. This is, besides the capability of producing skillful forecasts on the subseasonal timescale, a selling point of the combination of physical knowledge and statistical models. Finally, we compare the skill of our statistical models to subseasonal state-of-the-art NWP forecasts.

How to cite: Kiefer, S. M., Lerch, S., Ludwig, P., and Pinto, J. G.: Can Machine Learning Models be a Suitable Tool for Predicting Central European Cold Winter Weather on Subseasonal Timescales?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8918, https://doi.org/10.5194/egusphere-egu24-8918, 2024.

Extreme events of the stratospheric polar vortex can modulate subsequent surface weather at subseaonal to seasonal (S2S) timescales. Moreover, they are considered to form windows of opportunity for tropospheric forecasting. This study aims to improve understanding of how the canonical surface response of polar vortex events translates into modulated surface predictability. 

First, we confirm that in the ECMWF extended-range prediction model, the mean signal of weak (strong) polar vortex events projects onto a negative (positive) phase of the North Atlantic Oscillation. The associated equatorward (poleward) shift of the eddy-driven jet then enhances or suppresses synoptic variability in specific regions. By constructing a leadtime, seasonal and model version-dependent climatology of forecast ensemble spread, we link these regions to anomalous forecast uncertainty. For example, sudden stratospheric warmings (SSWs) are followed by a southerly jet shift, which translates into suppressed Rossby wave breaking over Northern Europe, resulting in anomalously high forecast confidence in that region.

In general, both signatures in the mean and spread can contribute to predictability. However, when forecasts are compared to reanalyses, they manifest differently in different skill scores, such as the Root-Mean-Squared Error or the Continuously Ranked Probability Skill Score. We therefore discuss how separate consideration of anomalies in the ensemble mean and ensemble spread may aid to interpret predictability following polar vortex events.

Finally, we apply the diagnostics also to tropical teleconnections. We find indications that windows of forecast opportunity might be dominated by stratospheric polar vortex variability over the Atlantic and by ENSO variability over the Pacific.

How to cite: Spaeth, J., Rupp, P., Garny, H., and Birner, T.: Stratospheric impact on subseasonal forecast uncertainty in the Northern extratropics , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9510, https://doi.org/10.5194/egusphere-egu24-9510, 2024.

Persistent summer weather can result in extreme events with enormous socio-economic impacts; recent summers in Europe have notably demonstrated this. The dynamics that cause persistent surface weather, as well as potential changes under anthropogenic climate change, are the subject of active scientific debate. Summertime atmospheric dynamics have nevertheless received less attention and we are far from obtaining a comprehensive understanding of the mechanisms involved in the formation of persistent weather conditions in summer. This study investigates the drivers responsible for making some surface extreme events more prone to being long-lasting than others.

Gaining a comprehensive understanding of such processes poses challenges due to the complex interactions of variables and fluxes operating at various timescales – from individual weather events (daily to weekly), to the general circulation of the atmosphere and its modulation by specific changes in sea surface temperature or soil moisture interactions (monthly, seasonal to interannual). Furthermore, studies are recently observing that persistent (quasi-stationary or recurrent) circulation patterns do not necessarily always translate to extreme events and persistence at the surface. This discussion extends to open questions about, such as the potential role of soil moisture preconditioning in extending the lifetime of these events.

Starting from an impact-based definition of persistent hot conditions for different European regions, we characterise their persistence by looking at the associated circulation patterns and surface conditions. Through a comparison of long-lived (persistent) and short-duration events, we discern dynamical differences and regional variations that shed light on the common ingredients and potential mechanisms influencing the persistence of extreme heat events in summer. We use the ERA5 reanalysis dataset to take advantage of its high spatiotemporal resolution and relatively long temporal coverage from the 1950s up to today.

A deeper investigation into the dynamical processes controlling persistent surface conditions over Europe in summer is essential for improved predictability at the sub-seasonal to seasonal (S2S) timescale, and it holds significant relevance for risk preparedness. Results from the study aim to advance the discussion on summer dynamics, weather persistence and climate impacts.

How to cite: Pappert, D., Tuel, A., Coumou, D., Vrac, M., and Martius, O.: The dynamics of persistent hotspells in European summers, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9738, https://doi.org/10.5194/egusphere-egu24-9738, 2024.

Seasonal predictions are essential in mitigating damage to people and nature as a result of climate change and extreme events by improving timely decision-making particularly for water and irrigation management. The newly constructed Grand Ethiopian Renaissance Dam, located in the Blue Nile (BN) Basin in Ethiopia at the border to Sudan, increases the urgency of optimized transboundary water management and improved seasonal predictions. However, the global seasonal forecasting systems have known limitations such as biases and drifts. Specifically at regional level, such as in the highlands of Ethiopia, the seasonal predictions need accurate post-processing. Recent developments have shown the large potential of Deep Learning (DL) applications to improve weather and climate predictions. The goal of this study is to improve the global seasonal forecasting system SEAS5 of ECMWF specifically for the BN Basin using DL approaches such as conventional Convolutional Neural Networks (CNN) or more advanced Adaptive Fourier Neural Operators (AFNO). We present first results for improving and downscaling SEAS5 global seasonal precipitation forecasts in the BN Basin with a particular emphasis on ensemble generation and calibration. The neural networks are trained with ERA5-Land-reanalysis data as a ground-truth, which has a higher resolution than SEAS5 (~9km compared to ~36km). This additional downscaling step allows us to consider the high variations in precipitation intensities in the Ethiopian highlands. The results show that the applied DL models have high potential in improving forecasting scores such as the continuous ranked probability skill score. They therefore allow for improved timely decision-making for water management in the transboundary BN Basin.

How to cite: Wiegels, R., Glawion, L., Polz, J., Chwala, C., Weber, J. N., Schober, T. C., Lorenz, C., and Kunstmann, H.: Deep Learning improved seasonal forecasts for the Blue Nile Basin, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11457, https://doi.org/10.5194/egusphere-egu24-11457, 2024.

Large-scale (synoptic to planetary), quasi-stationary circulation patterns in the atmosphere modulate the local weather dynamics from seasonal to sub-seasonal scale. These circulation patterns are known as Weather Regimes (WRs) and are a prominent feature in the midlatitudes. Most studies so far have focused on specific regions (such as the west coast of the United States or the European sector), and during a specific time of the year (namely the boreal winter season). Little work has been done on understanding the spatiotemporal characteristics (frequency, duration, and orientation) of seasonal North American WRs and how they affect local weather, especially in terms of extremes. This study aims to fill this knowledge gap with an investigation of North American WRs independently for all four seasons. Using a k-means clustering algorithm on daily geopotential height anomalies (de-seasonalized at monthly scale) at the 500-hPa pressure level, we identify five WRs in each of the four seasons across three independent reanalysis datasets: 1) MERRA2; 2) ERA5; and 3) NCEP-NCAR Reanalysis 1, for the period 1980-2022. Initial analysis shows that the spatial patterns of these WRs are robust but have non-trivial differences in the frequency and duration of occurrences across different reanalysis datasets. Additionally, we explore the occurrence of local extreme weather (precipitation and temperature) across the contiguous United States (CONUS) during the presence of these seasonal WRs. This study aims to improve the understanding of the seasonal to sub-seasonal variations of North American WRs and their influence on local extreme weather.

How to cite: Borkotoky, S. S., Schiro, K., and Grise, K.: Seasonal classification of North American weather regimes and their effect on extreme weather, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11637, https://doi.org/10.5194/egusphere-egu24-11637, 2024.

Sudden stratospheric warming events (SSWs)–in which the westerly polar vortex rapidly breaks down during winter–are  some of the most dramatic examples of dynamical variability in Earth’s atmosphere. It is now well established that SSWs are, on average, followed by large scale anomalies in near-surface circulation patterns, including an equatorward shift of the eddy driven jet that can persist for several months. These anomalies have, in turn, been related to an increase in the likelihood of a variety of high impact weather extremes. However, not all SSWs are followed by impactful weather events; equally, most winter weather extremes are not preceded by SSWs.

Here we will discuss the extent to which the occurrence of individual extreme weather events and their impacts can be attributed to polar stratospheric variability, drawing upon new results from the Stratospheric Nudging And Predictable Surface Impacts (SNAPSI) project (Working Group 2). This project involves a set of controlled subseasonal hindcast experiments, targeted at three SSW case study events, in which the stratospheric state can be either freely-evolving or nudged towards a climatological or observational state. These simulations reveal that the stratospheric evolution can more than double the regional risk of extreme temperature, rainfall, and snow events. We will go on to explore the attribution of the subsequent impacts of these weather extremes, including on the energy sector, health, and wildfires.  

How to cite: Seviour, W., Butler, A., Garfinkel, C., and Hitchcock, P. and the SNAPSI Working Group 2: Attributing the role of sudden stratospheric warming events in surface weather extremes and their impacts: insights from SNAPSI Working Group 2, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11702, https://doi.org/10.5194/egusphere-egu24-11702, 2024.

Land-atmosphere (L-A) coupling can significantly contribute to subseasonal-to-seasonal (S2S) prediction. During periods of strong L-A coupling, land-atmosphere feedbacks are expected to enhance the memory of the system and therefore also the predictability and prediction skill. This study aims to evaluate S2S prediction of ambient surface air temperature under conditions of strong versus weak L-A coupling in forecasts produced with NASA’s state-of-the-art GEOS S2S forecast system. Utilizing three L-A coupling metrics that together capture the connection between the soil and the free atmosphere, enhanced prediction skill for surface air temperature is observed for 3-4 week boreal summer forecasts across the eastern Great Plains when strong L-A coupling is detected at this lead by all three indices. The forecasts with strong L-A coupling in these “hot spot” regions exhibit warm and dry anomalies, signals that are well simulated in the model. Overall, this study provides insight into how better capturing relevant L-A coupling processes might improve prediction on subseasonal-to-seasonal timescales.

How to cite: Lim, Y., Molod, A., Koster, R., and Santanello, J.: Land-Atmosphere Coupling Simulation and Its Role in Subseasonal-to-Seasonal Prediction, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13143, https://doi.org/10.5194/egusphere-egu24-13143, 2024.

We explored the objective methods to improve long-range forecasting through enhanced forecast skills and integrated forecast information. The objective process we used in this study includes the selection of monitoring factors for more reliable monthly seasonal forecasts. Therefore, we chose the three most significant monitoring factors, i.e., ENSO, snow cover over Eurasia Continent and Arctic sea ice. We first examined the effect and response of the monitoring factors on the boreal winter temperature in South Korea. To improve the information related to the ENSO in seasonal forecasting, the impact of the tropical precipitation which act as an oceanic ENSO forcing was investigated. As one of the important monitoring factors for boreal winter temperature prediction, we analyzed the availability of the index describing austral Eurasian snow cover. We also analyzed the usage of Arctic conditions for predicting monthly temperature for boreal winter. We then investigated how well the effect and response of the factors are simulated in the operational seasonal models. Finally, the link between observation-based monitoring factors and model-based prediction is proposed for objective forecasting.

How to cite: Kim, O., Im, S.-H., and Kim, G.: Improved long-range forecasts in South Korea through integrated forecast information, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14585, https://doi.org/10.5194/egusphere-egu24-14585, 2024.

Weather forecasts at subseasonal-to-seasonal (S2S) timescales have little or no deterministic forecast skill in the troposphere. Individual ensemble members are uncorrelated and span a range of scenarios that are possible for the given set of boundary conditions. The uncertainty of such probabilistic forecasts is then determined by this range of scenarios – often quantified in terms of ensemble spread. For certain boundary conditions, the ensemble spread can be highly anomalous, with conditions associated with reduced spread sometimes referred to as „windows of opportunity“. Various dynamical processes can affect the ensemble spread within a given region, including extreme weather events present in individual members. For geopotential height forecasts over Europe, such extremes are mainly comprised of synoptic storms travelling on the North Atlantic storm track.

We use ECMWF re-forecasts from the S2S database to investigate the connection between storm characteristics and increases in ensemble spread in more detail. We find that the presence of storms in individual ensemble members at s2s time scales forms a major contribution to the geopotential height forecast uncertainty over Europe. In our study, we quantify the magnitude of this contribution and analyse the underlying dynamics, using both Eulerian and Lagrangian frameworks. We further show that certain atmospheric conditions, like various blocked weather regimes, are associated with reduced geopotential height ensemble spread over Europe due to changes in the North Atlantic storm track and associated anomalies in storm density. This connection sheds light on the occurrence of some “windows of opportunity” in the troposphere on S2S time scales.

How to cite: Rupp, P., Afargan-Gerstman, H., Spaeth, J., and Birner, T.: The impact of storm event likelihood on the forecast uncertainty over Europe at S2S time scales, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14626, https://doi.org/10.5194/egusphere-egu24-14626, 2024.

Extratropical cyclones influence midlatitude surface weather directly via precipitation and wind and indirectly via upscale feedbacks on the large-scale flow. Biases in cyclone frequency and characteristics in medium-range to sub-seasonal numerical weather prediction might therefore hinder exploiting the potential predictability on these timescales. We thus, for the first time, identify and track extratropical cyclones in 21 years (2000 – 2020) of sub-seasonal ensemble reforecasts from the European Centre for Medium-Range Weather Forecasts (ECMWF) in the Northern Hemisphere in all seasons. Overall, the reforecasts reproduce the climatology of cyclone frequency and life-cycle characteristics qualitatively well up to six weeks ahead. However, there are significant regional biases in cyclone frequency, which can result from a complex combination of biases in cyclone genesis (locally and upstream), size, location, lifetime, and propagation speed. Their magnitude is largest in summer, with the strongest deficit of cyclones of up to 15% in the North Atlantic, relatively large in spring, and smallest in winter and autumn. Moreover, the reforecast cyclones are too deep in both ocean basins during most seasons, although intensification rates are captured well. An overestimation of cyclone lifetime and differences between the native spatial resolutions of the reforecasts and the verification dataset might explain this intensity bias in some cases, but there are likely further so far unidentified processes involved. While the patterns of cyclone frequency and life cycle biases often appear in lead time weeks 1 and 2, their magnitudes typically grow further at sub-seasonal lead times and, in some cases, saturate in weeks 5 and 6 only. Most of the dynamical sources of these biases thus likely appear in the early medium range, but biases on longer timescales probably contribute to their further increase with lead time. Our study provides a useful basis to identify, better understand, and ultimately reduce biases in the large-scale flow and in surface weather in sub-seasonal weather forecasts. Given the considerable biases during summer, when sub-seasonal predictions of precipitation and surface temperature will become increasingly important, this season deserves particular attention for future research.

How to cite: Sprenger, M., Büeler, D., and Wernli, H.: Northern Hemisphere extratropical cyclone biases in ECMWF sub-seasonal forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14701, https://doi.org/10.5194/egusphere-egu24-14701, 2024.

Given the limited skill of precipitation forecasts, the question arises to what extent ensemble forecasting systems can be used for early warning systems that require longer lead times, such as drought early warning. In this study, we use ECMWF’s IFS extended range forecasts, statistically downscaled to a 2 km grid encompassing Switzerland, to quantify the spatially and seasonally stratified predictability of several precipitation statistics. Consistent with existing analyses we find the predictability of extratropical instantaneous precipitation to be limited to week 1. However, when considering accumulated precipitation and the standardized precipitation index (SPI) forecasts, which is commonly used for drought management, the forecasts are skillful well into week 3. This extension in predictability horizon is attributed to the characteristic of accumulated precipitation, which is less sensitive to differences in timing of precipitating systems. The enhanced predictability of SPI enhances the utility of extended range forecasts for monthly drought forecasts. We discuss the practical applicability of these findings in the context of the new Swiss drought early warning and monitoring platform, planned for operations in 2025. Leveraging the enhanced predictability of SPI, this platform stands to benefit from our research outcomes, providing stakeholders with tools for proactive drought management and response strategies. 

How to cite: Imamovic, A., Büeler, D., Pyrina, M., Humphrey, V., Spirig, C., Moret, L., and Domeisen, D.: How far in advance can we skillfully predict meteorological drought indices?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15134, https://doi.org/10.5194/egusphere-egu24-15134, 2024.

In recent years, deep learning (DL) models have been shown to be able to make competitive forecasts of El Niño Southern Oscillation (ENSO). In most cases, due to the short observational record, the outputs of global circulation models (GCM) are used to train DL models. However, GCMs themselves show biases when modeling ENSO dynamics, such as the lack of phase-locking behavior, shifted precipitation trends, or missing El Niño - La Niña asymmetry. The biases of the GCMs are likely inherited by the DL models during pre-training, raising the question of how we can obtain unbiased DL ENSO models while pre-training on GCM output. In this study, we contend that a possible solution to correct these biases is to use well-established domain adaptation methods, which allow DL models to account for shifts in data distribution between training and validation data sets. In particular, we use a ConvLSTM network trained on CESM2 simulations where we first use a supervised objective to fine-tune our model to reanalysis data. Secondly, we employ test-time training to adapt our model for the domain shift between CESM2 and reanalysis data. This study serves as a first step toward comparing domain adaptation techniques for data-driven seasonal-to-annual DL models in a limited data regime.

How to cite: Rodriguez, M.: Domain adaptation for deep learning ENSO forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15210, https://doi.org/10.5194/egusphere-egu24-15210, 2024.

Droughts, prolonged heat-waves, heavy precipitation events and large-scale flooding - the last years have demonstrated that global climate change is already hitting hard in many places of the Earth. This, inevitably, leads to increased water stress that requires a more sustainable and timely water management across scales. In particular, for optimized use of water resources for irrigation or hydropower generation, it is essential to know their expected availability in the coming months all over the world. This sub-seasonal to seasonal temporal domain, from weeks to months ahead, is addressed by seasonal forecasting systems such as SEAS5, developed by the European Centre for Medium-Range Weather Forecasts (ECMWF). These systems have the potential to provide essential data for enhancing water management practices. Without a bias correction though, the data exhibit a notable deficiency in skill. We have shown for several regions of the world that the “Bias Correction and Spatial Disaggregation” method (BCSD) can improve the forecasting skill substantially. Our next step is now to expand our efforts from the regional to the global scale, i.e., to provide the BCSD-forecasts for the entire globe. Here, the challenge lies in significantly reducing the computational demand for the bias correction: Presently, the BCSD requires several days to execute on a global scale. However, if such forecasts should be used as decision support, a timely provision is crucial.

We therefore present a method to achieve this task: The utilization of fixed Cumulative Distribution Functions (CDFs) rather than their recalculation for each pixel has the potential to enhance the computational efficiency of the bias correction. This approach not only significantly reduces the required data volume but also improves accessibility. To further achieve transferability of the system, we also demonstrate the performance of this system in a containerized environment. Our goal is to achieve a globally corrected SEAS5 forecasts within a time frame of ideally less than one day. With the provision of these bias-corrected data in near-real time, better estimations become available for direct utilization by water managers or as input for subsequent modeling processes.

How to cite: Weber, J. N., Lorenz, C., Schober, T., Wiegels, R., and Kunstmann, H.: Global bias-corrected seasonal forecasts: Towards efficient and near real-time solutions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16022, https://doi.org/10.5194/egusphere-egu24-16022, 2024.

The seasonal cycle (SC) anomalies of winter surface air temperature (SAT) over China mainly include three modes: consistent changes throughout the winter, inverse changes in the early and late winter, and opposite changes in the southern and northern China, respectively.  The positive EOF1 phase (i.e., uniformly warming throughout winter) can be attributed to global warming, especially in the North Atlantic and tropical Pacific. The EOF2 is mainly related to the dipole sea surface temperature (SST) pattern in the North Atlantic. In the early winter, the Rossby wave originating from North Atlantic strengthens Ural blocking high (UBH) and Siberian high (SH) in the early winter, resulting in cold SAT anomalies in most of China. While the large-scale zonal circulation with weakened SH has transformed SAT over China into a warm state in the later winter. The EOF3 can be attributed to the tripole SST in the North Atlantic and El Niño-like SST pattern in the tropical Pacific. In December, the Rossby wave train originating from the mid-latitudes of the North Atlantic Ocean enhances cold air activity in the Northern Hemisphere, causing cold SAT anomalies in Northeast China, while the dominating southerly winds in southern China cause warm SAT anomalies. In the late winter, the large-scale circulation resembles negative AO phase, resulting in the northerly winds and cold SAT anomalies in the northern China. Meanwhile, the anomalous anticyclonic circulation in the Northwest Pacific causes warm SAT anomalies in southern China. Therefore, the combined effects of tropical and extratropical SST should be considered when predicting interannual variability of winter SAT anomalies over China.

How to cite: Yu, M.: Diversity of seasonal cycle anomalies of surface air temperature in winter over China , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16753, https://doi.org/10.5194/egusphere-egu24-16753, 2024.

Combining ensemble forecasts from different models into a multi-model ensemble (MME) have been shown to improve forecast skill at different time-scales, including the sub-seasonal to seasonal (S2S) one. Here, we investigate a new method to build such MME based on barycenter.

Recognizing ensemble forecasts as discrete probability distributions, we work directly in the probability distribution space. This allows us to use existing tools in this space, and in particular the concept of barycenter. The barycenter of a collection of distributions (or the ensemble forecasts here) is the distribution that best represents them, based on a given metric. The barycenter can thus be seen as the combination of these distributions, and so used to build a MME. We compare two barycenters based on different metrics: the L2 and the Wasserstein distances. The Wasserstein distance corresponds to the cost of the optimal transport between two distributions and has interesting properties in the distribution space. We compare it to the L2-barycenter which is in fact shown to be equivalent to the well-known “pooling” MME method (i.e. the concatenation of the different ensembles members). Another interesting point of the barycenters is that they allow you to give different weights to the models and so to easily build a weighted-MME. The weights have an important impact on the skill of the MMEs. We are thus optimizing the weights by learning them from the data using cross-validation on the forecasts.

The two barycenter-based MMEs are applied to the combination of the models from the S2S project’s database. Their performances are evaluated for the prediction of weekly 2m-temperature during European winter with respect to different metrics. As a proof of concept, we first start with the combination of two models, namely the European Centre Medium-Range Weather Forecasts (ECMWF) and the National Center for Environmental Prediction (NCEP) models. We show that the two MMEs are generally able to perform as well or better than both the single-models, but that the best combination method depends on the chosen metric. We then extend the barycenter approach to the combination of more models, of which we will discuss preliminary results.

How to cite: Le Coz, C., Tantet, A., Flamary, R., and Plougonven, R.: A barycenter-based approach for the multi-model ensembling of subseasonal forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17318, https://doi.org/10.5194/egusphere-egu24-17318, 2024.

Seasonal forecasting has the potential to support agricultural activities by offering crop-yield forecasts and facilitating measures to mitigate weather-related damages. This study aims to enhance the application of subseasonal to seasonal (S2S) forecasts in agriculture by evaluating them through tailored verification methods that consider crop calendars and areas.

The verification employs the so-called 1-norm continuous ranked probability score (CRPS), which utilizes the absolute norm instead of a square to quantify forecast errors. While the 1-norm CRPS is not a proper score and does not suit for ensemble forecast verification, it offers an advantage in terms of user-friendliness. Specifically, the score is proportional to the expectation of the absolute error, and thus, it is easier to relate the outcomes of crop models under the assumption of linearity compared to other scores like the ordinal CRPS.

Crop regions and seasons for major commodity crops such as wheat, rice, and maize were identified using global datasets of crop yields and crop calendars. Using the crop calendar information, we can assess the within-season forecast performance in relation to crop growth stages globally. Reforecast data from seasonal forecasts archived by the EU-funded Copernicus Climate Change Service (C3S) were evaluated, allowing for a multi-model comparison of forecast skill. The presentation illustrates a set of example verification products targeted to the common commodity crops. A comprehensive overview of forecast skill for the target crops is anticipated to facilitate a dialogue between meteorological and agricultural experts, thereby enhancing the usability of the seasonal forecast.

How to cite: Takaya, Y., Iizumi, T., Masutomi, Y., and Nakaegawa, T.: Verification of seasonal forecast for facilitating agricultural applications, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17920, https://doi.org/10.5194/egusphere-egu24-17920, 2024.

Predictive skills of coupled sea-ice/ocean and atmosphere models are limited by the chaotic nature of the atmosphere. Assimilation of observational information on ocean hydrography and sea ice allows to obtain a coupled-system state that provides a basis for subseasonal-to-seasonal ocean and sea-ice forecast (Mu et al., 2022). However, if the atmosphere is not additionally constrained, the quasi-random atmospheric states within an ensemble forecast lead to a fast divergence of the ocean and sea-ice states, degrading the system’s performance with respect to the sea ice forecasts. As reported previously, imposing an additional constraint by nudging large-scale winds to the ERA5 reanalysis data (Sánchez-Benítez et al., 2021; Athanase et al., 2022) improves predictive skills of the AWI Coupled Prediction System (AWI-CPS, Mu et al. 2022) with regard to sea ice drift (Losa et al., 2023). Here we provide results based on a much more extensive set of ensemble-based data assimilation experiments spanning the time period from 2002 to 2023 and a series of long forecast experiments over 2010 – 2023, initialized in four different seasons. We compare the performance of forecasts initialized from two sets of data assimilation experiments, with and without atmospheric wind nudging. The additional relaxation of the large-scale atmospheric circulation to the ERA5 reanalysis data for the initialization leads to reasonable atmospheric forecast skill on weather timescales: Despite the simple technique, the coarse resolution compared to NWP systems, and the limited optimization efforts, 10-day forecasts of the 500 hPa geopotential height are about as skillful as the best performing NWP forecasts were about 10 –15 years ago. Among other aspects, this leads to significantly improved subseasonal-to-seasonal sea-ice concentration and thickness forecasts.

Athanase, M., Schwager, M., Streffing, J., Andrés-Martínez, M., Loza, S., and Goessling, H.: Impact of the atmospheric circulation on the Arctic snow cover and ice thickness variability , EGU General Assembly 2022, Vienna, Austria, 23–27 May 2022, EGU22-5836, https://doi.org/10.5194/egusphere-egu22-5836, 2022.

Losa, S. N., Mu, L., Athanase, M., Streffing, J., Andrés-Martínez, M., Nerger, L., Semmler, T., Sidorenko, D., and Goessling, H. F.: Combining sea-ice and ocean data assimilation with nudging atmospheric circulation in the AWI Coupled Prediction System, EGU General Assembly 2023, Vienna, Austria, 24–28 Apr 2023, EGU23-14227, https://doi.org/10.5194/egusphere-egu23-14227, 2023.

Mu, L. , Nerger, L. , Streffing, J. , Tang, Q. , Niraula, B. , Zampieri, L., Loza, S. N. and Goessling, H. F. (2022): Sea‐Ice Forecasts With an Upgraded AWI Coupled Prediction System , Journal of Advances in Modeling Earth Systems, 14 (12) . doi: 10.1029/2022ms003176

Sánchez-Benítez, A. , Goessling, H. , Pithan, F. , Semmler, T. and Jung, T. (2022): The July 2019 European Heat Wave in a Warmer Climate: Storyline Scenarios with a Coupled Model Using Spectral Nudging , Journal of Climate, 35 (8), pp. 2373-2390 . doi: 10.1175/JCLI-D-21-0573.1

How to cite: Loza, S., Athanase, M., Mu, L., Streffing, J., Sánchez-Benítez, A., Andrés-Martínez, M., Nerger, L., Semmler, T., Sidorenko, D., and Goessling, H.: Improving daily-to-seasonal sea ice forecasts of the AWI coupled prediction system with sea-ice and ocean data assimilation and atmospheric large-scale wind nudging., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-18260, https://doi.org/10.5194/egusphere-egu24-18260, 2024.

Accurate subseasonal forecast of East Asian summer monsoon precipitation (EASM) is pivotal, impacting the livelihoods of billions. However, the proficiency of state-of-the-art subseasonal-to-seasonal (S2S) models in forecasting precipitation remains constrained. We developed a convolutional neural network regression model, harnessing the more reliably predicted atmospheric variables from dynamic models to enhance their forecast skills for precipitation. The outcomes of the CNN model are promising: a 12% increase in accuracy and a 10% reduction in RMSE for precipitation forecast at the lead time of one week. The predictive skill of dynamic models for atmospheric variables shows a significant correlation with the performance of the CNN model. Ablation experiments on various predictors reveal that xx is the most influential factor affecting the CNN model's performance.

How to cite: Jiahui, Z. and Liu, F.: Improving sub-seasonal forecasting of East Asian monsoon precipitation with deep learning, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19908, https://doi.org/10.5194/egusphere-egu24-19908, 2024.

The year-to-year varying annual evolutions of the stratospheric polar vortex (SPV) have an important downward impact on the weather and climate from winter to summer and thus potential implications for seasonal forecasts. This study constructs a parametric elliptic orbit model for capturing the annual evolutions of mass-weighted zonal momentum at 60° N (MU) and total air mass above the isentropic surface of 400 K (M) over the latitude band of 60–90° N from 1 July 1979 to 30 June 2022. The elliptic orbit model naturally connects two time series of a nonlinear oscillator. As a result, the observed coupling relationship between MU and M associated with SPV as well as its interannual variations can be well reconstructed by a limited number of parameters of the elliptic orbit model. The findings of this study may pave a new way for short-time climate forecasts of the annual evolutions of SPV, including its temporal evolutions over winter seasons as well as the spring and fall seasons, and timings of the sudden stratospheric warming events by constructing its elliptic orbit in advance.

How to cite: Secor, M., Yu, Y., Sun, J., Cai, M., and Luo, X.: A Parametric Model of Elliptic Orbits for Annual Evolutions of Northern Hemisphere Stratospheric Polar Vortex and Their Interannual Variability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20725, https://doi.org/10.5194/egusphere-egu24-20725, 2024.

The Madden-Julian Oscillation (MJO) is a large-scale tropical phenomenon where fluctuations of clouds, rainfall, winds, and pressure propagate eastward around the globe every 30 to 90 days on average. The MJO has significant impacts on weather and climate both locally and globally. Despite its importance, forecasting the MJO remains challenging due to the limitations of traditional numerical and statistical methods. To address this, machine learning has emerged as a promising avenue for MJO forecasting (Martin et al. 2022, Silini et al. 2021, Delaunay and Christensen 2022). Apart from accurate forecasts emphasized in previous research, our study aims to get more physical insights: we build a predictive and interpretable convolution neural network (CNN) and unravel what tropical waves at which spatial scales are essential for MJO forecasting.

Our CNN model takes tropical reanalysis maps as input and predicts the MJO index, achieving forecast skills comparable to NCEP Climate Forecast System (CFSv2). This level of skill is state-of-art in interpretable neural networks. To understand what information is crucial to our MJO forecast, we decompose the output of each convolution layer into tropical waves at different zonal scales. We find that the CNN focuses on large-scale patterns whose zonal scale is above 2500 km. In fact, even when fed exclusively with large-scale features as input, the CNN achieves MJO forecasts akin to the skill of the original model. Furthermore, the CNN chooses to reconstruct large-scale features from input containing solely small-scale features instead of relying directly on small scales for forecasting. This reconstruction further emphasizes the critical role of large-scale patterns in MJO predictions.

In future research, we plan to perform a systematic analysis to evaluate the contribution of different tropical waves to MJO forecasting. We will also simplify the model architecture to facilitate better understanding. Additionally, we plan to incorporate more previous time steps as input memories to enhance forecast accuracy. This work represents a promising advance towards economic yet precise MJO forecasting.

How to cite: Yao, L., Yang, D., Duncan, J., Kumar Chattopadhyay, A., Hassanzadeh, P., Bhimji, W., and Yu, B.: Machine Learning Models Use Large Scale Signals to Forecast the MJO, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20993, https://doi.org/10.5194/egusphere-egu24-20993, 2024.

The latest assessment report (AR6) of the Intergovernmental Panel on Climate Change includes a new element to climate research, i.e. the Interactive Atlas (IA), which is very useful for users from different sectors. As the new CMIP6 global climate model simulations use the brand-new SSP-scenarios paired with the RCP-scenarios, the latest climate change projections should be evaluated in order to update the regional and national adaptation strategies. Keeping this in mind we focused on Europe, with a special emphasis on Hungary in our study.

Our aim was to analyse the potential future changes of different temperature indices for Europe, in order to recognize spatial patterns and trends that may shape our climate in the second half of the 21^st century. For this purpose, multi-model mean simulation data provided by the IPCC AR6 WG1 IA were downloaded on a monthly base. We chose two climate indices beside the mean temperature values, which represent temperature extremes, namely, the number of days with maximum temperature above 35 °C and the number of frost days (i.e. when daily minimum temperature is below 0 °C). We focused on the end of the 21^st century (2081–2100) with also briefly considering the medium-term changes of the 2041–2060 period (both compared to the last two decades of the historical simulation period, i.e. 1995–2014 as the reference period). For both future periods we used all scenarios provided in the IA, namely, SSP1-2.6, SSP2-4.5, SSP3-7.0 and SSP5-8.5.

Several zonal and meridional segments over the continent were defined, where we analysed the projected changes of the indices. The zonal segments provide an insight on two different effects that may induce spatial differences between future regional changes. (i) Continentality can be recognized as an increasing effect from the western parts of the segment towards the east. (ii) Topography also appears as the influence of mountains, plains, and basins emerge. The meridional segments provide information about the north-to-south differences as well, as the effects of sea cover. The changes in the indices are plotted on diagrams representing the different months, where the differences in the scenarios are also shown. These diagrams are compared to their respective landscape profiles, furthermore, statistical parameters were calculated. In addition, a monotony index was defined as the cumulative direction of differences between the neighbouring grid cells and analysed within the study.

Our results show that in the changes of mean temperature, both the zonal location and sea cover will play a key role in forming spatial differences within Europe. However, for the extreme temperature indices, topography and continentality are likely to become more dominant than sea cover, while the zonal location remains an important factor. 

Acknowledgements: This work was supported by the Hungarian National Research, Development and Innovation Fund [grant numbers PD138023, K-129162], and the National Multidisciplinary Laboratory for Climate Change [grant number RRF-2.3.1-21-2022-00014]. 

How to cite: Divinszki, F., Kis, A., and Pongrácz, R.: Analysing the projected monthly changes of temperature-related climate indices over Europe using zonal and meridional segments based on CMIP6 data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-389, https://doi.org/10.5194/egusphere-egu24-389, 2024.

PDO and ENSO are most prominent variability modes in the Pacific Ocean at decadal and interannual timescales respectively. Mutual independence between ENSO and PDO is questionable (Chen & Wallace, 2016). Linear combination of the first two orthogonal modes of SST variability in our Study Region (SR; 70^oN - 20^oS, 110^oE - 90^oW) i.e. mode 1 (interannual mode, we call it, IAM; ENSO like variability) and mode 2 (North Pacific Mode (NPM; Deser & Blackmon (1995)); a decadal mode) produces a PDO like variability (Chen & Wallace, 2016). It suggests that PDO is not independently hosted in the Pacific Ocean and can be represented by two linearly independent variability modes.

To produce credible and skillful climate information at multi-year to decadal timescales, Decadal Climate Prediction Project (DCPP), led by the Working Group on Subseasonal to Interdecadal Prediction (WGSIP), focuses on both the scientific and practical elements of forecasting climate by employing predictability research and retrospective analyses within the Coupled Model Intercomparison Project Phase 6 (CMIP6). Component A under DCPP experiments concentrates on hindcast experiments to examine the prediction skill of participating models with respect to actual observations.

As linear combination of  IAM and NPM in SR produces PDO pattern and timescales efficiently, we compared the  ability of DCPP-A hindcasts to predict  IAM, NPM, and  PDO. In this analysis we use output from 9 models (a total of 128 ensemble members), initialised every year from 1960 to 2010. To produce the prediction skill estimates.

At lead year 1 from initialisation, the prediction of NPM,  IAM and PDO is quite skillful as the models are initialised with observations. In subsequent years, skill of either IAM or NPM or both drop significantly and that leads to drop in skill of predicted PDO index. Both the deterministic estimates and probabilistic estimates of prediction skill for DCPP hindcast experiments suggest that the ability of hindcast experiments to predict NPM governs the prediction skill to predict PDO index.

Keywords: PDO, ENSO, NPM, CMIP6, DCPP, hindcast

References

Chen, X., & Wallace, J. M. (2016). Orthogonal PDO and ENSO indices. Journal of Climate, 29(10), 3883–3892. https://doi.org/10.1175/jcli-d-15-0684.1

Deser, C., & Blackmon, M. L. (1995). On the Relationship between Tropical and North Pacific Sea Surface Temperature Variations. Journal of Climate, 8(6), 1677–1680. https://doi.org/10.1175/1520-0442(1995)008<1677:OTRBTA>2.0.CO;2

How to cite: Dixit, J. and AchutaRao, K. M.: Relationship of the predictability of North Pacific Mode and ENSO with predictability of PDO, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-868, https://doi.org/10.5194/egusphere-egu24-868, 2024.

The signal-to-noise paradox is known to be a limitation in multiple seasonal and decadal forecast models where the model ensemble mean predicts observations better than individual ensemble members. This ‘paradox’ occurs for different parameters, like the NAO, temperature, wind speed or storm counts in multiple seasonal and decadal forecasts. However, investigations have not yet found the origin of the paradox. First hypotheses are that weak ocean – atmosphere coupling or a misrepresentation of eddy feedback in these models is responsible.

Our previous study found a stronger signal-to-noise error in windstorm frequency than for the NAO despite highly significant forecast skill. In combination with the underestimation of eddy feedback in multiple models, this led to the question: Might the signal-to-noise paradox over the North-Atlantic be driven by severe winter windstorms?

To assess this hypothesis, the signal-to-noise paradox is investigated in multiple seasonal forecast suites from the UK Met Office, ECMWF, DWD and CMCC. The NAO is used to investigate the changes in the paradox depending on the storminess of the season. The results show a significant increase of the NAO-signal-to-noise error in stormy seasons in GloSea5. Other individual models like the seasonal model of the DWD or CMCC do not show such a strong difference. A multi-model approach, on the other hand, shows the same tendency as GloSea5. Nevertheless, these model differences mean that more hindcasts are needed to conclusively demonstrate that the signal-to-noise error arises from Atlantic windstorms.

How to cite: Degenhardt, L., Leckebusch, G. C., Scaife, A. A., Smith, D., and Hardiman, S.: Is the NAO signal-to-noise paradox exacerbated by severe winter windstorms?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1757, https://doi.org/10.5194/egusphere-egu24-1757, 2024.

Sea surface temperature (SST) changes in the Mediterranean Sea have profound impacts on both the Mediterranean regions and remote areas. Previous studies show that the Mediterranean SST has significant decadal variability that is comparable with the Atlantic multidecadal variability (AMV). However, few studies have discussed the characteristics and sources of the decadal predictability of Mediterranean SST based on observations. Here for the first time we use observational datasets to reveal that the decadal predictability of Mediterranean SST is contributed by both external forcings and internal variability for both annual and seasonal means, except that the decadal predictability of the winter mean SST in the eastern Mediterranean is mostly contributed by only internal variability. Besides, the persistence of the Mediterranean SST is quite significant even in contrast with that in the subpolar North Atlantic, which is widely regarded to have the most predictable surface temperature on the decadal time scale. After the impacts of external forcings are removed, the average prediction time of internally generated Mediterranean SST variations is more than 10 years and closely associated with the multidecadal variability of the Mediterranean SST that is closely related to the accumulated North Atlantic Oscillation forcing.

How to cite: Yan, X., Tang, Y., and Yang, D.: Study of the Decadal Predictability of Mediterranean Sea Surface Temperature Based on Observations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1940, https://doi.org/10.5194/egusphere-egu24-1940, 2024.

The European North-West shelf seas (NWS) support economic interests and provide environmental services to several adjacent populous countries. Skilful seasonal forecasts of the NWS would be useful to support decision making. Here, we quantify the skill of an operational large-ensemble ocean-atmosphere coupled dynamical forecasting system (GloSea), as well as a benchmark persistence forecasting system, for predictions of NWS sea surface temperature (SST) at 2-4 months lead time in winter and summer. We also identify sources of- and limits to NWS SST predictability with a view to what additional skill may be available in the future. We find that GloSea NWS SST skill is generally high in winter and low in summer. Persistence of anomalies in the initial conditions contributes substantially to predictability. GloSea outperforms simple persistence forecasts, by adding atmospheric variability information, but only to a modest extent. Where persistence is low – for example in seasonally stratified regions – both GloSea and persistence forecasts show lower skill. GloSea skill can be degradeded by model deficiencies in the relatively coarse global ocean component, which lacks a tidal regime and likely fails to properly fine-scale NWS physics. However, using “near perfect atmosphere” tests, we show potential for improving predictability of currently low performing regions if atmospheric circulation forecasts can be improved, underlining the importance of development of atmosphere-ocean coupled models for NWS seasonal forecasting applications.

How to cite: Atkins, J., Tinker, J., Graham, J., Scaife, A., and Halloran, P.: Seasonal forecasting of the European North-West shelf seas: limits of winter and summer sea surface temperature predictability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3190, https://doi.org/10.5194/egusphere-egu24-3190, 2024.

Spring frost events occurring after budburst of grapevines can damage new shoots, disrupt plant growth and cause large economic losses to the viticultural sector. Frost protection practices encompass a variety of vineyard management actions across timescales, from seasonal to decadal and beyond. The cost-effectiveness of such measures depends on the availability of accurate predictions of the relevant climate hazards at the appropriate timescales.

In this work, we present a statistical downscaling method which predicts variations in the frequency of occurrence of spring frost events in the important winemaking region of Catalunya at the seasonal timescale. The downscaling method exploits the seasonal predictability associated with the predictable components of the atmospheric variability over the Euro-Atlantic region, and produces local predictions of frost occurrence at a spatial scale relevant to vineyard management.

The downscaling method is designed to address the specific needs highlighted by a representative stakeholder in the local viticultural sector, and is expected to deliver an actionable prototype climate service. The statistical procedure is developed in perfect prognosis mode: the method is trained with large-scale reanalysis data against a high-resolution gridded observational reference, and validated against multi-model seasonal hindcast predictions.

Our work spotlights the potential benefits of transferring climate predictability across spatial scales for the design and provision of usable climate information, particularly regarding extremes.

How to cite: Roncoroni, S., Athanasiadis, P., and Gualdi, S.: Statistical downscaling of extremes in seasonal predictions - a case study on spring frosts for the viticultural sector, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4538, https://doi.org/10.5194/egusphere-egu24-4538, 2024.

Skillful prediction of seasonal monsoons has been a challenging problem since the 1800s. However, significant progress has been made in Indian summer monsoon rainfall prediction in recent times, with skill scores reaching 0.6 and beyond, surpassing the estimated predictability limits. This phenomenon leads to what is known as the “Signal-to-noise Paradox.” To investigate this paradox, we utilized 52 ensemble member hindcast runs spanning 30 years.

Through the application of ANOVA and Mutual Information methods, we estimate the predictability limit globally. Notably, for the boreal summer rainfall season, the Indian subcontinent exhibited the paradox, among several other regions, while the Equatorial Pacific region, despite demonstrating high prediction skill, does not have the Signal-to-Noise paradox. We employed a novel approach to understand how sub-seasonal variability and their projection in association with predictors are linked to the paradoxical behavior of seasonal prediction skill.

We propose a new method to estimate predictability limits that is free from paradoxical phenomena and shows much higher seasonal predictability. This novel method provides valuable insights into the complex dynamics of monsoon prediction, thereby creating opportunities for expanded research and potential improvements in seasonal forecasting skill in the coming years.

How to cite: Shivamurthy, Y., Saha, S. K., Pokhrel, S., Konwar, M., and Chaudhari, H. K.: Why does the Signal-to-Noise Paradox Exist in Seasonal Climate Predictability?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4873, https://doi.org/10.5194/egusphere-egu24-4873, 2024.

Compound Dry and Hot Extremes (CDHE) have an adverse impact on socioeconomic factors during the Indian summer monsoon, and a future exacerbation is anticipated. The occurrence of CDHE is influenced by teleconnections, which play a crucial role in determining its likelihood on a seasonal scale. Despite the importance, there is a lack of studies unravelling the teleconnections of CDHE in India. Previous investigations specifically focused on teleconnections between precipitation, temperature, and climate indices. Hence, there is a need to unravel the teleconnections of CDHE. This study presents a framework combining event coincidence analysis (ECA) with complexity science. ECA evaluates the synchronization between CDHE and climate indices. Subsequently, complexity science is utilized to construct a driver-CDHE network to identify the critical drivers of CDHE. A logistic regression model is employed to evaluate the proposed drivers' effectiveness. The occurrence of CDHE exhibits distinct patterns from July to September when considering intra-seasonal variability. Our findings contribute to the identification of drivers associated with CDHE. The primary driver for Eastern, Western India and Central India is the indices in the Pacific Ocean and Atlantic Ocean, respectively, followed by the indices in the Indian Ocean. These identified drivers outperform the traditional Niño 3.4-based predictions. Overall, our results demonstrate the effectiveness of integrating ECA and complexity science to enhance the prediction of CDHE occurrences.

How to cite: Agarwal, A. and Guntu, R.: Towards the Predictability of Compound Dry and Hot Extremes through Complexity Science, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-7134, https://doi.org/10.5194/egusphere-egu24-7134, 2024.

The implementation of adaptation policies requires seamless and relevant information on the evolution of the climate over the next decades. Decadal climate predictions are subject to drift because of intrinsic model errors and their skill may be limited after a few years or even months depending on the region. Non-initialized ensembles of climate projections have large uncertainties over the next decades, encompassing the full range of uncertainty attributed to internal climate variability. Providing the best climate information over the next decades is therefore challenging. Recent studies have started to address this challenge by constraining uninitialized projections of sea surface temperature using decadal predictions or using a storyline approach to constrain uninitialized projections of the Atlantic Meridional Overturning Circulation using observations. Here, using a hierarchical clustering method, we select a sub-ensemble of non-initialized climate simulations based on their similarity to observations. Then, we try to further refine this sub-ensemble of trajectories by selecting a subset based on its consistency with decadal predictions. This study presents a comparison of these different methods for constraining surface temperatures in the North-Atlantic / Europe region over the next decades, focusing on CMIP6 non-initialized simulations.

How to cite: Bonnet, R., Boé, J., and Sanchez, E.: Constraining near to mid-term climate projections by combining observations with decadal predictions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8028, https://doi.org/10.5194/egusphere-egu24-8028, 2024.

Adaptation to climate change requires accurate and reliable climate information on decadal and multi-decadal timescales. Such near-term climate information is obtained from future projection simulations, which are strongly affected by uncertainties related to, among other things, internal climate variability. Here we present an approach to constrain variability in future projection simulations of the coupled model intercomparison project phase 6 (CMIP6). The constraining approach involves phasing in the simulated with the observed climate state by evaluating the area-weighted spatial pattern correlations of sea surface temperature (SST) anomalies in individual members and observations. The constrained ensemble, based on the top ranked members in terms of pattern correlations with observed SST anomalies, shows significant added value over the unconstrained ensemble in predicting surface temperature 10 and also 20 years  after the synchronization with observations, thus extending the forecast range of the standard initialised predictions. We also find that while the prediction skill of the constrained ensemble for the first ten years is similar to the initialized decadal predictions, the added value against the unconstrained ensemble extends over more regions than the decadal predictions. In addition, the constraining approach can also be used to attribute predictability of regional and global climate variations to regional SST variability.

How to cite: Mahmood, R., G. Donat, M., Ortega, P., and Doblas-Reyes, F.: Constraining internal variability in CMIP6 simulations to provide skillful near-term climate predictions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9049, https://doi.org/10.5194/egusphere-egu24-9049, 2024.

How to cite: Christiansen, B., Yang, S., and Drews, A.: The Atlantic Multi-decadal Variability in observations and in a large historical multi-model ensemble: Forced and internal variability, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9100, https://doi.org/10.5194/egusphere-egu24-9100, 2024.

The recent severe European summer heat waves of 2015 and 2018 co-occurred with cold subpolar North Atlantic (NA) sea surface temperatures (SSTs). However, a significant connection between this oceanic state and European heat waves was not yet established.

We investigate the effect of cold subpolar NA SSTs on European summer heat waves using two 100-year long AMIP-like model experiments: one that employs the observed global 2018 SST pattern as a boundary forcing and a counter experiment for which we removed the negative NA SST anomaly from the 2018 SST field, while preserving daily and small-scale SST variabilities. Comparing these experiments, we find that cold subpolar NA SSTs significantly increase heat wave duration and magnitude downstream over the European continent. Surface temperature and circulation anomalies are connected by the upper-tropospheric summer wave pattern of meridional winds over the North Atlantic European sector, which is enhanced with cold NA SSTs. Our results highlight the relevance of the subpolar NA region for European summer conditions, a region that is marked by large biases in current coupled climate model simulations.

How to cite: Bischof, S., Pilch Kedzierski, R., Hänsch, M., Wahl, S., and Matthes, K.: The Role of the North Atlantic for Heat Wave Characteristics in Europe, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9274, https://doi.org/10.5194/egusphere-egu24-9274, 2024.

Several studies show that the occurrence of summer extreme temperatures over Europe is increased since the middle of the twentieth century and is expected to further increase in the future due to global warming (Seneviratne et al., 2021). Thus, predicting heat extremes several months ahead is crucial given their impacts on socio-economic and environmental systems.

In this context, state-of-the-art dynamical seasonal prediction systems (SPSs) show low skills in predicting European heat extremes on seasonal timescale, especially in central and northern Europe (Prodhomme et al., 2022). However, recent studies have shown that our skills in predicting extratropical climate can be largely improved by subsampling the dynamical SPS ensemble with statistical post-processing techniques (Dobrynin et al., 2022).

This study assesses if the seasonal prediction skill of summer extreme temperatures in Europe in the state-of-the-art dynamical SPSs can be improved through subsampling. Specifically, we use a multi-model ensemble (MME) of SPSs contributing to the Copernicus Climate Change Service (C3S), analysing di hindcast period 1993—2016. The MME is subsampled by retaining a subset of members that predict the phase of the North Atlantic Oscillation (NAO) and the Eastern Atlantic (EA), typically linked to summer extreme temperatures in Europe. The subsampling relies on spring predictors of the weather regimes and thus allows us to retain only those ensemble members with a reasonable representation of summer heat extreme teleconnections.

Results show that by retaining only those ensemble members that accurately represent the NAO phase, it not only enhances the seasonal prediction skills for the summer European climate but also leads to improved predictions of summer extreme temperatures, especially in central and northern Europe. Differently, selecting only those ensemble members that accurately represent the EA phase does not improve either the predictions of summer European climate or the predictions of summer extreme temperatures. This can be explained by the fact that the C3S SPSs exhibits deficiencies in accurately representing the summer low-frequency atmospheric variability.

Bibliography

Dobrynin, M., and Coauthors, 2018: Improved Teleconnection-Based Dynamical Seasonal Predictions of Boreal Winter. Geophysical Research Letters, 45 (8), 3605—3614, https://doi.org/10.1002/2018GL07720

Prodhomme, C., S. Materia, C. Ardilouze, R. H. White, L. Batté, V. Guemas, G. Fragkoulidis, and J. Garcìa-Serrano, 2022: Seasonal prediction of European summer heatwaves. Climate Dynamics, 58 (7), 2149—2166, https://doi.org/10.1007/s00382-021-05828-3

Seneviratne, S., and Coauthors, 2021: Weather and Climate Extreme Events in a Changing Climate, chap. 11, 1513—1766. Cambridge University Press, https://doi.org/10.1017/9781009157896.013

How to cite: Famooss Paolini, L., Ruggieri, P., Pascale, S., Brattich, E., and Di Sabatino, S.: Hybrid statistical-dynamical seasonal prediction of summer extreme temperatures over Europe, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9690, https://doi.org/10.5194/egusphere-egu24-9690, 2024.

As a consequence of limited reliability of dynamical forecast systems, particularly over Europe, efforts in recent years have turned to exploiting the power of Machine Learning methods to extract information on drivers of extreme temperature from observations and reanalysis. Meanwhile, the diverse impacts of extreme heat have driven development of new indicators which take into account nightime temperatures and humidity. In the H2020 CLimate INTelligence (CLINT) project, a feature selection framework is being developed to find the combination of drivers which provides optimal seasonal forecast skill of European summer heatwave indicators. Here, we present the methodology, its application to a range of heatwave indicators and forecast skill compared to existing dynamical systems. First, a range of (reduced-dimensionality) drivers are defined, including k-means clusters of variables known to impact European summer (e.g. precipitation, sea ice content), and more complex indices like the NAO and weather regimes. Then, these drivers are used to train machine learning based prediction models, of varying complexity, to predict seasonal indicators of heatwave occurrence and intensity. A crucial and novel step in our framework is the use of the Coral Reef Optimisation algorithm, used to select the variables and their corresponding lag times and time periods which provide optimal forecast skill. To maximise training data, both ERA5 reanalysis and a 2000-year paleo-simulation are used; the representation of heatwaves and atmospheric conditions are validated with respect to ERA5. We present comparisons of forecast skill to the dynamical Copernicus Climate Change Service seasonal forecasts systems. The differences in timing, predictability and drivers of daytime and nighttime heatwaves across Europe are highlighted. Lastly, we discuss how the framework can easily be adapted to other extremes and timescales.

How to cite: McAdam, R., Peláez Rodríguez, C., Hansen, F., Pérez Aracil, J., Squintu, A., Cavicchia, L., Zorita, E., Saldez-Sanz, S., and Scoccimarro, E.: Optimization-based driver detection and prediction of seasonal heat extremes, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9905, https://doi.org/10.5194/egusphere-egu24-9905, 2024.

Long-term sea level rise and multiyear-to-decadal sea level variations pose substantial risks of flooding and erosion in coastal communities. The North Atlantic Ocean and the U.S. East Coast are hotspots for sea level changes under current and future climates. Here, we employ a machine learning technique, a self-organizing map (SOM)-based framework, to systematically characterize the North Atlantic sea level variability, assess sea level predictability, and generate sea level predictions on multiyear-to-decadal timescales. Specifically, we classify 5000-year North Atlantic sea level anomalies from the Seamless System for Prediction and EArth System Research (SPEAR) model control simulations into generalized patterns using SOM. Preferred transitions among these patterns are further identified, revealing long-term predictability on multiyear-to-decadal timescales related to shifts in Atlantic meridional overturning circulation (AMOC) phases. By combining the SOM framework with “analog” techniques based on the simulations and observational/reanalysis data, we demonstrate prediction skill of large-scale sea level patterns comparable to that from initialized hindcasts. Moreover, additional source of short-term predictability is identified after the exclusion of low-frequency AMOC signals, which arises from the wind-driven North Atlantic tripole mode triggered by the North Atlantic Oscillation. This study highlights the potential of machine learning methods to assess sources of predictability and to enable efficient, long-term climate prediction.

How to cite: Gu, Q., Jia, L., Zhang, L., Delworth, T., Yang, X., Zeng, F., and Li, S.: Exploring multiyear-to-decadal North Atlantic sea level predictability using machine learning and analog methods, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10539, https://doi.org/10.5194/egusphere-egu24-10539, 2024.

The inter-annual to multi-decadal variability of recurrent, synoptic-scale atmospheric circulation patterns in the Northern Hemisphere extratropics, as represented by the Jenkinson-Collison classification scheme, is explored in reanalysis data spanning the entire 20th century, and in global climate model (GCM) data from the historical, AMIP and DCPP experiments conducted within the framework of CMIP6. The aim of these efforts is to assess the effect of coupled vs. uncoupled and initialised vs. non-initialized GCM simulations in reproducing the observed low-frequency variability of the aforementioned circulation patterns.

Results reveal that the observed annual counts of typical recurrent weather patterns, such as cyclonic or anticyclonic conditions and also situations of pronounced advection, exhibit significant oscillations on multiple time-scales ranging between several years and several decades. The period of these oscillations, however, is subject to large regional variations. This is in line with earlier studies suggesting that the extratropical atmospheric circulation’s low frequency variability is essentially unforced, except in the Pacific-North American sector where the forced variability is enhanced due to ENSO teleconnections. Neither the periods obtained from historical nor those obtained from AMIP experiments align with observations. Likewise, not even the periods obtained from different runs of the same GCM and experiment correspond to each other. Thus, in an non-initialized model setup, ocean-atmosphere coupling or the lack thereof essentially leads to the same results. Whether initialization and/or augmenting the ensemble size can improve these findings, will also be discussed.

Acknowledgement: This work is part of project Impetus4Change, which has received funding from the European Union’s Horizon Europe research and innovation programme under grant agreement No 101081555.

How to cite: Brands, S., Cimadevilla, E., and Fernández, J.: Low-frequency variability of synoptic-scale atmospheric circulation patterns in the Northern Hemisphere extratropics and associated hindcast skill of decadal forecasting systems, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10551, https://doi.org/10.5194/egusphere-egu24-10551, 2024.

Global mean surface temperature is the key metric by which our warming climate is monitored and for which international climate policy is set. At the end of each year the Met Office makes a global mean temperature forecast for the coming year. Following on from the new record 2023, we predict a high probability of another record year in 2024 and a 35% chance of exceeding 1.5 C above pre-industrial. Whilst a one-year temporary exceedance of 1.5 C would not constitute a breech of the Paris Agreement target, our forecast highlights how close we are now to breeching this target. We show that our 2024 forecast can be largely explained by the combination of the continuing warming trend of +0.2 C/decade and the lagged warming affect of a strong tropical Pacific El Nino event. We further highlight 2023 was significantly warmer than forecast and that much of this warming signal came from the southern hemisphere and requires further understanding.

How to cite: Dunstone, N., Smith, D., Scaife, A., Hermanson, L., Colman, A., and Folland, C.: Will 2024 be the first year above 1.5 C?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10574, https://doi.org/10.5194/egusphere-egu24-10574, 2024.

The Mediterranean region has been identified as an important climate change hotspot, over the 21st century both air temperature and its extremes are projected to rise at a rate surpassing that of the global average and a significant decrease of average summer precipitation is projected, particularly for the western Mediterranean. On average, Mediterranean droughts have become more frequent and intense in recent years and are expected to become more widespread in many regions. These prolonged dry spells pose a substantial threat to agriculture and impact several socio-economic sectors. In this context, long-range weather forecasting has emerged as a promising tool for seasonal drought risk assessment. However, the interpretation of the forecasting products is not always straightforward due to their inherent probabilistic nature. Therefore, a rigorous evaluation process is needed to determine the extent to which these forecasts provide a fruitful advantage over much simpler forecasting systems, such as those based on climatology. 

In this study, we use the latest version of ECMWF’s seasonal prediction system (SEAS5) to understand its skill in predicting summer droughts. The Standardized Precipitation Evapotranspiration Index (SPEI) aggregated over different lead times is employed to mark below-normal dryness conditions in August. We use a comprehensive set of evaluation metrics to gain insight into the accuracy, systematic biases, association, discrimination and sharpness of the forecast system. Our findings reveal that up to 3 months lead time, seasonal forecasts show stronger association and discrimination skills than the climatological forecast, especially in the Southern Mediterranean, although the prediction quality in terms of accuracy and sharpness is limited. On the other hand, extending the forecast range up to 6 months lead time dramatically reduces its predictability skill, with the system mostly underperforming elementary climatological predictions. 

This approach is then extended to examine the full ensemble of seasonal forecasting systems provided by the Copernicus Climate Change Service (C3S) to test their skill in predicting droughts. Our findings can help an informed use of seasonal forecasts of droughts and the development of related climate services.

How to cite: Cerato, G., Bellomo, K., and von Hardenberg, J.: Summer drought predictability in the Mediterranean region in seasonal forecasts, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11485, https://doi.org/10.5194/egusphere-egu24-11485, 2024.

Reliable probabilistic information at the seasonal time scale is essential across various societal sectors, such as agriculture, energy, or water management. Current applications of seasonal predictions rely on General Circulation Models (GCMs) that represent dynamical processes in the atmosphere, land surface, and ocean while capturing their linear and nonlinear interactions. However, GCMs come with an inherent high computational cost. In an operational setup, they are typically run once a month and at a lower temporal and spatial resolution than the ones needed for regional applications. Moreover, GCMs suffer from significant drifts and biases and can miss relevant teleconnections, resulting in low skill for particular regions or seasons. 

In this context, the use of generative AI methods that can model complex nonlinear relationships can be a viable alternative for producing probabilistic predictions with low computational demand. Such models have already demonstrated their effectiveness in different domains, i.e. computer vision, natural language processing, and weather prediction. However, although requiring less computational power, these techniques still rely on big datasets in order to be efficiently trained. Under this scenario, and with sufficiently high-quality global observational datasets spanning at most 70 years, the research trend has evolved into training these models using climate model output. 

In this work, we build upon the work presented by Pan et al., 2022, which introduced a conditional Variational Autoencoder (cVAE) to predict global temperature and precipitation fields for the October to March season starting from July initial conditions. We adopt several pre-processing changes to account for different biases and trends across the CMIP6 models. Additionally, we explore different architecture modifications to improve the model's performance and stability. We study the benefits of our model in predicting three-month anomalies on top of the climate change trend. Finally, we compare our results with a state-of-the-art GCM (SEAS5) and a simple empirical system based on the linear regression of classical seasonal indices based on Eden et al., 2015.

Pan, Baoxiang, Gemma J. Anderson, André Goncalves, Donald D. Lucas, Céline J.W. Bonfils, and Jiwoo Lee. 'Improving Seasonal Forecast Using Probabilistic Deep Learning'. Journal of Advances in Modeling Earth Systems 14, no. 3 (1 March 2022). https://doi.org/10.1029/2021MS002766.

Eden, J. M., G. J. van Oldenborgh, E. Hawkins, and E. B. Suckling. 'A Global Empirical System for Probabilistic Seasonal Climate Prediction'. Geoscientific Model Development 8, no. 12 (11 December 2015): 3947–73. https://doi.org/10.5194/gmd-8-3947-2015.

How to cite: Palma, L., Peraza, A., Duarte, A., Civantos, D., Materia, S., Nandi, A., Peña-Izquierdo, J., Tufis, M., Vilella, G., Romero, L., Soret, A., and Donat, M.: A global empirical system for probabilistic seasonal climate prediction based on generative AI and CMIP6 models , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11930, https://doi.org/10.5194/egusphere-egu24-11930, 2024.

The Earth has warmed significantly over the past 40 years, and the fastest rate of warming has occurred in and around the Arctic. The warming of northern high latitudes at a rate of almost four times the global average (Rantanen et al., 2022), known as Arctic amplification, is associated with sea ice loss, glacier retreat, permafrost degradation, and expansion of the melting season. Since the mid-2000s, summer sea ice has exhibited a rapid decline, reaching record minima in September sea ice area in 2007 and 2012. However, after the early 2010s, the downward trend of minimum sea ice area appears to decelerate (Swart et al., 2015; Baxter et al., 2019). This apparent slowdown and the preceding acceleration in the rate of sea ice loss are puzzling in light of the steadily increasing rate of greenhouse gas emissions of about 4.5 ppm yr⁻¹ over the past decade (Friedlingstein et al., 2023) that provides a constant climate forcing. Recent studies suggest that low-frequency internal climate variability may have been as important as anthropogenic influences on observed Arctic sea ice decline over the past four decades (Dörr et al., 2023; Karami et al., 2023). Here, we investigate how unusual this decade-long pause in Arctic summer sea ice decline is within the context of internal climate variability. To do so, we first assess how rare this is deceleration of Arctic sea ice loss is by comparing it to trends in CMIP6 historical simulations. We also use simulations from the Decadal Climate Prediction Project (DCPP) contribution to CMIP6 to determine if initializing decadal prediction systems from estimates of the observed climate state substantially improves their performance in predicting the slowdown in Arctic sea ice loss over the past decade. As the DCPP does not specify the data or the methods to be used to initialize forecasts or how to generate ensembles of initial conditions, we also assess how different formulations affect the skill of the forecasts by analyzing differences between models. This work provides an opportunity to attribute this pause in Arctic sea ice retreat to interannual internal variability or radiative external forcings, something that observation analysis alone cannot achieve.

How to cite: DeRepentigny, P., Massonnet, F., Bilbao, R., and Materia, S.: How unusual is the recent decade-long pause in Arctic summer sea ice retreat?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12969, https://doi.org/10.5194/egusphere-egu24-12969, 2024.

Compound dry and hot events, characterized by elevated temperatures and reduced precipitation, pose interconnected challenges to human social economics, necessitating comprehensive strategies for mitigation and adaptation. This study focuses on the Pacific-Japan (PJ) pattern, a significant climate variability influencing summer climates in East Asia. While previous research has explored its impact on Japan and Korea, our investigation delves into its effects on Taiwan, a mountainous subtropical island with a population of approximately 24 million. Utilizing long-term temperature and rainfall data, along with reanalysis dynamic downscaling datasets, we examine the interannual impacts of the PJ pattern on summer temperature and compound heat and dry events. Our findings reveal a significant temperature increase during the positive phase of the PJ pattern, characterized by anticyclonic anomalous circulation over Taiwan. Additionally, both the Standardized Precipitation Index and soil water exhibit a decline during this phase, reflecting meteorological and hydrological drought conditions. A robust negative correlation (-0.7) between drought indices and temperature emphasizes the compound effect of heat and dry events during the PJ positive phase. This study enhances the understanding of the PJ pattern as a climate driver, describing its role in hot and dry summers over Taiwan. The insights gained, when integrated into seasonal prediction and early warning systems, can aid vulnerable sectors in preparing for potential heat and dry stress hazards.

How to cite: Lin, S.-Y., Tseng, W.-L., Wang, Y.-C., and Lo, M.: Compound Heat and Dry Events Influenced by the Pacific–Japan Pattern over Taiwan in Summer, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14341, https://doi.org/10.5194/egusphere-egu24-14341, 2024.

Over the past century, Taiwan has gradually recognized the hazards posed by extreme heat events (EHT), prompting the development of mid-term adaptation strategies to address challenges in the coming decades. However, our understanding of decadal-scale temperature variations remains insufficient, requiring further research into influencing factors. Our study reveals the crucial role of the Pacific Meridional Mode (PMM) in modulating decadal-scale variations in summer temperatures in Taiwan. During the positive phase of PMM, warm sea surface temperature anomalies trigger an eastward-moving wave train extending into East Asia. This leads to the development of high-pressure circulations near Southeast Asia and Taiwan, enhancing the temperature increase. This mechanism has been reproduced in experiments using the Taiwan Earth System Model. Moreover, our study utilizes the calendar day 90th percentile of maximum temperature (CTX) as the threshold for extreme high-temperature events (EHT), while also employing the heatwaves magnitude scale (HWMS) as the criterion for defining heatwaves. During the positive phase of PMM, the frequency and duration of EHT increase, with variations observed across different regions. The overall intensity of heatwave events also strengthens, primarily due to extended durations. Notably, in a single city, this results in exposure of up to 800,000 person-days to EHT, presenting a tenfold increase compared to the annual effect observed in the long-term warming trend. These findings on the decadal-scale relationship between summer temperatures in Taiwan and PMM contribute to a deeper understanding of EHT and heatwaves events impacts, providing more nuanced insights for future regional strategies in mitigating heatwave disasters.

How to cite: Tsai, C.-T., Tseng, W.-L., and Wang, Y.-C.: Linkage between Temperature and Heatwaves in Summer Taiwan to the Pacific Meridional Mode, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14379, https://doi.org/10.5194/egusphere-egu24-14379, 2024.

Storm surges and elevated water levels regularly challenge coastal protection and inland water management along the low-lying coastline of the German Bight. Skillful seasonal-to-decadal (S2D) predictions of the local storm surge climate would be beneficial to stakeholders and decision makers in the region. While storm activity has recently been shown to be skillfully predictable on a decadal timescale with a global earth system model, surge modelling usually requires very fine spatial and temporal resolutions that are not yet present in current earth system models. We therefore propose an alternative approach to generating S2D predictions of the storm surge climate by training a neural network on observed water levels and large-scale atmospheric patterns, and apply the neural network to the available model output of a S2D prediction system. We show that the neural-network-based translation from large-scale atmospheric fields to local water levels at the coast works sufficiently well, and that several windows of predictability for the German Bight surge climate emerge on the S2D scale.

How to cite: Krieger, D., Brune, S., Baehr, J., and Weisse, R.: Exploring ML-based decadal predictions of the German Bight storm surge climate, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14688, https://doi.org/10.5194/egusphere-egu24-14688, 2024.

Atlantic meridional overturning circulation (AMOC) is one of the mechanisms for climate predictability and one of the properties that decadal climate predictions are attempting to predict. The starting point for AMOC decadal predictions is sensitive to the underlying data assimilation and/or initialization procedure. This means that different choices during the data assimilation procedure (e.g., assimilation method, assimilation window, data sources, resolution, nudging terms and strength, full field vs anomaly initialization/assimilation, etc) can result in a different mean and even variability of reconstructed ocean circulation. How coherent the AMOC initial states should be among the CMIP-like decadal prediction experiments? How good in general should the initial AMOC be for decadal predictions? And do initialization issues of the ocean circulation influence the prediction skill of other variables that are of interest for application studies? These are the questions that we were attempting to address in our study, where we analyzed twelve decadal prediction systems from the World Meteorological Organization Lead Centre for Annual-to-Decadal Climate Prediction project. We identify that the AMOC initialization influences the quality of predictions of the subpolar gyre (SPG). When predictions show a large initial error in their AMOC, they usually have low skill for predicting the internal variability of the SPG five years after the initialization.

How to cite: Polkova, I. and the Co-Authors: Initialization shock in the ocean circulation reduces skill in decadal predictions of the North Atlantic subpolar gyre, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15358, https://doi.org/10.5194/egusphere-egu24-15358, 2024.

Using a large ensemble of initialised retrospective forecasts (hindcasts) from a seasonal prediction system, we explore various statistics relating to sudden stratospheric warmings (SSWs). Observations show that SSWs occur at a similar frequency during both El Niño and La Niña northern hemisphere winters. This is contrary to expectation, as the stronger stratospheric polar vortex associated with La Niña years might be expected to result in fewer of these extreme breakdowns. We show that this similar frequency may have occurred by chance due to the limited sample of years in the observational record. We also show that in these hindcasts, winters with two SSWs, a rare event in the observational record, on average have an increased surface impact. Multiple SSW events occur at a lower rate than expected if events were independent but somewhat surprisingly, our analysis also indicates a risk, albeit small, of winters with three or more SSWs, as yet an unseen event.

How to cite: Ineson, S., Dunstone, N., Scaife, A., Andrews, M., Lockwood, J., and Pang, B.: Statistics of sudden stratospheric warmings using a large model ensemble, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15476, https://doi.org/10.5194/egusphere-egu24-15476, 2024.

How to cite: Bachelery, M.-L., Brajard, J., Patacchiola, M., and Keenlyside, N.: Predicting Atlantic and Benguela Niño events with deep learning , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15709, https://doi.org/10.5194/egusphere-egu24-15709, 2024.

The German Meteorological Service's (Deutscher Wetterdienst DWD) climate predictions website  (www.dwd.de/climatepredictions) offers a centralized platform for accessing post-processed climate predictions, including subseasonal forecasts from ECMWF's IFS and seasonal and decadal predictions from the German climate prediction system. The website design was developed in collaboration with various sectors to ensure uniformity across all time frames, and users can view maps, tables, and time series of ensemble mean and probabilistic predictions in combination with their skill. The available data covers weekly, 3-month, 1-year, and 5-year temperature means, precipitation sums and soil moisture for the world, Europe, Germany, and particular German regions. To achieve high spatial resolution, the DWD used the statistical downscaling method EPISODES. Moreover, within the BMBF project KIMoDIs (AI-based monitoring, data management and information system for coupled forecasting and early warning of low groundwater levels and salinisation) the DWD provides climate prediction data of further hydrological variables (e.g. relative humidity) with corresponding prediction skill on a regional scale.

However, all predictions on these time scales can suffer from inherent systematic errors, which can impact their usefulness. To address these issues, the recalibration method DeFoReSt was applied to decadal predictions, using a combination of 3^rd order polynomials in lead and start time, along with a boosting model selection approach. This approach addresses lead-time dependent systematic errors, such as drift, as well as inaccuracies in representing long-term changes and variability.

This study highlights the improved accuracy of the recalibration approach on decadal predictions due to an increased polynomial order compared to the original approach, and its different impact on global and regional scales. It also explores the feasibility of transferring this approach to predictions with shorter time horizons of the provided variables.

How to cite: Pasternack, A., Mannig, B., Paxian, A., Hoff, A., Pankatz, K., Lorenz, P., and Früh, B.: Recalibrating DWD’s operational climate predictions: towards a user-oriented seamless climate service, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-15974, https://doi.org/10.5194/egusphere-egu24-15974, 2024.

The Mediterranean region is highly sensitive to climate change, having experienced an intense warming and drying trend in recent decades, primarily due to the increased concentrations of anthropogenic greenhouse gases. In the context of decision-making processes, there is a growing interest in understanding the near-term climate evolution of this region.

In this study, we explore the climatic fluctuations of the Mediterranean region in the near-term range (up to 10 years ahead) using two different products: projections and decadal predictions. The former are century-scale climate change simulations initialized from arbitrary model states to which were applied anthropogenic and natural forcings. A major limitation of climate projections is their limited information regarding the current state of the Earth’s climate system. Decadal climate predictions, obtained by constraining the initial conditions of an ensemble of model simulations through a best estimate of the observed climate state, provide a better understanding of the next-decade climate and thus represent an invaluable tool in assisting climate adaptation.

Using retrospective forecasts from eight decadal prediction systems contributing to the CMIP6 Decadal Climate Prediction Project (CMIP6 DCPP) and the corresponding ensemble of non-initialized projections, we compare the capabilities of the state-of-the-art climate models in predicting future climate changes of the Mediterranean region for some key quantities so as to assess the added value of initialization. 

Beyond the contribution of external forcings, the role of internal variability is also investigated since part of the detected predictability arises from internal climate variability patterns affecting the Mediterranean. The observed North Atlantic Oscillation, the dominant climate variability pattern in the Euro-Atlantic domain, as well as its  impact on wintertime precipitation over Europe are well reproduced by decadal predictions, especially over the Mediterranean, outperforming projections. We also apply a sub-sampling method to enhance the respective signal-to-noise ratio and consequently improve precipitation skill over the Mediterranean.

How to cite: Nicolì, D., Gualdi, S., and Athanasiadis, P.: Decadal predictions outperform projections in forecasting winter precipitation over the Mediterranean region, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16366, https://doi.org/10.5194/egusphere-egu24-16366, 2024.

The current generation of seasonal forecast models struggle to skilfully predict dynamical circulation over the North Atlantic and European region in boreal summer.  Using two different state-of-the-art seasonal prediction systems, we show that tropical rainfall anomalies drive a circulation signal in the North Atlantic/Europe via the propagation of Rossby waves.  The wave, however, is shifted eastwards compared to observations, so the signal does not contribute positively to model skill.  Reasons for the eastward shift of the Rossby wave are investigated, as well as other drivers of the signal in this region.  Despite the errors in the waves, the fact that seasonal forecast models do predict dynamical signals over the North Atlantic/Europe signifies seasonal predictability over this region beyond the climate change trend, and understaning the cause of the errors could lead to skilful predictions.

How to cite: Lockwood, J., Dunstone, N., Fröhlich, K., Fuentes Franco, R., Maidens, A., Scaife, A., Smith, D., and Thornton, H.: Investigating signals in summer seasonal forecasts over the North Atlantic/European region, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16985, https://doi.org/10.5194/egusphere-egu24-16985, 2024.

In this presentation, we introduce the WISER MENA projects SeaFOAM (Seasonal Forecasting Across MENA) and SeaSCAPE (Seasonal Co-Production and Application in MENA). These projects explore both the improvement to the regional-level seasonal forecast in the MENA region, as well as how to tailor the information in ways useful to a range of climate information stakeholders. SeaFOAM works alongside Maroc Meteo, Morocco's National Meteorological and Hydrological Service (NMHS) and the Long Range Forecasting node of the Northern Africa WMO Regional Climate Centre (RCC), to develop a framework for objective seasonal forecasting. This approach will blend techniques such as bias correction via local linear regression and canonical correlation analysis (CCA), with skill-assessed sub-selected models, to improve forecasting accuracy. Multiple drivers of rainfall variability, including the North Atlantic Oscillation (NAO) and Mediterranean Oscillation (MO), are investigated for their calibration potential. SeaSCAPE works with the WMO and various partners across MENA to understand the use of seasonal information in multiple sectors, exploring existing gaps and needs. Through stakeholder engagement workshops, training and bespoke support for the Arab Climate Outlook Forum (ArabCOF), SeaSCAPE operates collaboratively to tailor regional and national-level climate information to improve accessibility and usability of climate information on seasonal timescales.

How to cite: Lines, S., Savage, N., Parfitt, R., Colman, A., Chamberlain-Clay, A., Norris, L., Howard, H., and Ticehurst, H.: Strengthening seasonal forecasting in the Middle East & North Africa (MENA) through the WISER Programme., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17418, https://doi.org/10.5194/egusphere-egu24-17418, 2024.

Accurate forecasts of the natural resources of renewable energy production have become not only a valuable but a crucial tool for managing the associated risks of specific events, such as wind droughts. Wind energy, alongside with solar power, now provide a substantial part to the renewable energy share of the global energy production and growth in this sector will most likely further increase. The naturally given fluctuations of wind resources, however, pose a challenge for maintaining a stable energy supply, which, at the end of the chain, can have an impact on the energy market prices.
Operational short-term forecasting products for the wind energy sector (multiple days) are already commonly available and seasonal to sub seasonal forecasting solutions (multiple months) can provide valuable skill and are gaining in popularity. On the other side of the spectrum, typically on a time scale of multiple decades, we find risk assessment based on climate change projections. In between the long and short term time scales, however, there is a gap that still needs to be filled to achieve seamless prediction of risks that are relevant for the energy sector: decadal predictions.

Here, we present the results of an evaluation study of a multi-model decadal prediction ensemble (DCPP) for a selection of wind development regions in Europe. The evaluation is based on multiple decades long hindcasts and carried out with a focus on the skill of predicting specific event types of wind resource availability in a probabilistic context, alongside with basic deterministic skill measures. We further investigate specific event constellations and their large-scale drivers that, in combination, can provide windows of opportunity with enhanced predictive skill. We conclude with a discussion on how this hybrid approach can be used to potentially increase not only forecast skill but also the trust of the end user.

How to cite: Lochbihler, K., Lopez, A., and Lizcano, G.: Skill of wind resource forecasts on the decadal time scale, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17585, https://doi.org/10.5194/egusphere-egu24-17585, 2024.

Seasonal prediction is a field of research attracting growing interest beyond the scientific community due to its strong potential to guide decision-making in many sectors (e.g. agriculture and food security, health, energy production, water management, disaster risk reduction) in the face of the pressing dangers of climate change.

Among the various techniques being considered to improve the predictive skill of seasonal prediction systems, increasing the horizontal resolution of GCMs is a promising avenue. There are several indications that higher resolution versions of the current generation of climate models might improve key air-sea teleconnections, decreasing common biases of global models and improving the skill to predict certain regions at seasonal scales, e.g. in tropical sea surface temperature.

In this study, we analyze the differences in the predictive skill of two different seasonal prediction systems, based on the same climate model EC-Earth3 and initialized in the same way but using two different horizontal resolutions. The standard (SR) and high resolution (HR) configurations are based on an atmospheric component, IFS, of ~100 km and ~40 km of resolution respectively and on an ocean component, NEMO3.6, of ~100 km and ~25 km respectively. We focus in particular on the Tropical Pacific region where statistically significant improvements are found in HR with respect to SR for predicting ENSO and its associated climate teleconnections. We explore some processes that can explain these differences, such as the simulation of the tropical ocean mean state and atmospheric teleconnections between the Atlantic and Pacific tropical oceans. 

A weaker mean-state bias in the HR configuration, with less westward extension of ENSO-related SST anomalies, leads to better skill in ENSO regions, which can also be linked to better localization of the atmospheric teleconnection with the equatorial Atlantic Ocean. It remains to be assessed if similar improvements are consistently identified for HR versions in other forecast systems, which would prompt their routine use in seasonal climate prediction.

How to cite: Carreric, A., Ortega, P., Lapin, V., and Doblas-Reyes, F.: Comparing the seasonal predictability of Tropical Pacific variability in EC-Earth3 at two different horizontal resolutions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19229, https://doi.org/10.5194/egusphere-egu24-19229, 2024.

Accurate forecasts of the climate of the coming season and years are highly desired by many sectors of society. The skill of near-term climate prediction in winter in the North Atlantic and European region has improved over the last decade associated with larger ensembles, improving models and boosting of the prediction signal using intelligent post processing. International collaboration has improved the availability of forecasts and promoted the uptake of forecasts by different sectors. However, significant challenges remain, including summer prediction, understanding the risk of extremes within a season, multi-seasonal extremes and how best to post process the forecasts to aid decision making. This talk will summarise recent near-term climate prediction research activities at the UK Met Office and will detail our experience of providing such forecasts to the energy and water sectors.  

How to cite: Thornton, H.: The opportunities and challenges of near-term climate prediction, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19251, https://doi.org/10.5194/egusphere-egu24-19251, 2024.

This study focuses on applying machine learning techniques to bias-correct the seasonal temperature forecasts provided by the Copernicus Climate Change Service (C3S) models. Specifically, we employ bias correction on forecasts from five major models: UK Meteorological Office (UKMO), Euro-Mediterranean Center on Climate Change (CMCC), Deutscher Wetterdienst (DWD), Environment and Climate Change Canada (ECCC), and Meteo-France. Our primary objective is to assess the performance of our bias correction model in comparison to the original forecast datasets. We utilise temperature-based indices recommended by the Expert Team on Climate Change Detection and Indices (ETCCDI) to evaluate the effectiveness of the bias-corrected seasonal forecasts. These indices served as valuable metrics to gauge the predictive capability of the models, especially in forecasting natural cascading hazards such as wildfires, droughts, and floods. The study involved an in-depth analysis of the bias-corrected forecasts, and the derived indices were crucial in understanding the models' ability to predict temperature-related extreme events. The results of this research contribute valuable information for decision-making and planning across various sectors, including disaster risk management and environmental protection. Through a comprehensive evaluation of machine learning-based bias correction techniques, we enhance the accuracy and applicability of seasonal temperature forecasts, thereby improving preparedness and resilience to climate-related challenges. 

How to cite: Mbuvha, R. and Nikraftar, Z.: Machine Learning Approaches to Improve Accuracy in Extreme Seasonal Temperature Forecasts: A Multi-Model Assessment , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19297, https://doi.org/10.5194/egusphere-egu24-19297, 2024.

The ground temperature variability in the Northern Hemisphere winter is greatly influenced by the state of the polar vortex. When the vortex collapses during sudden stratospheric warmings (SSWs), rapid changes in stratospheric circulations propagate downward to the troposphere in the subsequent weeks. The ground effect following SSWs is typically manifested as the negative phase of the North Atlantic Oscillation. Our findings reveal a higher frequency of cold temperature anomalies in the Northern part of Eurasia during winters with SSWs, and conversely, warm anomalies in winters with a strong and stable vortex. This behavior is particularly evident when temperature anomalies are categorized into three equal subgroups, or terciles. Recently, we developed a statistical model that successfully predicts SSW occurrences with an 86% accuracy rate. The model utilizes the stratospheric Quasi-Biennial Oscillation (QBO) phase and two parameters associated with solar activity: the geomagnetic aa-index as a proxy for energetic particle precipitations and solar irradiance. In this study, we explore the model's potential to provide a seasonal forecast for ground temperatures. We assess the probabilities of regional temperature anomalies falling into the lowest or highest terciles based on the predicted weak or strong vortex state. Additionally, we demonstrate that the QBO phase further enhances the forecast quality. As the model provides SSW predictions as early as preceding August, our results carry significant societal relevance as well, e.g., for the energy sector, which is highly dependent on prevailing weather conditions.

How to cite: Vokhmianin, M., Salminen, A., Mursula, K., and Asikainen, T.: Seasonal forecast of the late boreal winter temperature based on solar forcing and QBO, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19359, https://doi.org/10.5194/egusphere-egu24-19359, 2024.

Most global climate models with convective parameterization have trouble in simulating the observed diurnal cycle of convection. Maximum precipitation usually happens too early during local summertime, especially over land. Observational analyses indicate that deep convection over land cannot keep pace with rapid variations in convective available potential energy (CAPE), which is largely controlled by boundary layer forcing. In this study, a new convective closure in which shallow and deep convection interact strongly, out of equilibrium, is implemented in atmosphere-only and ocean-atmosphere coupled models developed at the NOAA Geophysical Fluid Dynamics Laboratory (GFDL). The diurnal cycles of convection in both simulations are significantly improved without altering their mean states. These improvements in the diurnal cycle of these climate models are consistent with those obtained by Peter Bechtold and colleagues in the ECMWF Integrated Forecasting System. The new closure shifts maximum precipitation over land later by about three hours. Compared to satellite observations, the diurnal phase biases are reduced by half. Shallow convection to some extent equilibrates rapid changes in the boundary layer at sub-diurnal time scales. Future model improvement will focus on the remaining biases, which may be further reduced by including stochastic entrainment and cold pools.

How to cite: Zhang, B., Donner, L., Zhao, M., and Tan, Z.: Improved Diurnal Cycle in GFDL Earth System Models with Non-Equilibrium Convection, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-33, https://doi.org/10.5194/egusphere-egu24-33, 2024.

Mesoscale Convective Systems (MCSs) that become large or have long lifespans contribute disproportionately to extreme rainfall. Gaining a better understanding of the factors that determine whether a system will become large could improve our understanding of extreme weather phenomena. The recent emergence of high-resolution global simulations from the DYAMOND project, coupled with a storm tracking algorithm called TOOCAN, provides a groundbreaking opportunity to study the factors controlling the maximum area of MCSs. In this study we use machine learning algorithms to predict the maximum area of convective systems based on their early development stages and initial environmental conditions. The results reveal that the initial evolution of the system anticipates its maximum area. Factors such as the presence of ice in the system's environment, proximity to surrounding systems, intensity of vertical velocity at 500 hPa, and the migration distance, have been identified as significant factors in improving the accuracy of the prediction. Using a linear model, we investigate the relative role of the environment and of the system itself, in the growth of the system. 

How to cite: Muller, C., Abramian, S., Risi, C., Roca, R., and Fiolleau, T.: Is the fate of Mesoscale Convective Systems written from the start?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1875, https://doi.org/10.5194/egusphere-egu24-1875, 2024.

Future changes in tropical convection will be closely tied to changes in the underlying sea surface temperature (SST) pattern. To understand the convective response to warming in a coupled atmosphere-ocean system, we perform a series of idealized, 20-year radiative-convective equilibrium experiments with a 2D cloud-resolving model coupled to a 25-m slab ocean. The domain length is that of the tropical Pacific basin, and different climates are achieved by varying the parameterized ocean heat transport (q-flux). The simulations are characterized by two distinct regimes of  convection-SST coupling: an oscillatory regime that occurs when the mean SST is near that of the present-day tropical Pacific (27-30 °C), and a non-oscillatory regime at warmer temperatures (>36 °C).

The oscillatory regime is defined by internal, 3°C oscillations in mean SST driven by variations in low cloudiness. During the warming phase of the cycle, SSTs are homogeneous, deep convection occurs in two regions, and low clouds are sparse. During the cooling phase, there are well-defined warm and cold pools, deep convection aggregates into a single region, and expansive low cloud decks act to decrease the mean SST.  

In the warmer, non-oscillating regime, distinct warm and cold pools still form, but convection is no longer limited to the warmest SSTs. Rather, convection develops over cooler SSTs and is then advected to the warm pool by the mean flow. The expansion of deep convection to cooler SSTs impedes low cloud formation over the cold pool and inhibits the low cloud-driven oscillations in mean SST. Changes in sub-cloud buoyancy explain the expansion of the convectively unstable region.

Both regimes (oscillatory and non-oscillatory) can be achieved for the same q-flux depending on initial conditions. Intermediate SSTs (30-36 °C) are unstable on long timescales and eventually revert to one regime or the other. While certain aspects of this behavior are likely sensitive to simulation design, our broader set of experiments suggests potential shifts in convection-SST coupling as the climate warms.

How to cite: Sokol, A., Munteanu, V., and Hartmann, D.: Internal variability, multiple equilibria, and convection-SST coupling in a cloud-resolving model with an interactive ocean, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3218, https://doi.org/10.5194/egusphere-egu24-3218, 2024.

During the Australian fire season 2019/2020, an unprecedented amount of smoke aerosol was not only released, but also transported upwards and injected into the tropopause region by so-called pyro-cumulonimbus clouds (pyroCb). The resulting lower stratospheric aerosol loads in early 2020 were comparable to those of the largest volcanic eruptions of the twentieth century. PyroCbs have been identified as the main pathway for biomass burning aerosol into the stratosphere. To study the phenomenon of PyroCbs, simulations of the so-called Australian New Year Super Outbreak are performed with the numerical weather model ICON. Simulations were run in a nested, limited area mode setup, with the smallest domain reaching down to a horizontal grid spacing of 500 m. Within the domain, an idealised fire perturbation was applied for which an additional constant surface sensible heat and water vapour flux was introduced to represent the thermodynamical impacts of the fire. Simulations with this setup were successful in producing fire-induced deep convection with subsequent smoke injection into the lower stratosphere. Preliminary sensitivity experiments show a high sensitivity of the PyroCb properties to initial and boundary conditions. We can show, that especially water vapour emissions, which would originate from evaporating surface water as well as from combustion of organic materials, have a decisive, enhancing impact on the pyro-convection. Moreover, besides the fire intensity, the plume characteristics and smoke injection heights are also closely linked to the background meteorology, in particular. In the long term, the goal is to incorporate the effects of extreme biomass burning emission into large scale climate simulations by taking into account PyroCb activity. However, this will require a very deep understanding of wildfire triggered convection and PyroCb dynamics.  

How to cite: Müller, J., Senf, F., and Tegen, I.: Modelling the formation of an extreme Australian pyro-convection event and its sensitivities, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3972, https://doi.org/10.5194/egusphere-egu24-3972, 2024.

The elimination of rain evaporation in the planetary boundary layer (PBL) has been found to lead to convective self-aggregation (CSA) even without radiative feedback (frequently referred to as “moisture memory aggregation”), but the precise mechanisms underlying this phenomenon remain unclear. We conducted cloud-resolving simulations with two domain sizes (L = 128 and 256 km; Δx = 1 and 4 km) with homogenised radiation and progressively reduced rain evaporation in the PBL by multiplying it with a factor 𝛼 = [1.0, 0.8, 0.6, 0.4, 0.2, 0]. Surprisingly, self aggregation only occurred when rain evaporation was almost completely removed (𝛼 ≈ 0). Similar to conventional radiatively-driven aggregation (RDA), a shallow circulation that leads to an upgradient moist static energy transport is present, but in this case it is the additional convective heating resulting from the reduction of evaporative cooling in the moist patch that triggers this circulation, thereafter a dry subsidence intrusion into the PBL in the dry patch takes over and intensifies aggregation. Hence, this type of aggregation should be more appropriately referred to as “convectively-driven aggregation” (CDA). Contrary to RDA, in CDA temperature and moisture anomalies oppose each other in their buoyancy effects, hence explaining the need for near-zero 𝛼 values: only when rain evaporation is almost completely removed can the additional heating trigger aggregation. Lastly, we found radiative cooling and not cold pools to be the leading cause of the domain size dependence of CDA. Runs with similar amounts of cold pools aggregate in the large but not small domain due to stronger radiative cooling rates and concomitant broadening of the range of precipitable water in the larger domain. 

How to cite: Hwong, Y.-L. and Muller, C.: The Unreasonable Efficiency of Total Rain Evaporation Removal in Triggering Convective Self-Aggregation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4139, https://doi.org/10.5194/egusphere-egu24-4139, 2024.

Convection frequently initiates in proximity to cold fronts during the European warm-season and can also be associated with hazards such as flooding, rain, and hail. Despite this, the frequency and underlying processes that drive such events are not well-understood. To understand the typical nature, frequency and forcing mechanisms of convection depending on the region relative to the front, automatic front detection methods, a convective cell detection and tracking dataset (KONRAD), and lightning data are combined between 2007–2016.

The climatology shows that convective cells are most frequent in Germany marginally ahead of the surface front. Furthermore, the 700 hPa frontal line marks the minimum frequency of convection and a shift in regime between cells with a strong diurnal cycle on the cold-side of the 700 hPa front and a weakened diurnal cycle on the warm-side of the 700 hPa front. The results are consistent for lightning data on a sub-European domain. Given cell detection ahead of the surface front, cells are up to 3 times more likely to be associated with a mesocyclone compared to non-cold-frontal cells in Germany. Cells with 55 dBZ cores are over 1.5 times more likely.

To unravel the complex relationships between different predictor variables and the probability of convection a logistic regression model is developed. Feature importance techniques are utilised to understand which variables carry the most importance depending on the region relative to the front. We find solar heating carries more importance towards the model’s predictive power behind the 700 hPa front than ahead of the 700 hPa front. The opposite is true for the elevation term, which acts as a proxy for the influence of orography on convective initiation. By giving the model information on the number of surrounding grid points associated with convection, a proxy for cell interactions, the most skill is added near the surface front.

These results are an important step towards a deeper understanding of the underlying processes that drive cold-frontal convection and improved forecasting.

How to cite: Pacey, G., Pfahl, S., Schielicke, L., and Wapler, K.: Climatology, characteristics and forcing mechanisms of warm-season cold-frontal convection in Europe, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5759, https://doi.org/10.5194/egusphere-egu24-5759, 2024.

Seeley and Wordsworth (2021) showed that in small-domain cloud-resolving simulations the temporal pattern of precipitation transforms in extremely hot climates (≥ 320 K) from quasi-steady to organized episodic deluges, with outbursts of heavy rain alternating with several dry days. They proposed a mechanism for this transition involving increased water vapor greenhouse effect and solar radiation absorption leading to net lower-tropospheric radiative heating. This heating inhibits lower-tropospheric convection and decouples the boundary layer from the upper troposphere during the dry phase, allowing lower-tropospheric moist static energy to build until it discharges, resulting in a deluge. We perform cloud-resolving simulations in polar night and show that the same transition occurs, implying that some revision of their mechanism is necessary. We perform further tests to show that episodic deluges can occur even if the lower-tropospheric radiative heating rate is negative, as long as the magnitude of the upper-tropospheric radiative cooling is about twice as large. We find that in the episodic deluge regime the period can be predicted from the time for radiation and reevaporation to cool the lower atmosphere.

How to cite: Song, X., Abbot, D., and Yang, J.: Critical role of vertical radiative cooling contrast in triggering episodic deluges in small-domain hothouse climates, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5781, https://doi.org/10.5194/egusphere-egu24-5781, 2024.

Spontaneous aggregation of deep convection is a common feature of idealized numerical simulations of the tropical atmosphere in Radiative-Convective Equilibrium (RCE). However, at coarse grid resolution where deep convection is not fully resolved, the occurrence of this phenomenon is highly sensitive to subgrid-scale processes. This study investigates the role of mixing and entrainment, provided by either the turbulence model or the implicit numerical dissipation, in this phenomenon. The results of two different models, WRF and SAM, have been analysed and compared using different configurations by varying the turbulence models, initial conditions, and horizontal spatial resolution. At a coarse grid resolution of 3 km, the occurrence of Convective Self-Aggregation (CSA) is prevented in models with low numerical diffusivity due to the removal of turbulent mixing, while it is preserved in models with high numerical diffusivity. When refining the horizontal grid resolution to 1 km, which reduces the implicit numerical dissipation, CSA can only be achieved by increasing explicit turbulent mixing. Even with a small amount of shallow clouds, CSA was found to occur in this case. Therefore, this study suggests that the sensitivity of CSA to horizontal grid resolution is not primarily due to the corresponding decrease in shallow clouds. It has been found that the amplitude of initial humidity perturbations introduced by convection in the free troposphere is regulated by turbulent mixing and dissipation at small scales. The size and strength of humidity perturbations in the free troposphere that can destabilize the RCE state increase with greater dissipation at small scales.

How to cite: Silvestri, L., Saraceni, M., and Bongioannini Cerlini, P.: Numerical diffusion and turbulent mixing in convective self-aggregation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5959, https://doi.org/10.5194/egusphere-egu24-5959, 2024.

Improving weather and climate prediction cannot avoid accurately representing organized convection, as its convective and stratiform components distinctly reshape large-scale circulations via redistributing momentum and heat. For latent heating, the stratiform heating in organized convection shifts to higher altitudes compared to convective regions, presenting a significant challenge for representation in models across scales. The Multiscale Coherent Structural Parameterization (MCSP), introduced by Moncrieff et al. (2017), offers a promising solution by generating the top-heavy profile from convective heating in slantwise layer overturning scenarios. As part of the MCS: PRIME project, the PRIME-MCSP implementation by Zhang et al. (submitted, 2024) couples MCSP with the CoMorph-A convection scheme in the UK Met Office Unified Model with the following improvements: 1) CoMorph permits unstable air to rise from any height, diverging from the conventional CAPE trigger for deep convection, thereby enhancing continuity and facilitating storm tracking. 2) We activate MCSP selectively for deep mixed-phase clouds, recognizing the limited ability of shallow clouds to produce a stratiform component. 3) We configure the global model runs to include both a fixed convective-stratiform heating fraction and a fraction proportional to cloud top temperature.

MCS tracks in ensembles of weather runs show that PRIME-MCSP suppresses cloud deepening and reduces precipitation areas by dampening low-level updrafts. 20-year climate simulations show that PRIME-MCSP improves the precipitation seasonal cycle over the Indian Ocean, while increasing the warm-season wet bias over the Western Pacific. Additionally, PRIME-MCSP intensifies the Madden Julian Oscillation (MJO). The model run using a variable convective-stratiform fraction more accurately represents the MJO frequency and aligns better with reanalysis. Future plans focus on the stochastic representation of stratiform effects, steered by insights from data assimilation increments.

How to cite: Zhang, Z., Christensen, H., Muetzelfeldt, M., Woollings, T., Plant, B., Stirling, A., Whitall, M., Moncrieff, M., and Chen, C.-C.: Improving and Assessing Organized Convection Parameterization in the Unified Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8367, https://doi.org/10.5194/egusphere-egu24-8367, 2024.

Organised thunderstorm clusters known as Mesoscale Convective Systems (MCSs) can bring high impact hazards such as flash floods, lighting and destructive winds. It is crucial for the forecasting and mitigation of these hazards to understand the processes that influence the characteristics of storms and thereby contribute to extreme events. Soil moisture is known to influence the initiation of MCSs in several regions of the world, but the influence of soil moisture on the later stages of MCS lifecycles is less well understood. Work in West Africa has revealed that dry soil moisture structures on scales > 200 km can increase the scale and longevity of propagating, mature afternoon MCSs, but this has not been investigated for other regions. In the current work we simultaneously analyse seven global MCS hotspot regions where storms may be sensitive to soil moisture, the US Great Plains, China, India, West Africa, Australia, South Africa and South America, to gain a more global perspective of the impact of soil moisture conditions on mature MCS characteristics. Using a combination of global datasets, storm tracks, satellite data, reanalysis data and CMIP6 simulations, we reveal that large-scale soil moisture gradients (100s of km) can intensify storms by driving favourable shear conditions through the strengthening of low-level atmospheric temperature gradients. By separating storms by soil moisture conditions, we show an increase in precipitation feature area and rainfall production on days with favourable gradients compared to days with unfavourable gradients. This is a newly identified mechanism through which soil moisture can influence storm hazards globally, which has implications for the forecasting and future projection of extreme events under climate change.

How to cite: Barton, E., Klein, C., Taylor, C., Marsham, J., Parker, D., Maybee, B., Feng, Z., and Leung, L. R.: Storm intensification driven by soil moisture gradients in global hotspot regions, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8856, https://doi.org/10.5194/egusphere-egu24-8856, 2024.

Representing land-ocean heterogeneity via convective
adjustment timescale
Bidyut Goswami¹ , Andrea Polesello¹ , Caroline Muller¹ .
¹Department of Earth Science, Institute of Science and Technology Austria, Klosterneuburg, Austria
January 2024

Abstract

The time needed by deep convection to bring the atmosphere back to equilibrium
is called convective adjustment timescale or simply adjustment timescale, typically
denoted by τ . In the Community Atmospheric Model version 6 (CAM6), convection
is parameterized through the Zhang-McFarlan scheme [1], where CAPE undergoes
an exponential consumption, of which τ is the time constant. τ is a tunable pa-
rameter in CAM6 and it has a default value of 1 hour, worldwide, on both ocean
and land. Albeit, there is no justified reason why one adjustment timescale value
should work over land and ocean both. Continental and oceanic convection is dif-
ferent in terms of the vigor of updraft and hence can have different durations.[2, 3]
So it is logical to investigate the prescription of two different convective adjustment
timescales for land (τ_L ) and ocean (τ_L ). To understand the impact of representing
land-ocean heterogeneity via τ , we investigated CAM climate simulations for two
different convective adjustment timescales for land and ocean in contrast to having
one value globally.
Following a comparative analysis of 5-year-long climate simulations, we find
τ_O =4hr and τ_L =1hr to yield the best results. In particular, we obtain a better
description of the Madden-Julian Oscillation (MJO). Although these τ values were
chosen empirically and require further tuning, the conclusion of our finding remains
the same, which is, to use two different τ values for land and ocean.
References
[1] G. Zhang and N. A. McFarlane, “Sensitivity of climate simulations to the parameterization of
cumulus convection in the canadian climate centre general circulation model,” Atmosphere-
Ocean, vol. 33, no. 3, pp. 407–446, 1995.
[2] C. Lucas, E. J. Zipser, and M. A. Lemone, “Vertical Velocity in Oceanic Convection off
Tropical Australia,” Journal of the Atmospheric Sciences, vol. 51, pp. 3183–3193, 11 1994.
[3] R. Roca, T. Fiolleau, and D. Bouniol, “A Simple Model of the Life Cycle of Mesoscale
Convective Systems Cloud Shield in the Tropics,” Journal of Climate, vol. 30, pp. 4283–
4298, 6 2017.

How to cite: Polesello, A., Goswami, B. B., and Muller, C.: Representing land-ocean heterogeneity via convective adjustment timescale, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9561, https://doi.org/10.5194/egusphere-egu24-9561, 2024.

Within a diurnal cycle, the transition from shallow to deep convection takes several hours, despite having large environmental instability at the onset of shallow convection. During this period, the cloud environment remains rather steady, while the convection exhibits a rapid development. Properly predicting the timing of this rapid shallow-to-deep transition within a diurnal cycle is still a major shortcoming of weather and climate models that employ the so-called mass-flux parameterization of atmospheric convection, as they typically predict the onset of deep convection too early, not allowing for a gradual convective deepening. In this work, it is argued that the problem of correctly representing the diurnal cycle of deep convection comes from the fundamental assumptions of the mass-flux formulation, in which it is considered that the clouds, represented by steady-state plumes, only interact with a spatially homogeneous environment. However, in the rapid shallow-to-deep transition, the convection still requires several hours to deepen, even if the environment remains steady, so some interactions must be missing. Here, a conceptual model for cloud development is introduced, in which a cloud is formed due to the sum of water transport from the boundary layer by multiple updrafts during its life-time, allowing for cloud-cloud interactions. This process captures local preconditioning, in which the clouds themself provide favorable conditions for the development of subsequent updrafts. It is also argued that the cold pools act as a reinforcement of this process, organizing the updrafts, and thus, allowing for a greater degree of local preconditioning. Based on this new conceptual model, it is argued that the shallow-to-deep transition can be seen as a predator-prey problem, in which the cloud population at the cloud base acts as prey, while the surface precipitation rate acts as predators. This simple predator-prey model is then tested against an idealized large-eddy simulation, showing that indeed, the rapid shallow-to-deep transition of atmospheric convection exhibits predator-prey characteristics. Moreover, it is shown how easily the simple predator-prey model can be implemented in current mass-flux schemes, leading to improved representation of deep convection within a diurnal cycle. Overall, this suggests that better representing the spatial organisation of clouds can lead to improvements in the timing of cloud and precipitation properties, thanks to a better convective memory.

How to cite: Vraciu, C.-V., Savre, J., and Colin, M.: Predator-prey characteristics of the rapid shallow-to-deep transition of atmospheric convection, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10474, https://doi.org/10.5194/egusphere-egu24-10474, 2024.

Deep convection over Amazonia can manifest in various forms, from scattered convective cells to mesoscale organizations like squall lines and cloud clusters. This diversity significantly influences vertical convective transport, impacting not only large-scale circulation but also the poorly understood cycle of gases and aerosols emitted by the forest. Monitoring convective systems over Amazonia during the CAFE-Brazil field campaign (Dec 2022-Jan 2023) involved the HALO aircraft and the ATTO-Campina ground site, employing meteorological, aerosol, and chemical measurements.

On January 18, a 500-km wide mesoscale system dissipated, giving rise to new convective cells initially disorganized and later organized into a large squall line. This event was measured by ATTO-Campina and HALO during the local afternoon. To understand the processes driving organizational changes and their impact on transport, 24-hour simulations with the Meso-NH model were conducted over an 800-km wide domain, ranging from horizontal resolutions of 1600 m down to 200 m, ultimately resulting in large-eddy simulations.

The simulations revealed a strong resolution sensitivity in mesoscale convective organization, with a distinct emergence of squall lines at the finest resolutions only. Surprisingly, at fine resolution, organized convection exhibited larger transport due to increased updraft size, rather than intensity. Cloud cluster organization exhibited a delayed onset compared to convective cell organization, aligning with expectations. Ongoing investigations are currently focusing on gravity waves and cold pools to better understand their impact on convective organization.

How to cite: Dauhut, T., Gonthier, H., Viala, B., and Haytzmann, G.: Organization of convection over Amazonia and its impact on transport, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10894, https://doi.org/10.5194/egusphere-egu24-10894, 2024.

Satellite infrared (IR) cloud imagery has proven valuable in the identification of Pyrocumulonimbus (pyroCb) clouds. The substantial brightness temperature difference observed between warm shortwave IR wavelengths (~4 μm) and window IR wavelengths (~11 μm) has served as a reliable marker for detecting daytime pyroCb. However, this indicator becomes ineffective during nocturnal hours when the enhanced brightness temperature at 4 μm is solely a daytime phenomenon, arising from PyroCb microphysics that increase solar reflectivity of clouds. We have developed a machine learning model designed to detect pyroCb events during nighttime using IR channels from the Advanced Baseline Imager (ABI) aboard GOES-16. The model leverages the distinctive characteristics of daytime IR channels as its training data. We applied the trained model to five intense pyroCb events in western North America during August 2017. Furthermore, we have employed an established cloud-tracking tool known as Tracking and Object-Based Analysis of Clouds (tobac) to analyze the evolution of the clouds plumes and infer their lifetimes. Our research aims to extend this case study on a global scale, with the objective of creating a comprehensive database for the lifetimes of pyroCb events. Such a database will enhance our understanding of pyroCb dynamics, which is helpful for investigating the radiative implications and the potential impact on stratospheric chemistry.

How to cite: Liu, F. and da Silva, A.: Detecting diurnal cycle and lifetime of pyrocumulonimbus using GOES-16 infrared data with a machine learning model , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12581, https://doi.org/10.5194/egusphere-egu24-12581, 2024.

The launch of the joint ESA JAXA Earth Cloud Aerosol and Radiation Explorer (EarthCARE) mission (May, 2024) marks the beginning of a new era of spaceborne radar measurements that target atmospheric convection. In addition to the EarthCARE mission that features the first Cloud Profiling Radar (CPR) with Doppler capability, NASA’s Investigation of Convective Updrafts (INCUS) and Atmosphere Observing System (AOS) missions aim to provide unique observations of convective dynamics. Prior to this upcoming decade of the study of atmospheric convection from space, the CloudSat CPR collected remarkable data of convective cores over a period of 15 years. Despite its high frequency that results in significant attenuation and multiple scattering effects, the 94-GHz CloudSat CPR offers a relatively small footprint (compared to the TRMM/GPM radar footprint of 5 km) and collocated radar-radiometer (passive) brightness temperatures (Tb). Here, we propose a refined deep convective core (DCC) identification scheme by first selecting the CPR profiles with continuous echoes between below 2 and above 10 km. The 10-dBZ echo top height is also required to exceed 10 km and located within 2 km from cloud top. Additionally, profiles with stratiform precipitation flags in the CloudSat products are not included in the analysis.

We investigated the CloudSat observations from 2006 to 2019 globally and also with a focus over 4 convective basins where model simulations are performed by the NASA’s INCUS science team. The four deep convection basins are Amazon, Congo, Philippines, and Western Pacific, which represent a decent spectrum of atmospheric environments. It is found that the DCCs over the Congo basin are featured with larger size and likely more intensified updrafts, while the Western Pacific is characterized with finer-scale cores. The analysis shows that the DCCs with size below 5 km predominate, implying the narrow cores can be under detected by the large-footprint radars such as GPM. The distinct depressions of 94-GHz Tb due to the presence of high-density ice particles lend complementary information on DCC classifications. In addition, multiple scattering can be a confounding factor in interpretating the CPR measurements within deep convective clouds. Our preliminary calculations suggest the impact of multiple scattering becomes significant at ~2.5 km from radar cloud top on average and is subject to the DCC updraft intensity. Moreover, profiles of 94-GHz radar reflectivity and Tb are forward calculated from the high-resolution model simulation outputs to understand the constraints that such observations can afford on key measures such as convective mass fluxes.

How to cite: Xu, Z., Kollias, P., Battaglia, A., Treserras, B. P., and Marinescu, P.: Investigating Deep Convective Cores Combining CloudSat Observations and Model Simulations , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13870, https://doi.org/10.5194/egusphere-egu24-13870, 2024.

Over continental plains, precipitation tends to peak in the late afternoon or during nighttime. The accurate simulation of the land precipitation diurnal cycle in GCMs has been a long-standing challenge. Nighttime surface cooling tends to yield a stable layer with large convective inhibition (CIN). However, CIN arises from traditional parcel considerations—measuring the inhibition for an infinitesimal parcel. Here, we argue that the CIN layer is less effective in inhibiting convection than previously thought for convective entities of typical horizontal cloud size.

A time-dependent process model for anelastic convective entities (ACE) is formulated to consistently include dynamic entrainment/detrainment as well as a representation of nonhydrostatic perturbation pressure. Spatially nonlocal effects mediated by the pressure field imply that horizontal feature size becomes a factor in the vertical conditional instability problem. ACE simulations using nighttime GoAmazon soundings with strong surface inversion demonstrate that the vertically nonlocal pressure response and its interaction with the surface boundary condition make the CIN layer ineffective for convective features of substantial horizontal size. Within the convective column, buoyancy of different signs offset each other via the nonlocal interaction over vertical scales comparable to the typical horizontal scale. Furthermore, the interaction with the surface tends to downweight the effectiveness of negative buoyancy contributions at low levels. This implies that a much smaller vertical velocity perturbation (or more generally, nonlocal buoyancy forcing from neighboring disturbances) can tunnel through the CIN layer. The same effect yields smaller magnitude for the mass flux above the CIN layer compared with steady plume models. 

A related implication of including spatially nonlocal interactions is that the vertical acceleration due to deep-convective buoyancy tends to extend above the level of neutral buoyancy (LNB). This results in cloud top much higher than the LNB, exhibiting the convective cold-top feature previously noted in observations. Results here point to revision for convective parameterizations. 

How to cite: Kuo, Y.-H., Tan, Z., Zhao, M., and Neelin, J. D.: Why can nighttime convection occur despite strong convective inhibition?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14373, https://doi.org/10.5194/egusphere-egu24-14373, 2024.

Data-driven approaches using machine learning to parameterizing model physical processes in Earth System Models have been actively explored in recent years. Deep-learning-based convection parameterization is one such example. While significant progress has been made in emulating convection using neural networks (NN), serious roadblocks remain, including generalization of the NN-based scheme trained on model data from current climate to future climate and integration instability when it is implemented into the model for long-term integrations. This study uses a deep residual convolutional network to emulate convection simulated by a superparameterized global climate model (GCM). The NN uses the current environmental state variables and advection tendencies, as well as the history of convection to predict the GCM grid-scale temperature and moisture tendencies, cloud liquid and ice water contents from moist physics processes. Independent offline tests show that the NN-based scheme has extremely high prediction accuracy for all output variables considered. In addition, the scheme trained on data in the current climate generalizes well to a warmer climate with +4K sea surface temperature in an offline test, with high prediction accuracy as well. Further tests on different aspects of the NN architecture are performed to understand what factors are responsible for its generalization ability to a warmer climate. We are also able to perform multi-year integrations, without encountering any integration instability, when the scheme is implemented into the NCAR CAM5. The details will be presented at the meeting.

How to cite: Zhang, G., Han, Y., and Wang, Y.: Using Deep Learning for Convection Parameterization, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14599, https://doi.org/10.5194/egusphere-egu24-14599, 2024.

Tropical precipitation extremes and their changes with surface warming are investigated using global storm resolving simulations and high-resolution observations. The simulations demonstrate that the spatial organization of convection at mesoscale, a process that cannot be physically represented by conventional global climate models, is important for the variations of tropical daily precipitation extremes (total accumulations over a day). In both the simulations and observations, daily precipitation extremes increase in a more organized state, in association with larger, but less frequent, storms. Repeating the simulations for a warmer climate results in a robust increase in monthly-mean daily precipitation extremes. Higher precipitation percentiles have a greater sensitivity to convective organization, which is predicted to increase with warming. Without changes in organization, the strongest daily precipitation extremes over the tropical oceans increase at a rate close to Clausius-Clapeyron (CC) scaling. Thus, in a future warmer state with increased organization, the strongest daily precipitation extremes over oceans increase at a faster rate than CC scaling. Moreover, as the precipitation distribution becomes more uneven with increased organization, the tropics may not only face heavier precipitation extremes, but experience more extensive drying.

How to cite: Bao, J., Stevens, B., Kluft, L., and Muller, C.: Intensification of daily tropical precipitation extremes from more organized convection, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-14717, https://doi.org/10.5194/egusphere-egu24-14717, 2024.

As the atmosphere gets warmer, it is expected to hold more water vapor, thereby fueling stronger storms. At the same time, the condensation of this vapor increases the combined load of liquid water and ice aloft, forcing convection to do more work to lift water to the level where it precipitates. This takes away from the generation of kinetic energy, thereby creating a “brake” on atmospheric motions. The evolution of this precipitation “brake” with warming determines the magnitude of future storm intensification, with important societal implications. The new generation of kilometer-scale climate models is capable of projecting this evolution. In this presentation, we show how the NOAA/GFDL X-SHiELD experimental global storm-resolving model can be used to estimate the total mechanical work done by convection and the work done to lift water which is then subsequently dissipated by friction during precipitation. The statistics are computed in year-long simulations of the present climate and of a 4K warmer climate.

We find that the ratio of kinetic energy generation vs work spent to lift water is respectively 30% vs 70% of the total mechanical work done by convection on global average, with a relative stability across regions and in the present vs future climate.

Moving beyond regional averages, when we organize the space by decreasing values of dissipation, we find that the ratio of work spent to lift water to total mechanical work strongly increases in the most convective percentiles, that is, most of the work done by convection is used to lift water in the extremes, showing that water loading strongly opposes kinetic energy generation. We also find that the total work done by convection, the work spent to lift water and the precipitation-induced dissipation all increase similarly with warming in the most convective percentiles. This suggests that, as the Earth warms, the updrafts tend to “kill” themselves in situ from increased water loading instead of generating a response at larger scale.

How to cite: Bolot, M., Pauluis, O., Harris, L., Cheng, K., Merlis, T., Clark, S., Kaltenbaugh, A., Zhou, L., and Fueglistaler, S.: Increase of a precipitation “brake” to stronger storms in kilometer-scale global warming simulations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16757, https://doi.org/10.5194/egusphere-egu24-16757, 2024.

Mesoscale convective systems are the building block of tropical precipitation, as more than 40% of global precipitation and more than 80% of extreme rainrates are produced by these organized systems. However, when investigating the sensitivity of global rain extremes, the behavior and morphology of organized storm systems are typically ignored and corresponding dynamics are instead interpreted using the textbook framework of a convecting parcel. Indeed, despite rich observational and case studies describing the internal dynamics and structures of MCSs, no conceptual framework exist to this day to bridge the gap between global hydrologic sensitivity and MCS behavior.

This work introduces new approaches to link extreme precipitation rates in the tropics to the occurrence, internal dynamics and lifecycle of individual MCSs. Individual storms are idenditifed based on by the Lagrangian tracking algorithm TOOCAN which tracks storm anvils over their lifecycle, and which has been applied to satellite observations and to global storm resolving models in the DYAMOND experiment. We first use this rich dataset to develop a numerical interface that maps the occurrence of extreme precipitation rate onto the MCS cloud shield. We then introduce a novel conceptual framework to decompose the sensitivity of precipitation extremes to the change in storm occurrence and change in internal dynamics within this cloud shield. 

Results are threefold. We demonstrate a robust phasing in the timing of global extreme rainrates within the storm lifecycle, robustly occurring at 25-30% of the storm's lifetime for the models and regions analyzed. The analytical decomposition confirms that in a given climate state, variability in the heaviest rainrates across regions mostly occur through changes in MCS frequency, rather than changes in their efficiency at producing rain. We finally argue that the sensitivity of extremes to climate state may occur through both a change in occurrence and a change in internal MCS dynamics.

How to cite: Fildier, B., Carenso, M., Roca, R., and Fiolleau, T.: Mapping km-scale global extreme rainfall onto mesoscale convective systems lifecycle, frequency and dynamics , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16951, https://doi.org/10.5194/egusphere-egu24-16951, 2024.

In this study, we compare convection characteristics in three models that are at the forefront of global km-scale modelling, the ICON model developed by the Max Planck Institute for Meteorology (MPI-M) and German Weather Service (DWD), the IFS developed by the European Centre for Medium-Range Weather Forecasts (ECMWF), and the NICAM model developed by the University of Tokyo, the Japan Agency for Marine-Earth Science and Technology (JAMSTEC) and the National Institute for Environmental Studies (NIES). For IFS and ICON, we analyse 1-year coupled simulations at 4.4 and 5 km resolution, respectively, which stem from Cycle 3 of the H2020 Next Generation Earth Modelling Systems (nextGEMS) project. For NICAM, we analyse a 1-year AMIP-type simulation at 3.5 km resolution. Convection schemes have been switched off in ICON and NICAM, while in the IFS the deep convection scheme’s cloud base mass flux is strongly reduced. 

Modelling convection at km-scale resolutions is both exciting and challenging because some important processes are already resolved at these scales (e.g., deep convection) but other important processes remain under-resolved (e.g., mixing of grid-scale updrafts with their environment). Thus, we analyse in this study what common issues exist in ICON, IFS and NICAM with respect to the convection characteristics in the tropics, in what respects all models do well and where there are substantial inter-model differences.

Specifically, we analyse local convection characteristics and show that compared to satellite observations, the models tend to overestimate precipitation intensity (NICAM and ICON), while they underestimate precipitation cell size and precipitation duration. We study mesoscale organisation by using different organisation metrics and show that the models tend to underestimate organisation, even though they all consistently show that when organisation is enhanced, heavy precipitation is enhanced as well. We also investigate moisture-convection relationships and show that the models generally do not moisten enough during a convective event compared to ERA5 reanalysis data. Consistently, the sensitivity of lower-tropospheric moisture variations to the life cycle of deep convection over ocean looks too weak in ICON and IFS.

Finally, we look at land-ocean differences of the convection characteristics and show that while all models capture the diurnal cycle of precipitation over ocean well, there are some substantial differences over land, even though biases are not consistent between the models. Over coastal regions of the Maritime Continent, ICON has too strong mean precipitation and a too strong diurnal cycle, whereas IFS overall underestimates both, connected to a too weak propagation of convection onto the ocean during nighttime, potentially connected to too weak cold pools. Meanwhile, NICAM has more realistic convection characteristics in these coastal regions.

How to cite: Becker, T., Takasuka, D., and Bao, J.: Characteristics of precipitating convection and moisture-convection relationships in global km-scale simulations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-17683, https://doi.org/10.5194/egusphere-egu24-17683, 2024.

Flash floods arising from short-duration precipitation extremes are costly for the population, and their frequency and intensity could increase with global warming (Fowler et al., 2021). Understanding the mechanisms leading to extreme precipitation is thus essential. A common hypothesis for precipitation extremes is that they scale with temperature according to the thermodynamic Clausius-Clapeyron (CC) law. However, increases in short-duration precipitation extremes beyond the CC expectation (or super-CC) were reported in multiple regions. The super-CC scaling is currently understood as the combination of two effects: (1) an invigoration of convective precipitation through convective cloud feedbacks; (2) a statistical effect resulting from a shift in rain type, from light stratiform to heavier convective-type precipitation, with increasing temperatures.

This work revisits these hypotheses by identifying convective precipitation at an unprecedented high resolution (5 km spatially and 10 min temporally). For this, we employ the EUropean Cooperation for LIghtning Detection (EUCLID) lightning dataset to define convective precipitation and combine it with weather station data from the German weather service (Deutscher Wetterdienst, DWD). We show that while (total) extreme precipitation increases with a super-CC rate, the scaling of both convective and stratiform-type precipitation extremes is in accordance with the CC law. We thus conclude that the super-CC rate is explained by the statistical shift in rain type alone and refute any mechanistic origin. 

Mesoscale Convective Systems (MCSs), which dominate extreme precipitation events in Europe (Da Silva & Haerter, 2023), are known to contain both a convective and stratiform region (Houze, 1997). By tracking MCSs over Germany, we show that MCS extreme precipitation also features a super-CC rate, which we relate to a dramatic increase in their convective fraction for dew point temperatures exceeding 14 degrees Celsius. 

References:

Da Silva, N. A., & Haerter, J. O. (2023). The precipitation characteristics of mesoscale convective systems over Europe. Journal of Geophysical Research: Atmospheres, 128, e2023JD039045. https://doi.org/10.1029/2023JD039045

Fowler, H.J., Lenderink, G., Prein, A.F. et al. Anthropogenic intensification of short-duration rainfall extremes. Nat Rev Earth Environ 2, 107–122 (2021). https://doi.org/10.1038/s43017-020-00128-6

Houze, R. A. Stratiform precipitation in regions of convection: A meteorological paradox? Bulletin of the American Meteorological Society 78, 2179 – 2196 (1997). https://doi.org/10.1175/1520-0477(1997)078<2179:SPIROC>2.0.CO;2

How to cite: Da Silva, N. and Haerter, J.: Convective precipitation extremes may not increase beyond the Clausius-Clapeyron expectation, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20249, https://doi.org/10.5194/egusphere-egu24-20249, 2024.

Geophysical convection is usually characterized by Reynolds number in the range typical for turbulent flow. Despite that, it displays features of organization.  
Thermal vortex rings are considered candidates for the basic elements of that order (Yano 2023, ch. 16).
In this work, the process of their formation from a spherical buoyancy anomaly is studied numerically. The buoyancy distribution is assumed to be uniform with a discontinuity at the interface.
The rising anomaly experiences a collapse at the bottom, and initially spherical shape is transformed into a torus. Neglecting diffusive processes, the system is uniquely defined by the vortex sheet coincident with the interface. For that reason, its evolution is considered on the grounds of vorticity dynamics with Lagrangian approach.
The vortex sheet is intensified by buoyancy and further subjected to Kelvin-Helmholtz instability. This starts in high wavenumbers increasing the effective thickness by purely advective mechanism. A similar instability is then launched in lower wavenumbers, and the phenomenon repeats hierarchically. As a result, the energy is transferred from small to large scales. The same mechanism also drives the interfacial mixing by applying stretching and folding repetitively. This makes it a good starting point for further studies on the entrainment rate and order emerging out of chaos. 

How to cite: Jędrejko, P. and Yano, J.-I.: Formation of thermal vortex rings, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-20512, https://doi.org/10.5194/egusphere-egu24-20512, 2024.

Mesoscale convective systems (MCSs), the largest type of deep convective storms are formed when convection aggregates and grows upscale, forming a distinct mesoscale circulation through the interaction of multiple storms. Thus, storms play an important role in MCS organization. Due to their large size, longer duration, and larger precipitation, MCSs cause high-impact extreme weather events like lightning, damaging hail, gusty winds, and flooding. During the Indian summer monsoon (June-September), synoptic-scale weather systems move across the monsoon zone (MZ), causing MCSs to form frequently. MCSs often produce widespread and heavy rain throughout the MZ. Hence, studies on structure and evolution of MCSs highlighting the organization of convection are needed for an improved understanding of MCS.

In this study, we used an object-based cloud-tracking method (Feng et al., 2018) to identify and track MCSs and embedded storms in remote sensing observations and numerical simulations. The work is divided into three parts. In the first part, we tracked MCSs over the monsoon zone using geostationary satellite infrared brightness temperature (IRTb) and GPM IMERG precipitation from June-September, 2014 to 2019 and examined various aspects of observed MCSs (n=2092) such as spatial coverage, diurnal cycle, rainfall amount, and land-ocean contrast. The majority of MCSs are positioned in the monsoon trough's southeast-northwest stretch and account for more than 60% of total precipitation. For MCSs with short and long lifespans, there was a clear land-ocean divide and varied lifecycle trends. Oceanic MCSs last longer, are deeper, and provide more rainfall over a larger area than land-based MCSs.

In the second part of the study, we explored embedded storm structures for those MCSs that exist within the radar domain (n=65) by applying a storm classification algorithm to the S-band Doppler radar observations during June-September 2015. We observed that an MCS contains many precipitation features, especially during early stages of development when multiple convective clusters begin to amalgamate. Furthermore, we investigated the co-evolution of numerous storm parameters (e.g., areas of convective/stratiform precipitation, convective core length, and top heights) as a function of MCS lifetime. Distinct vertical structures are observed for the convective, stratiform, and anvil components of MCSs.

In the third part of this work, we examine the ability of a convection-permitting Weather Research Forecast (WRF) model in simulating MCSs and their characteristics (initiation, size, intensity, lifetime, propagation) during June-September 2015. A similar cloud-tracking algorithm is applied to WRF-simulated data (reflectivity, IRTb, and precipitation) to identify and track MCS in the simulation. Although the model underestimated the number of observed MCSs, the composite evolution and frequency distribution of convective area, precipitation amount, MCS propagation speed produces reasonable agreement with observations but underestimate stratiform areas. Consistent with observations, the simulated MCS properties showed a gradual increase from convective initiation to around the first half of the MCS lifetime. We observed that an MCS contains multiple precipitation features, particularly during the initial development stage when multiple convective clusters begin to aggregate. More details on observed MCSs and embedded storm structures, as well as their representation in simulation, will be presented.

How to cite: Tupsoundare, M., Deshpande, S., Feng, Z., Das, S. K., Deshpande, M., and Hanmante, H.: Object-Based Analyses of Mesoscale Convective Systems and Embedded Storms over the Indian Monsoon Zone Using Datasets from Satellite, Radar and Model Simulations               , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-328, https://doi.org/10.5194/egusphere-egu24-328, 2024.

Ensemble sensitivity is a statistical tool applied within an ensemble that reveals the atmospheric flow features (e.g. position of a jet streak, or magnitude of a low-level moisture plume) at early forecast times that are related to a chosen forecast response later in the forecast window.  The response function is chosen to diagnose high-impact forecast features such as maximum updraft helicty over a specified area, or number of grid points of simulated reflectivity exceeding 40 dBZ in a chosen region.  Since ensemble sensitivity highlights the features early in a forecast important to the prediction of high-impact features later in the forecast, a subset of members with the smallest errors in sensitive regions can be chosen that might improve probabilistic forecasts of the response relative to the full ensemble. Similar to ensemble data assimilation, this process incorporates observational information to beneficially update forecast distributions.  The subsetting procedure can be done quickly once an ensemble has been run, and sensitivity-based subsets can typically be generated well before the next extended forecast can be run within a cycling storm-scale data assimilation and forecasting system. In turn, the subsetting procedure, if shown to improve forecasts, could be a unique and useful operational forecasting tool.

Ensemble sensitivity-based subsetting has been tested within the Texas Tech University operational ensemble system in both an idealized framework and in more operational settings in real time during several years of the National Oceanic and Atmospheric Administration (NOAA) Hazardous Weather Testbed (HWT).  Response functions that diagnose severe convective hazards, such as updraft helicity, hail size, and simulated reflectivity have been tested to gain an understanding of both the general capability of the technique and the perception of forecasters regarding its value in a real-time forecasting environment.  Here we discuss this effort and its associated results, the technique’s current status, and future plans toward ultimate operational implementation.

How to cite: Ancell, B. and Coleman, A.: Ensemble Sensitivity-Based Subsetting for Convection: Progress Toward Operational Use, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1300, https://doi.org/10.5194/egusphere-egu24-1300, 2024.

Applications of Machine Learning (ML) in the different stages of weather forecasting have considerably developed recently. Such progress is likely to change the landscape and offer new perspectives to speed up and improve forecast performances, at different spatio-temporal scales. In this context, Météo-France engaged more actively in this new area of research, with the objective to further explore the capabilities and opportunities of ML for operational forecasting. Major ongoing projects include ML to significantly enhance the size of convective-scale ensemble forecasts, high-resolution statistical downscaling and the development of data-driven kilometre-scale forecasting systems. Early results will be presented and our short-term roadmap will be discussed.

How to cite: Raynaud, L., Brochet, C., and Moldovan, G.: ML for weather prediction at Météo-France : current status and future plans, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1675, https://doi.org/10.5194/egusphere-egu24-1675, 2024.

A major aspiration of operational and research meteorology is to relate the average behaviour of convective-scale flows to the more predictable, larger-scale flows in which they occur. This goal is difficult, partly because convective flows often self-organize at mesoscales, with the dynamics of such mesoscale convective systems (MCSs) distinct from those at convective and synoptic scales. In this study we use a tracking algorithm to detect MCSs in Australian operational radar data, revealing regional, seasonal and sub-seasonal, i.e. synoptic, differences in organizational characteristics. Restricting to MCS observations with nominally two-dimensional mean system-relative flows, spatio-temporal organizational differences are generally well explained by theoretical ideas regarding the breakdown of two-dimensional overturning flows. Theoretically, breakdown is characterised by a single non-dimensional convective Richardson number R, which provides the ratio of thermodynamic potential energy to inflow kinetic energy. Specifically, 76% of MCS relative trailing-stratiform, up-shear tilted observations, nominally associated with primarily non-overturning system-relative flows, occur when R>5, whereas 72% of relative leading-stratiform, down-shear tilted observations, nominally indicating primarily overturning system-relative flows, occur when R<5. Spatiotemporal variations in observed organizational characteristics are broadly consistent with spatiotemporal variations in median R. These results likely have implications for convective parametrisation, and operational convective permitting model testing and development.

How to cite: Short, E. and Lane, T.: Mesoscale Convective Systems across Australia, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1936, https://doi.org/10.5194/egusphere-egu24-1936, 2024.

The Advanced Geostationary Radiation Imager (AGRI) onboard the FY-4A geostationary satellite provides high spatiotemporal resolution visible reflectance data since March 12th, 2018. Data assimilation experiments under the framework of observing system simulation experiment have shown great potential of these data to improve the forecasting skills of numerical weather prediction (NWP) models. To effectively assimilate the AGRI data, it is important to address the quality the observations. In this study, the FY-4A/AGRI channel 2 (0.55 μm - 0.75 μm) reflectance was evaluated by the equivalents derived from the short-term model forecasts of the China Meteorological Administration Mesoscale Model (CMA-MESO) using the Radiative Transfer for TOVS (RTTOV, v 12.3). It is shown that the observation minus background (O – B) statistics could be used to reveal the abrupt changes related to the measurement calibration processes. In addition, O - B statistics are negatively biased. Potential causes include measurement errors, the unresolved processes, forward-operator errors, etc. The relative mean biases of O-B computed for cloud-free and cloudy pixels were used to correct the systematic differences for cloudy and clear pixels separately. Results indicate that the bias correction method could effectively reduce the biases and standard deviations of O-B. In addition, an ensemble forecast has advantages over a deterministic forecast in correcting the biases in FY-4A/AGRI visible reflectance data. The finding suggests an effective method to monitor the performance of FY-4A/AGRI visible measurements and to correct the biases in the observations. 

How to cite: Zhou, Y., Liu, Y., Zeng, Y., and Han, W.: Evaluation of FY-4A/AGRI visible reflectance using the equivalents derived from the forecasts of CMA-MESO using RTTOV, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-2793, https://doi.org/10.5194/egusphere-egu24-2793, 2024.

The historic 22-26 May 2015 flood event in Texas and Oklahoma was caused by anomalous clustered mesoscale convective systems (MCSs) that produced record-breaking rainfall and $3 billion of damage in the region. A month-long regional convection-permitting simulation is conducted to reconstruct multiple clustered MCSs that lead to this flood event. We further use the pseudo global warming approach to examine how a similar event may unfold in a warmer climate and the driving physical factors for the changes. Tracking of MCSs in observations and simulations shows that the historical simulation reproduces the salient characteristics of the observed MCSs. In a warmer climate under a high-emission (SSP5-8.5) scenario, the Southern Great Plains is projected to experience a near surface warming of 4-6 K, accompanied by enhanced moisture transport by the strengthened Great Plains low-level jet. A warmer and moister lower troposphere leads to 36-59% larger convective available potential energy, supporting wider and more intense convective updrafts and rainfall production. Consistently, MCSs have wider convective areas and stronger rainfall intensities, producing 50% larger rain volumes during the mature stage. Extreme (99.5%) MCS rainfall frequency and amount will increase by threefold (Fig. 1). However, MCS stratiform rain area decreases as a result of elevated stratiform cloud bases that lead to stronger sublimation and evaporation of precipitation in response to warming, resulting in reduced weak-to-moderate surface precipitation. Results suggest that global warming greatly increases precipitation intensity of clustered MCS events under strong synoptic influence, with much higher potential to produce serious floods without additional climate adaptation.

Figure 1. (a) Frequency distribution of MCS grid-point hourly rain rates, and (b) normalized cumulative distribution of rainfall amount by hourly rain rates. The region of the data included is show in the inset.

How to cite: Feng, Z., Chen, X., and Leung, R.: How Might the May 2015 Flood in the U.S. Southern Great Plains Induced by Clustered MCSs Unfold in the Future?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3210, https://doi.org/10.5194/egusphere-egu24-3210, 2024.

This study examines the urban impacts associated with a developed city belt on generating an afternoon heavy rainfall event over a coastal developing city that is 70–100 km downwind from the city belt over the Yangtze River Delta region. Observational analyses show pronounced urban heat island (UHI) effects along the upstream city belt prior to convection initiation (CI). A series of cloud-permitting model simulations with the finest grid spacing of 1 km are performed to examine the impacts of urbanization on CI and the subsequent heavy rainfall event. Results reveal the generation of warm anomalies and low-level convergence in the planetary boundary layer along the upstream city belt, thereby inducing upward motion for CI. The southwesterly flows of the monsoonal warm-moist air, enhanced by the UHI effects along the city belt, allow the development of convective cells along the belt. Some of the cells merge during their downstream propagation, promoting to the ultimate generation of the distinct heavy rainfall centers in favor of local convective clusters over the coastal city where atmospheric columns are more moist and potentially unstable under the influences of sea breezes. Sensitivity simulations show small contribution of the downstream city but more influences from the upstream city belt on the heavy rainfall event. The above findings help elucidate how the UHI effects could assist the CI in a weak-gradient environment, and explain why urbanization can contribute to increased downwind mean and extreme precipitation under the influences of favorable regional forcing conditions. These findings have been published in Monthly Weather Review.

How to cite: Jiang, X., Zhang, D.-L., and Luo, Y.: Influences of urbanization on an afternoon heavy rainfall event over the Yangtze River Delta region in East China, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3714, https://doi.org/10.5194/egusphere-egu24-3714, 2024.

The impact of urbanization and the sensitivity of urban canopy parameters (UCPs) on a typical summer rainfall event in Hangzhou, China, is investigated using three groups of ensemble experiments. In this case, urbanization leads to higher temperatures, lower mixing ratios, lower wind speeds before precipitation, and more precipitation in and around the urban area. Both the thermal and dynamical effects of urbanization contribute to an increase in temperature and precipitation, with thermal effects contributing 71.2% and 63.8% to the temperature and precipitation increase, respectively, while the thermal and dynamical impacts cause the opposite changes to the mixing ratio and wind speed. Compared to the other three meteorological elements, the model has the largest uncertainty in the simulation of precipitation, which includes the sensitivity of the different parameterization schemes to the simulation of precipitation in urban areas, and the uncertainty brought by the urban effect on precipitation is not confined within the city but extends to the surrounding areas as well. Temperature and mixing ratio are more sensitive to thermal-related UCPs, while the wind speed is mainly affected by the structural parameters. These variations, however, are sometimes contradictory to precipitation changes, which further adds to the complexity of precipitation simulation.

How to cite: Wu, M., Dong, M., Chen, F., and Yang, X.: Impacts of Urbanization and Its Parameters on Thermal and Dynamic Fields in Hangzhou: A Sensitivity Study Using the Weather Research and Forecasting Urban Model, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-3735, https://doi.org/10.5194/egusphere-egu24-3735, 2024.

Convection-permitting numerical weather prediction (NWP) is crucial for forecasting high-impact weather events such as heavy precipitation, storms, floods, wind gusts and fog. The assimilation of observations plays a significant role in improving the forecasting skill of these weather events. To make better use of existing observations and guide the design of future observation networks, accurately assessing the influence of assimilated observations is essential. The degrees of freedom for signal (DFS) has long been used to assess the influence of observations on the analysis. While various methods exist for calculating the DFS in variational data assimilation (DA) systems, calculating the DFS in ensemble-based DA systems (e.g., the ensemble transform Kalman filter) is a largely unexplored area. Since ensemble-based DA systems are becoming increasingly dominant for convection-permitting NWP, practical implementation of the DFS in such DA systems is needed. Unlike in variational DA systems, the background error covariance matrix is not static in ensemble-based DA methods. Consequently, the DFS calculated at each assimilation step measures the observation influence for a certain background error covariance matrix. This means that the DFS estimates are flow dependent. In addition, domain localisation of observations is often used in ensemble-based DA systems (e.g., local ensemble transform Kalman filter). This implies that the DFS should be calculated locally. In this work, we propose novel approaches for calculating the DFS in ensemble-based DA systems and investigate existing approaches applicable to such systems. We establish their consistency under idealised conditions and discuss their differences in practical applications. To validate our theoretical findings, we conduct simple numerical experiments using JEDI (Joint Effort for Data assimilation Integration) developed by JCSDA (Joint Center for Satellite Data Assimilation).  Our results provide useful information for assessing the influence of observations in ensemble-based DA systems. This work is financially supported by the Met Office and is fully in line with the Met Office’s strategy and its ongoing development of the next generation data assimilation and observation processing system.

How to cite: Hu, G., Dance, S., Fowler, A., Simonin, D., and Waller, J.: Assessing the influence of observations on the analysis in ensemble-based data assimilation systems, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4082, https://doi.org/10.5194/egusphere-egu24-4082, 2024.

Frequent air pollution episodes pose severe health and environmental challenges in Tehran, Iran. Despite recent efforts, pollutant levels often exceed WHO-based national standards. This study addresses the pressing need for accurate air quality prediction by leveraging advanced satellite data and machine learning techniques. Our methodology integrates Sentinel-5P satellite data with optical depth remote sensing information. We systematically evaluated five machine learning algorithms to identify the most effective approach for AQI prediction. This study aims to advance air quality prediction in Tehran by integrating Sentinel-5P satellite data with machine learning algorithms. We examined the efficacy of various algorithms, including Decision Tree, K-Nearest Neighbors, Random Forest, Support Vector Machine, and Logistic Regression, in correlating air pollutant levels with the Air Quality Index (AQI). The selection criteria focused on algorithmic efficiency and accuracy in handling diverse environmental datasets. The Random Forest algorithm, utilizing Sentinel-5P and optical depth data, achieved a remarkable accuracy of 74% in predicting AQI. Further enhancement was observed by incorporating climatic data, COVID-19 status, and environmental parameters; the model achieved a significant predictive accuracy of up to 75.6%. These findings underscore the critical impact of nitrogen dioxide, ozone, and aerosol optical depth on Tehran's AQI, with notable variations observed post-COVID-19 restrictions. The increase in AQI following the lifting of COVID-19 restrictions suggests a significant correlation between human activity and air quality. These insights can inform targeted environmental policies in Tehran. We demonstrate the potential of integrating satellite data with machine learning to predict AQI accurately. Our approach offers a scalable model for urban air quality management with implications for environmental policy and public health initiatives.

How to cite: Kafi, A. M., Hosseinipoor, M., Zare Shahne, M., and Jamaat, A.: Integrating Sentinel-5P Satellite Data and Machine Learning Algorithms for Air Quality Index Prediction in Tehran: A Comprehensive Study on Factors Influencing Air Quality, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4506, https://doi.org/10.5194/egusphere-egu24-4506, 2024.

How to cite: Jyoti, K., Weissmann, M., Griewank, P., and Meier, F.: Testing Hybrid-3DEnVar in the convective scale NWP model AROME-Austria, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5305, https://doi.org/10.5194/egusphere-egu24-5305, 2024.

Mesoscale convective systems (MCSs) are crucial in modifying the water cycle and frequently induce high-impact weather events over eastern China. Radar and Climate Prediction Center (CPC)-4 km satellite-derived infrared cloud top temperature (Tb) data were used to thoroughly analyze the long-term climatology of MCSs over eastern China, particularly in the Yangtze–Huaihe River Basin (YHR) in the warm season from 2013 to 2018. For the first time, we contrasted the effects of data set selection and threshold setting on research outcomes. The large-scale environments of MCSs initiation were also investigated using the latest global reanalysis data ERA5. It is found that striction of thresholds, including duration, reflectivity/Tb, area, and linearity, would lead to a greater proportion of early-morning MCSs. Satellite-identified MCSs differed from radar-derived ones, exhibiting afternoon diurnal peaks, faster movement speeds, longer travel distances, and expansive impact areas. The center of MCS and related precipitation shifted northward from Pre-Meiyu to Post-Meiyu seasons, contributing to up to 20% of total rainfall, with most MCSs moving along eastward trajectories. MCSs typically had the most substantial impact in the Meiyu season because of the most prolonged duration, largest convective core area, and strongest precipitation intensity. Warm-season MCSs initiated ahead of midlevel troughs and were related to strong anomalous low-level convergence and midlevel upward. The circulation anomalies were the strongest in the Pre-Meiyu season among the three subseasons, with most moisture sourced from the southwest.

How to cite: Lu, Y., Tang, J., Xu, X., Tang, Y., and Fang, J.: Characteristics of Warm‐Season Mesoscale Convective Systems Over the Yangtze–Huaihe River Basin (YHR): Comparison Between Radar and Satellite, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5399, https://doi.org/10.5194/egusphere-egu24-5399, 2024.

Using the FengYun (FY) satellite products and hourly rain gauge data, the east-moveing regional rainfall events (RREs) with long duration and large areas originated in the northeastern Tibetan Plateau (TP) were identified. Our findings reveal that 70% of heavy and long-duration(≥6h) RREs originating in the northeastern TP have the potential to move a thousand kilometers eastward during the warm-season. We noted distinct differences in the speed and spatial location of rainfall for the two types of eastward-moving RREs under investigation.. For the long-distance eastward-moving RREs, three local enhancements of precipitation centers, corresponding to the center moving out of 105°E, 110°E and 115°E are evident. In contrast, for the short-distance eastward-moving RREs, the precipitation centers mainly reach the second topographical terrace without further eastward moving. The evolution of mid-level trough and upper troposphere warm anomalies are closely related to the eastward-moving RREs. With the eastward movements of middle troposphere trough, coupled with the synergistic effects of the convergence and a change in wind orientation at the lower level, and the divergence at the upper-level, collectively contribute to the long-distance eastward moving RREs. The short-distance eastward moving RREs, influenced by the ridge of western Pacific subtropical high over North China and the low-level anomalous anticyclone, remains west of 110°E. This study offers an in-depth understanding of how upstream precipitation events influences the downstream rainfall.

How to cite: Chen, H. and He, M.: The characteristics of eastward-moving regional rainfall events originating in the northeastern of Tibetan plateau, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5848, https://doi.org/10.5194/egusphere-egu24-5848, 2024.

The hydro-climate of the Sahel is dominated by organised Mesoscale Convective Systems (MCSs), which typically bring intense rain every few days during the West African Monsoon season. MCSs leave a swath of wet soil often hundreds of kilometres across, which in turn create strong spatial patterns of surface fluxes of heat and water back into the atmosphere. Previous studies have shown that soil moisture patterns exert a strong control on the initiation and propagation of MCSs, significantly enhancing the predictability of convection on scales of 10 – 100s km. Here, we use satellite observations to examine how this strong, locally negative, soil moisture-precipitation feedback evolves and impacts rainfall patterns over a series of storms.

We track the response of the surface and atmosphere to over 5,000 MCS events from the period 2004-2020, using a combination of satellite-derived products (Land Surface Temperature; LST, soil moisture, Vegetation Optical Depth, rainfall, cloud-top temperature). Initial anomalies in LST and soil moisture weaken rapidly in the 3-4 days after the MCS, particularly in climatologically wetter regions. However, a statistically significant memory of the original MCS event still remains in surface anomalies out to 20 days. In terms of rainfall, we see a strong suppression of convection in the first 48 hours after the MCS in areas which initially received heavy rain. There is also some evidence of enhanced MCS activity around the edges of the original swath in the first 4 days. The persistence over several days of mesoscale rainfall patterns anti-correlated with the original MCS point to an important role for surface-atmosphere feedbacks. Synoptic forcing cannot explain the finer scale rainfall response, whilst post-MCS cold pool effects are too short-lived. On longer time scales (5-20 days) in climatologically drier areas, we also find a weak but statistically significant enhancement of rainfall around the original initiation zone.

These results have important implications for rainfall forecasting on scales of tens to several hundred kilometres. Pre-existing soil moisture heterogeneity provides strong predictability of where future convection will occur under favourable synoptic conditions. This provides skill out to 2-4 days, but strongly depends on regional rainfall frequencies. Because new MCSs create new soil moisture patterns, the combination of storms every few days and a strong negative land feedback at the mesoscale actively degrades longer term predictability within the rainy season, effectively limiting intra-seasonal to seasonal forecast skill for severe weather.

How to cite: Taylor, C., Klein, C., and Harris, B.: Multiday mesoscale soil moisture persistence and atmospheric predictability – an illustration from the Sahel, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5880, https://doi.org/10.5194/egusphere-egu24-5880, 2024.

Shallow cumulus cloud fields often organize and deepen within regions of mesoscale ascent associated with orographic flows. However, there is significant uncertainty in the relative roles of mesoscale and cloud-scale factors ultimately controlling the location of orographic deep convection initiation (DCI). These factors include spatial heterogeneity of the magnitude of mesoscale vertical mass flux associated with orographic convergence, near-cloud convective ingredients (e.g. CAPE, CIN, LFC, and shear), and entrainment effects. More fundamentally, it is not well understood how these factors influence the initial width and strength of low-level cloudy updrafts, which are increasingly cited as important governors of their ultimate depth potential. Thus, it is important to better understand these relationships for increased predictability of DCI.

Numerous DCI events observed along the Sierras de Córdoba range during the Cloud, Aerosol, and Complex Terrain Interactions (CACTI) project were modeled by the U.S. Department of Energy’s LES ARM Symbiotic Simulation and Observation (LASSO) team. In this study, we examine the connection between low-level cloudy updrafts, DCI, and the ascent associated with the mesoscale orographic circulation using LES with 100-m and 500-m grid spacing across multiple days. We hypothesize that the width and strength of low-level cloudy updrafts and the probability of DCI events along the ridge are proportional to the width, strength, and depth of the local orographic convergence. To test this, we examine correlations between the width and depth of developing cloudy updrafts and the: i) 3D structure of the evolving orographic ascent, and ii) convective meteorological ingredients (e.g., convective available potential energy, convective inhibition, level of free convection, moisture, etc.).

Preliminary results indicate that DCI events do not always occur in regions of the strongest or widest orographic ascent along the mountain range. Further, the strength and width of low-level cloudy updrafts that precede DCI are only weakly correlated with most orographic ascent metrics. Overall, the apparent relative roles of mesoscale ascent and convective sounding parameters governing DCI varied significantly across case days. Near-cloud relative humidity located near and just above the level of free convection steadily increased with time during each afternoon, likely owing to orographic vertical moisture flux and/or cloud detrainment. Thus, in addition to highly varied roles of the background conditions, the fate of individual growing cloudy updrafts may further depend on complex cloud-scale factors, such as entrainment and microphysical processes.

How to cite: Marquis, J., Varble, A., Feng, Z., Jo, E., and Gustafson, W.: Relationships between growing cloudy updrafts, deep convection initiation, and orographic flow, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6302, https://doi.org/10.5194/egusphere-egu24-6302, 2024.

  Recently, the increase in convective storms that develop rapidly within a short period of time and on a very small area causes severe damage to property and human life. Thus, it is important to understand the characteristics of convective activities and to provide the information about severity of the developing storms.  In order to address these issues, object-based analysis of convective systems is essential to provide severity information on convective precipitation systems including their life-cycle from initiation to dissipation. 
  In this study, we analyzed the developing stage of convective storms by using the statistics of storms detected by the Fuzzy Logic Algorithm for Storm Tracking (FAST). The Column Maximum (CMAX) was used to provide the information on detection and severity of storms. A convective storm was defined as a CMAX values above 35dBZ and small convective cells with an area less than 20km2 were filtered out. The identified storm was tracked on a fuzzy basis using storm speed and its morphological characteristics. Within the detected storm area, we analyzed the characteristics of the storm by averaging variables such as reflectivity (ZH), echo top height corresponding to ZH, rainfall rate at 1.5km altitude, VIL (Vertical Integrated Liquid) contents, etc.
  This study aims to provide quantitative information on severity of individual storms by using these radar variables and storm characteristics. We calculated and modified the threshold values of each predictor for determining the severity of the convective storms. Furthermore, we plan to analyze the intensity and frequency of severe precipitation storms in associated with the occurrence or absence of lightning event during their life cycle.

Key words : Weather Radar, convective storms, Radar parameter, storm severity

※ This research was supported by the "Development of radar based severe weather monitoring technology (KMA2021-03121)" of "Development of integrated application technology for Korea weather radar" project funded by the Weather 

How to cite: Kang, E., Kwon, S., and Lee, S.: Analysis of convective storm characteristics to classify the storm severity information using weather radar variables, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6900, https://doi.org/10.5194/egusphere-egu24-6900, 2024.

The forecast accuracy of numerical weather prediction models is strongly determined by the precision of the initial conditions, especially for storm and convective-scale weather prediction. Since radars allow to capture the internal structure and important microphysical and dynamical processes in convective systems, they are crucial instrument for improvement of weather forecasts on these scales. Dual-polarization radar, in contrast to a prevalent single-polarization radar, also provides information on the types and sizes of hydrometeor particles. As a result, polarimetric radar data (PRD) proves to be a valuable data source for data assimilation. However, direct assimilation of PRD is not used in current operational non-hydrostatic convection-permitting numerical models. This is associated with several difficulties, such as model error estimation, which require further study.

The current focus of our study is to directly assimilate PRD in an idealized setup. For that purpose, observation system simulation experiments (OSSEs) were performed that simulate the development of a long-lived supercell using the ICON model with two-moment microphysics scheme. In OSSE, the Kilometer-scale Ensemble Data Assimilation (KENDA) system, which comprises the Local Ensemble Transform Kalman Filter (LETKF) was used. The new polarimetric radar forward operator EMVORADO-POL developed at Deutscher Wetterdienst (DWD) was incorporated in the setup. The first steps towards the direct assimilation of differential reflectivity, in addition to non-polarimetric variables, have been implemented and will be presented. Proper thresholds and model equivalents of polarimetric data were examined. Results were compared to reference experiments assimilating non-polarimetric variables such as reflectivity and radial velocity.

How to cite: Bardachova, T., Ramezani Ziarani, M., and Janjic, T.: Towards the assimilation of dual-polarization radar data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8069, https://doi.org/10.5194/egusphere-egu24-8069, 2024.

Mesoscale convective systems (MCSs) frequently occur over Asia during the warm season, often producing intense precipitation with associated socioeconomic impacts. Here we reveal significant trends in MCS occurrence frequency and related precipitation in Asia during the warm season (March–September) in 2001–2020, using a tracking method that combines cloud and precipitation criteria with high-resolution satellite data from the Global Precipitation Measurement mission. To examine whether there are differences between MCSs of different scales, both meso-α scales (MαCSs) and meso-β scales (MβCSs), with horizontal scales of 200–2,000 km and 20–200 km, are tracked. The distribution pattern of frequency and related precipitation of both MαCSs and MβCSs are quite similar and manifest positive trends over East Asia (EA) and Northeast Asia, and negative trend over Southeast Asia (SEA). The MCS precipitation trend contributes significantly to total precipitation trend, with MαCSs contributing the most. Our analysis indicates the trend in lower-tropospheric water vapor flux convergence has a similar spatial pattern to the MCS frequency and related precipitation trend. Based on an atmospheric moisture flux decomposition analysis, the water vapor flux convergence trend can largely be explained by the change in horizontal wind convergence, while the specific humidity trend driven largely by temperature change plays a minor role. The trend in wind convergence in EA and SEA is possibly related to the evident trend in the lower-tropospheric anticyclone over the western North Pacific and SEA, which might be due to the relatively stronger warming in the Indian Ocean during the past two decades.

How to cite: Guo, Y., Fu, Q., Leung, L. R., Na, Y., and Lu, R.: Trends in Warm Season Mesoscale Convective Systems OverAsia in 2001–2020, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8739, https://doi.org/10.5194/egusphere-egu24-8739, 2024.

Convective clouds play a crucial role for understanding the Earth’s climate. Current advancements of remote sensing instruments allow us to obtain valuable information on the spatio-temporal dynamics of convective clouds on multiple scales. Nevertheless, a continuous coverage of high-resolved 4D observational data to investigate the 3D properties of rapidly developing convective clouds is generally not available.

In this study, we leverage 4D radar reflectivities (in dBZ) derived from the extrapolation of passive and active remote sensing sensors with machine learning to close this gap. Using data with a spatial resolution of 3 km and a temporal resolution of 15 minutes, we receive a continuous perspective on the evolution of the cloud vertical column along the different stages of the cloud life-cycle. For this purpose, we apply an object-based algorithm to detect the centroid of convective cores and their anvil at each time step. Based on these centroids, we extract the 3D cloud field and track the horizontal and vertical movement through space and time. Afterwards, we filter all tracks using the vertical extension and maximum reflectivity of the associated cloud field to exclude erroneous features.

Here, we present an evaluation of the algorithm and its ability to investigate the 4D spatio-temporal properties of convective clouds. We set out to compare convective systems of different sizes over both oceans and continents to analyze the impact of varying environmental conditions on the cloud vertical motions along the cloud life-cycle.

How to cite: Brüning, S. and Tost, H.: Investigating the life-cycle of convective clouds from 4D observational data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8810, https://doi.org/10.5194/egusphere-egu24-8810, 2024.

This study aims to improve the localization and assimilation of satellite observations in the visible and infrared spectral ranges to enhance predictions of clouds and convective processes. Understanding correlation structures between satellite observations and atmospheric state variables is crucial for successful data assimilation. We focus on examining vertical ensemble-based correlations from Himawari-8 channels (VIS0.64 or IR7.35) and tackle the challenge of vertical observation-space localization. Traditional distance-based localization methods are often suboptimal due to the multi-layered origin of observed radiation. We present empirical optimal localization (EOL) functions derived from a 1000-member ensemble convective-scale simulation to address this issue. Our research highlights the need for channel-specific and variable-specific localization strategies, emphasized by our analysis of two summer case studies that exhibit substantial situational variability in correlation structures, especially in the visible spectral range. Further, we explore various predictors for formulating dynamic, situation-specific vertical localization strategies, offering insights into their effectiveness and potential for advancing convective-scale satellite data assimilation.

How to cite: Necker, T., Honda, T., Griewank, P., Miyoshi, T., and Weissmann, M.: Situation-Dependent Localization for All-Sky Satellite Observations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8868, https://doi.org/10.5194/egusphere-egu24-8868, 2024.

Designing a weather station network is a demanding, multi-objective optimisation problem and usually constrained to local geographies. In this study, the authors deviate from typical approaches that focus the design of weather station networks on a small or country-wide area and present a method that is applicable on a global scale.

Prior art suggests that weather networks should exhibit high density, often at 1-3km or finer resolution, especially when deployed over complex topographies and urban landscapes. High station density is usually required to support research on urban micrometeorology, agricultural applications and to capture intricate meteorological mesoscale phenomena such as convective precipitation and sea breeze. High density is also required due to the persistence of temperature inversions at near-surface layers is significantly influenced by topography, leading to prolonged periods of temperature inversion.

In this novel approach, the authors suggest the design of a global weather network distributed over millions of hexagons covering the entire world. The number of weather stations per hexagon is determined by the topology (e.g. maximum elevation difference, aspects, water formations, etc.) and the land use (urban coverage, green areas, etc.) of the covered area.

The method is materialized via an open-source software tool (available on GitHub) which utilizes freely available elevation data (Copernicus DEM) and land use data (OpenStreetMap) and is capable of preparing the global weather station network in reasonable computation time (~24 hours on a 16-core CPU).

Finally, the authors present their findings, discuss the effect of various hexagon sizes and suggest that the design of a global weather station network is viable and computationally feasible.

How to cite: Keppas, S., Balis, H., Dravilas, I., and Pagonis, J.: Designing a Global Weather Station Network, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9241, https://doi.org/10.5194/egusphere-egu24-9241, 2024.

Forecasting high impact weather events is a major challenge for numerical weather prediction. Initial condition uncertainty plays an important role but so do uncertainties arising from the representation of subgrid-scale processes, e.g. cloud microphysics. Here, we investigate the impact of cloud microphysical parameter uncertainties on the forecast of a selected severe convective storm over South-Eastern Germany in 2019, which is generally referred to as the Munich hailstorm (Wilhelm et al., 2020).
The storm is simulated using the ICON model (2-moment cloud microphysics, 1 km grid-spacing) with perturbed microphysical parameters related to graupel and hail formation. Combinations of parameter perturbations are chosen according to a Latin hyper cube design and one-at-a-time parameter perturbations for the smallest and largest parameter values. Important impacts on surface (hail) precipitation are found for parameters pertaining to (i) CCN and INP activation, (ii) diffusional growth of ice, and (iii) the mass-diameter and mass-fall velocity relations for graupel. The behavior of graupel particles are thereby controlled by their density.
The one-at-a-time parameter perturbation simulations are used to track microphysical process rates. By closing the hydrometeor mass budgets we explore changes in precipitation formation pathways (based on the approach by Barrett and Hoose, 2023) arising from perturbations of the most impactful parameters. Preliminary results show a strong influence of graupel density on the hail particle size distribution as well as total precipitation, but less so on surface hail amount.
The analysis allows us to draw conclusions about the most impactful cloud microphysical parameters for hail forecast uncertainty as well as the underlying mechanisms.

How to cite: Kuntze, P., Miltenberger, A., Hoose, C., Kunz, M., and Frey, L.: Impact of aerosol and microphysical uncertainty on the evolution of a severe hailstorm, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-9266, https://doi.org/10.5194/egusphere-egu24-9266, 2024.

Convective storms over South America and Australia are among the most intense worldwide (e.g., Zipser 2006). However, they are less researched compared to US and Europe. This study analyses the thunderstorm climatology over South America and Australia based on over 20 years of overshooting cloud top (OT) satellite detections (Khlopenkov et al. 2021). These OTs serve as robust, horizontally homogeneous indicators of strong updrafts and hence intense thunderstorms. Furthermore, we focus on the frequency of severe storms and hail by using ERA5 Reanalysis data to exclude OTs in unfavorable environments (e.g., Punge et al. 2023).
The resulting climatologies of intense thunderstorms and hail are largely consistent with existing literature, showing strong thunderstorm activity in tropical regions but more severe (e.g., hail-producing) storms in south-central South America and southeast Australia. Some notable details will also be discussed, such as the discrepancy with observational hotspots near the coast in South America and a surprisingly strong signal over northwest Australia. Furthermore, regarding a climate change signal, preliminary analysis indicates no significant trend for South America. However, the multi-year variations are strongly linked to the El Ninjo-Southern Oscillation (ENSO).

How to cite: Fischer, J., Kunz, M., and Bedka, K.: Thunderstorm and Hail Frequencies in South America and Australia Based on Overshooting Tops, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11456, https://doi.org/10.5194/egusphere-egu24-11456, 2024.

Convective storms are an important weather hazard in Europe as shown by the high number of severe wind gusts, large hail, tornadoes, and flash floods recorded each year in the European Severe Weather Database. A recent innovation to the ESWD was the introduction of the new tornado International Fujita scale for rating tornado and wind intensity from damage. In 2023, no fewer than 62182 new reports were entered, and reinsurer Munich Re estimated severe thunderstorms to account for the majority of weather-related losses in Europe in 2023 with a total damage of € 10 billion.

At the core of ESSL’s mission is conducting and facilitating research on severe weather at a European level. Over the years, the organisation has grown with support from its members which include most of Europe’s weather services and commercial sector partners. In addition to research ESSL is active in the area of forecaster training and the evaluation of novel forecasting and nowcasting applications at the ESSL Testbed.

The recorded multi-year trends of severe weather apparent in the ESWD are often dominated by non-meteorological factors, but for large hail indications are strong that its frequency is changing, illustrated by the new hailstone size record of 19 cm diameter in northern Italy in July 2023. ESSL’s recent models of large hail climatology across Europe and the world support these trends. A key challenge for the research community is to develop methods to estimate trends from ever higher-resolution reanalyses and climate models. This is not straightforward as even the highest resolution models do not resolve tornadoes or microbursts, let alone hailstones, and already show biases at coarse scales.

The mentioned work modelling severe weather has given new insights into which environmental characteristics are important to severe weather occurrence. For hail, we additionally studied the conditions under which individual hailstorms in 2021, 2022, and 2023 that were particularly severe. We show the importance of the vertical distribution of buoyancy and wind in a storm-centred reference framework, defined using radar-observed storm motion.

High vertical wind shear above the boundary layer and high CAPE above the -10 °C isotherm for hail, and a combination of vertical vorticity and strong streamwise vorticity for tornadoes. ESSL is collaborating with ECMWF to develop forecast tools based on these concepts. That said, many questions remain, for example regarding the pre-convective environment and mountain ranges, and with the developing storms. For instance, an important concentration of severe weather is evident surrounding the Alps. To address related questions, ESSL has taken the initiative for a multi-year multi-national field campaign in central Europe called TIM (Thunderstorm Intensification from Mountains to plains), in which it will collaborate with a large number of research institutes.

How to cite: Groenemeijer, P., Battaglioli, F., Púčik, T., Holzer, A., and Taszarek, M.: Severe Storms Research at ESSL, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12158, https://doi.org/10.5194/egusphere-egu24-12158, 2024.

Despite the abundance of satellite observations, their assimilation in all-sky scenarios remains difficult, which hinders the use of high-resolution information in forecast models. In this work, we focus on the direct assimilation of satellite radiances (visible 0.6 μm, infrared IR 6.2 and 7.3 μm of the Seviri instrument) under all-sky conditions into the convection-permitting Numerical Weather Prediction (NWP) model AROME, which is in operation at Geosphere Austria, the Austrian weather service. Our research aims to exploit the potential of assimilating visible and IR channels under all-sky conditions making use of convection-permitting weather models that can explicitly resolve deep convection. In particular, we are looking at the use of the Radiative Transfer for TOVS (RTTOV), as an observational operator, to generate synthetic images for each channel. We endeavor to optimize the operator settings for running simulations within the convective-scale AROME model. Our first experiment focuses on testing IR synthetic images generated under all-sky conditions during a summer period (August) over Austria. 

How to cite: Chkeir, S., Weissmann, M., Griewank, P., Meier, F., and Neduncheran, A.: Preparing AROME assimilation experiments for cloud-affected satellite observations, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12221, https://doi.org/10.5194/egusphere-egu24-12221, 2024.

While ensemble methods to estimate the impact of assimilated observations on forecast error have been widely used (known as EFSO), similar methods to estimate the benefit of potential observations not assimilated have received less attention. For this presentation we use a toymodel to illustrate these methods and highlight their strengths and weaknesses. We show that these methods work well over a wide range of lead times and for different types of observations, but only when the localization used in ensemble data assimilation to mitigate sampling errors is accounted for. While previous studies struggled to achieve quantitative results because they treated the localization inconsistently, we found three methods to overcome this limitation. 

How to cite: Griewank, P., Necker, T., and Weissmann, M.: Testing ensemble-based estimates of potential observation impact, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-12544, https://doi.org/10.5194/egusphere-egu24-12544, 2024.