Urban canopies consist of buildings and trees that are aligned along a street in the horizontal direction. These canopies in cities and forests modulate the local climate considerably in a complex way. Canopies constitute very fine subgrid features that actually have a significant impact on other components of earth system models but their feedbacks on larger scales are by now represented in rather heuristic ways. The problem in simulating their impact is twofold: First, their local modeling is delicate and, secondly, the numerical modeling of the scale interaction between fine and large scales is complicated since the fine scale structure is global. We will mostly focus on the second aspect.

Multiscale finite element methods (MsFEM) in their classical form have been applied to various porous media problems but the situation in climate, and hence flow-dominated regimes is different from porous media applications. In order to study the effect of various parameters like the concentration of pollutants, or the dynamics of the background velocity and of the temperature in the atmospheric boundary layer, a semi-Lagrangian reconstruction based multiscale finite element framework (SLMsR) developed by [1, 2] for passive tracer transport modeled by an advection-diffusion equation with high-contrast oscillatory diffusion is applied.

These methods are composed of two parts: a local-in-time semi-Lagrangian offline phase that pre-computes basis functions and an online phase that uses these basis functions to compute the solution on a coarse Eulerian simulation mesh. The overhead of pre-computing the basis functions in each coarse block can further be reduced by parallelization. The online phase is approximately as fast as a low resolution standard FEM but using the modified basis that carries subgrid information still allows to reveal the fine scale features of a highly resolved solution and is therefore accurate. This approach is studied in order to reveal the feedback of processes in the canopy layer on different scales present in climate simulation models and in particular on the atmospheric boundary layer.

We will show the results of massively parallel simulations for passive tracer transport in an urban region using the new multiscale approach and compare them to classical approaches.

References :

[1] Simon, Konrad, and Jörn Behrens. "Semi-Lagrangian Subgrid Reconstruction for Advection-Dominant Multiscale Problems.", Springer Journal of Scientific Computing (JOMP) (provisionally accepted), 2019

[2] Simon, Konrad, and Jörn Behrens. "Multiscale Finite Elements for Transient Advection-Diffusion Equations through Advection-Induced Coordinates.", Multiscale Modeling & Simulation 18.2 (2020): 543-571.

How to cite: Patel, H., Simon, K., and Behrens, J.: Massively Parallel Multiscale Simulations of the Feedback of Urban Canopies, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2507, https://doi.org/10.5194/egusphere-egu21-2507, 2021.

Running simulations on high-performance computers faces new challenges due to e.g. the stagnating or even decreasing per-core speed. This poses new restrictions and therefore challenges on solving PDEs within a particular time frame in the strong scaling case. Here, disruptive mathematical reformulations, which e.g. exploit additional degrees of parallelism also along the time dimension, gained increasing interest over the last two decades.

This talk will cover various examples of our current research on (parallel-in-)time integration methods in the context of weather and climate simulations such as rational approximation of exponential integrators, multi-level time integration of spectral deferred correction (PFASST) as well as other methods.

These methods are realized and studied with numerics similar to the ones used by the European Centre for Medium-Range Weather Forecasts (ECMWF). Our results motivate further investigation for operational weather/climate systems in order to cope with the hardware imposed restrictions of future super computer architectures.

I gratefully acknowledge contributions and more from Jed Brown, Francois Hamon, Terry S. Haut, Richard Loft, Michael L. Minion, Pedro S. Peixoto, Nathanaël Schaeffer, Raphael Schilling

How to cite: Schreiber, M.: Next-Generation Time Integration targeting Weather and Climate Simulations, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2734, https://doi.org/10.5194/egusphere-egu21-2734, 2021.

Fog and low stratus pose a major challenge for numerical weather prediction (NWP) models. Despite high resolution in the horizontal (~1 km) and vertical (~20 m), operational NWP models often fail to accurately predict fog and low stratus. This is a major issue at airports which require visibility predictions, or for energy agencies estimating day-ahead input into the electrical grid from photovoltaic power.

Most studies dedicated to fog and low stratus forecasts have focused on the physical parameterisations or grid resolutions. We illustrate how horizontal advection at the cloud top of fog and low stratus in a grid with sloping vertical coordinates leads to spurious numerical diffusion and subsequent erroneous dissipation of the clouds. This cannot be prevented by employing a higher-order advection scheme. After all, the formulation of the terrain-following vertical coordinate plays a crucial role in regions which do not exhibit perfectly flat orography. We suggest a new vertical coordinate formulation which allows for a faster decay of the orographic signal with altitude and present its positive impact on fog and low stratus forecasts. Our experiments indicate that smoothing of the vertical coordinates at low altitudes is a crucial measure to prevent premature dissipation of fog and low stratus in high-resolution NWP models.

How to cite: Westerhuis, S. and Fuhrer, O.: A new vertical coordinate formulation to improve forecasts of fog and low stratus in high-resolution numerical weather prediction models, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4429, https://doi.org/10.5194/egusphere-egu21-4429, 2021.

We introduce a new representation of the rotating shallow water equations based on a stochastic transport principle. The derivation relies on a decomposition of the fluid flow into a large-scale component and a noise term that models the unresolved small-scale flow. The total energy of such a random model is demonstrated to be preserved along time for any realization. Thus, we propose to combine a structure-preserving discretization of the underlying deterministic model with the discrete stochastic terms. This way, our method can directly be used in existing dynamical cores of global numerical weather prediction and climate models. For an inviscid test case on the f-plane we use a homogenous noise and illustrate that the spatial part of the stochastic scheme preserves the total energy of the system. Finally, using an inhomogenous noise, we show  that the proposed random model better captures the structure of a large-scale flow than a comparable deterministic model for a barotropically unstable jet on the sphere.

How to cite: Brecht, R., Li, L., Bauer, W., and Mémin, E.: Rotating shallow water flow under location uncertainty with a structure-preserving discretization, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7539, https://doi.org/10.5194/egusphere-egu21-7539, 2021.

The Cubed Sphere is a grid commonly used in numerical simulation in climatology. In this talk we present recent progress
on the algebraic and geometrical properties of this highly symmetrical grid.
First, an analysis of the symmetry group of the Cubed Sphere will be presented: this group 
is identified as the group of the Cube, [1]. Furthermore, we show how to construct a discrete Spherical Harmonics (SH) basis associated to 
the Cubed Sphere. This basis displays a truncation scheme relating the zonal and longitudinal 
mode numbers reminiscent of the rhomboidal truncation on the Lon-Lat grid.
The new analysis allows to derive new quadrature rules of  interest for applications in any kind of spherical modelling. In addition,
we will comment on applications in mathematical climatology and meteorology, [2].

[1] J.-B. Bellet, Symmetry group of the equiangular Cubed Sphere, preprint, IECL, Univ. Lorraine, 2020, submitted

[2] J.-B. Bellet, M. Brachet and J.-P. Croisille, Spherical Harmonics on The Cubed Sphere, IECL, Univ. Lorraine, 2021, Preprint.

How to cite: Croisille, J.-P., Bellet, J.-B., and Brachet, M.: Interpolating data on the Cubed Sphere with Spherical Harmonics, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7990, https://doi.org/10.5194/egusphere-egu21-7990, 2021.

Semi-Lagrangian advection schemes are accurate and efficient and retain accuracy and stability even for large Courant numbers but are not conservative. Flux-form semi-Lagrangian is conservative and in principle can be used to achieve large Courant numbers. However this is complicated and would be prohibitively expensive on grids that are not logically rectangular. 

Strong winds or updrafts can lead to localised violations of Courant number restrictions which can cause a model with explicit Eulerian advection to crash. Schemes are needed that remain stable in the presence of large Courant numbers. However accuracy in the presence of localised large Courant numbers may not be so crucial.

Implicit time stepping for advection is not popular in atmospheric science because of the cost of the global matrix solution and the phase errors for large Courant numbers. However implicit advection is simple to implement (once appropriate matrix solvers are available) and is conservative on any grid structure and can exploit improvements in solver efficiency and parallelisation. This talk will describe an implicit version of the MPDATA advection scheme and show results of linear advection test cases. To optimise accuracy and efficiency, implicit time stepping is only used locally where needed. This makes the matrix inversion problem local rather than global. With implicit time stepping MPDATA retains positivity, smooth solutions and accuracy in space and time.

How to cite: Weller, H., Woodfield, J., and Kuehnlein, C.: Long Time Steps for Advection: MPDATA with implicit time stepping, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8815, https://doi.org/10.5194/egusphere-egu21-8815, 2021.

In this work we consider the surface quasi-geostrophic (SQG) system under location uncertainty (LU) and propose a Milstein-type scheme for these equations. The LU framework, first introduced in [1], is based on the decomposition of the Lagrangian velocity into two components: a large-scale smooth component and a small-scale stochastic one. This decomposition leads to a stochastic transport operator, and one can, in turn, derive the stochastic LU version of every classical fluid-dynamics system. 

    SQG is a simple 2D oceanic model with one partial differential equation, which models the stochastic transport of the buoyancy, and an operator which relies the velocity and the buoyancy.

    For this kinds of equations, the Euler-Maruyama scheme converges with weak order 1 and strong order 0.5. Our aim is to develop higher order schemes in time: the first step is to consider Milstein scheme, which improves the strong convergence to the order 1. To do this, it is necessary to simulate or estimate the Lévy area [2].

    We show with some numerical results how the Milstein scheme is able to capture some of the smaller structures of the dynamic even at a poor resolution. 

References

[1] E. Mémin. Fluid flow dynamics under location uncertainty. Geophysical & Astrophysical Fluid Dynamics, 108.2 (2014): 119-146. 

[2] J. Foster, T. Lyons and H. Oberhauser. An optimal polynomial approximation of Brownian motion. SIAM Journal on Numerical Analysis 58.3 (2020): 1393-1421.

How to cite: Fiorini, C., Li, L., and Mémin, É.: Higher order schemes in time for the surface quasi-geostrophic system under location uncertainty, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10935, https://doi.org/10.5194/egusphere-egu21-10935, 2021.

Large-scale instability in incompressible fluid driven by the so called Anisotropic Kinetic Alpha (AKA) effect satisfying the incompressible Navier-Stokes equation with Coriolis force is considered. The external force is periodic; this allows applying an unusual for turbulence calculations mathematical method developed by Frisch et al [1]. The method provides the orders for nonlinear equations and obtaining large scale equations from the corresponding secular relations that appear at different orders of expansions. This method allows obtaining not only corrections to the basic solutions of the linear problem but also provides the large-scale solution of the nonlinear equations with the amplitude exceeding that of the basic solution. The fluid velocity is obtained by numerical integration of the large-scale equations. The solution without the Coriolis force leads to constant velocities at the steady-state, which agrees with the full solution of the Navier-Stokes equation reported previously. The time-invariant solution contains three families of solutions, however, only one of these families contains stable solutions. The final values of the steady-state fluid velocity are determined by the initial conditions. After account of the Coriolis force the solutions become periodic in time and the family of solutions collapses to a unique solution. On the other hand, even with the Coriolis force the fluid motion remains two-dimensional in space and depends on a single spatial variable. The latter fact limits the scope of the AKA method to applications with pronounced 2D nature. In application to 3D models the method must be used with caution.

[1] U. Frisch, Z.S. She and P. L. Sulem, “Large-Scale Flow Driven by the Anisotropic Kinetic Alpha Effect,” Physica D, Vol. 28, No. 3, 1987, pp. 382-392.

How to cite: Rutkevich, P., Golitsyn, G., and Tur, A.: Coriolis force influence on the AKA effect, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13421, https://doi.org/10.5194/egusphere-egu21-13421, 2021.

Radial basis functions enable the use of unstructured quasi-uniform nodes on the sphere. Iteratively generated nodes such as the minimum energy nodes may not converge due to exponentially increasing local minima as the number of nodes grows. By contrast, deterministic nodes, such as those made with a spherical helix, are fast to generate and have no arbitrariness. It is noteworthy that the spherical helix nodes are more uniform on the sphere than the minimum energy nodes. Semi-Lagrangian and Eulerian models are constructed using radial basis functions and validated in a standard advection test of a cosine bell by the solid body rotation. With Gaussian radial basis functions, the semi-Lagrangian model found produces significantly smaller error than the Eulerian counterpart in addition to approximately three times longer time step for the same error. Moreover, the ripple-like noise away from the cosine bell found in the Eulerian model is significantly reduced in the semi-Lagrangian model. It is straightforward to parallelize the matrix–vector multiplication in the time integration. In addition, an iterative solver can be applied to calculate the inverse of the interpolation matrix, which can be made sparse.

How to cite: Enomoto, T. and Ogasawara, K.: Semi-Lagrangian advection models for quasi-uniform nodes on the sphere, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13687, https://doi.org/10.5194/egusphere-egu21-13687, 2021.

In this work, a stochastic representation [Bauer2020a, Bauer2020b] based on a physical transport principle is proposed to account for mesoscale eddy effects on the the large-scale oceanic circulation. This stochastic framework [Mémin2014] arises from a decomposition of the Lagrangian velocity into a time-smooth component and a highly oscillating noise term. One important characteristic of this random model is that it conserves the energy of any transported tracer. Such an energy-preserving representation has been successfully implemented in a well established multi-layered quasi-geostrophic dynamical core (http://www.q-gcm.org). The empirical spatial correlation of the small-scale noise is estimated from the eddy-resolving simulation data. In particular, a sub-grid correction drift has been introduced in the noise due to the bias ensuing from the coarse-grained procedure. This non intuitive term seems quite important in reproducing on a coarse mesh the meandering jet of the wind-driven double-gyre circulation. In addition, a new projection method has been proposed to constrain the noise living along the iso-surfaces of the vertical stratification. The resulting noise enables us to improve the intrinsic low-frequency variability of the large-scale current. From some statistical studies and energy transfers analysis, this improvement is well demonstrated.

• [Bauer2020a] W. Bauer, P. Chandramouli, B. Chapron, L. Li, and E. Mémin. Deciphering the role of small-scale inhomogeneity on geophysical flow structuration: a stochastic approach. Journal of Physical Oceanography, 50(4):983-1003, 2020a.       
• [Bauer2020b] W. Bauer, P. Chandramouli, L. Li, and E. Mémin. Stochastic representation of mesoscale eddy effects in coarse-resolution barotropic models. Ocean Modelling, 151:101646 (2020b).    
• [Mémin2014] E. Mémin. Fluid flow dynamics under location uncertainty. Geophysical & Astrophysical Fluid Dynamics, 108(2):119-146, 2014.     

How to cite: Li, L., Deremble, B., Lahaye, N., and Mémin, E.: Stochastic modeling of the oceanic mesoscale eddies, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15010, https://doi.org/10.5194/egusphere-egu21-15010, 2021.

This talk focusses on progress in ensemble forecasting methodology (Part I) and ensemble verification methodology (Part II).

Operational ECMWF ensemble forecasts are global predictions from days to months ahead. At all forecast ranges, model uncertainties are represented stochastically with the Stochastically Perturbed Parametrization Tendency scheme (SPPT). Recently, considerable progress has been made in developing the Stochastically Perturbed Parametrization scheme (SPP). The SPP scheme offers improved physical consistency by naturally preserving the local conservation properties for energy and moisture of the unperturbed version of the corresponding parametrization. In contrast, the SPPT scheme lacks such local conservation properties, mainly because the scheme does not perturb fluxes at the surface and at the top of the atmosphere consistently with the tendency perturbations in the column.

NWP research and development relies on scoring rules to judge whether or not a change to the forecast systems results in better ensemble forecasts. A new tool will be presented that can improve the understanding of score differences between sets of forecasts for a widely used proper score, the Continuous Ranked Probability Score (CRPS). An analytical expression has been derived for the CRPS when a homogeneous Gaussian (hoG) forecast-observation distribution is considered. This leads to an approximation of the CRPS when actual verification data are considered, which deviate from a homogeneous Gaussian distribution. The hoG approximation of the CRPS permits a useful decomposition of score differences. The methodology will be illustrated with verification data for medium-range weather forecasts.

How to cite: Leutbecher, M., Ben Bouallegue, Z., Haiden, T., Lang, S., and Lock, S.-J.: Progress in ensemble forecasting and verification methodologies at ECMWF, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-108, https://doi.org/10.5194/egusphere-egu21-108, 2021.

The GLM instruments aboard the GOES-16 and 17 satellites provides nearly uniform spatiotemporal coverage of total lightning over the Americas and adjacent vast oceanic regions of the western hemisphere. This work summarizes recent efforts from our group at CIMMS/NSSL geared towards the evaluation of the potential added value of assimilating GLM-observed total lightning data on short-term, convection-allowing scale (dx = 2-3 km) forecasts for higher impact weather events. Results using data assimilation (DA) approaches ranging from single deterministic three-dimensional variational (3DVAR) methods applied in real time to experimental ensemble-based VAR hybrid methods (3DEnVAR) will be highlighted. 
The lightning data assimilation (DA) scheme in these frameworks follow the same core philosophy wherein background water vapor mass mixing ratio is adjusted (increased) locally at or around observed lightning locations, either throughout the entire atmospheric column or within a fixed, confined layer above the lifted condensation level. Toward a more systematic assimilation of real GLM data, emphasis will be directed toward: (i) sensitivity tests with deterministic 3DVAR experiments aimed at evaluating the impact of the horizontal decorrelation length scale, DA cycling frequency as well the length of the accumulation window for the lightning data, (ii) aggregate statistics from real time CONUS-scale experiments over the Spring 2020 and (iii) preliminary results employing ensemble of 3DEnVARs with hybrid (static + flow dependent) background error covariances. 
Aggregate statistical results from all deterministic 3DVAR exercises in (i) and (ii) revealed that the assimilation of either radar (radial wind and reflectivity factor) or total lightning (GLM) resulted in overall notably more skillful, shorter term (0-3 h) forecast of composite reflectivity fields, accumulated rainfall, as well as individual storm tracks – with optimal skill obtained when both radar and lightning data were assimilated. In (iii) forecast impacts related to the following will be summarized: (1) the respective weights assigned to the flow-dependent component and static components of the background error covariances, (2) the inclusion of three time-level sampling for each member during each cycle and (3) the usage of Gaussian noise coupled with a fixed 3 to 12 h spin-up period prior to the beginning of the cycled 3DVAR.

How to cite: Fierro, A., Hu, J., Wang, Y., Gao, J., and Mansell, E.: Evaluation of the impact of assimilating spaceborne (GLM) total lightning data and radar data on short-term forecasts of convective events in the 3DVAR framework, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-133, https://doi.org/10.5194/egusphere-egu21-133, 2021.

As the computational capacity has been largely improved in the last decades, the grid configuration of numerical weather prediction models has stepped into microscale resolutions. Even if mesoscale models are not originally designed to reproduce fine scale phenomena, a large effort is being made by the research community to improve and adapt these systems. However, reasonable doubts exist regarding the ability of the models to forecast this type of events, due to the unfit parametrizations and the appearance of instabilities and lack of sensitivity in the variables. Here, the Weather Research and Forecasting (WRF) model effective resolution is evaluated for several situations and grid resolutions. This is achieved by assessing the curve of dissipation for the wind kinetic energy. Results show that the simulated energy spectrum responds to different synoptic conditions. Nevertheless, when the model is forced into microscale grid resolutions the dissipation curves present an unrealistic atmospheric energy. This may be a partial explanation to the aforementioned issues and imposes a large uncertainty to forecasting at these resolutions.

How to cite: Bolgiani, P., Díaz-Fernández, J., Quitián-Hernández, L., Sastre, M., Santos-Muñoz, D., Farrán, J. I., González-Alemán, J. J., Valero, F., and Martín, M. L.: On the effective resolution of WRF simulations at microscale grid resolution., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-229, https://doi.org/10.5194/egusphere-egu21-229, 2021.

This study investigates sensitivity of  cloud and precipitation parameterized microphysics  to stochastic representation of parameter uncertainty as formulated by the stochastically perturbed parameterization (SPP) scheme.  SPP is applied to multiple microphysical parameters within a lagrangian column model, used in several prior published studies to characterize  parameter uncertainty by means of multivariate nonlinear inversions using remote sensing observations. The 1D column microphysics model is forced with prescribed time-varying profiles of temperature, humidity and vertical velocity.  This modeling framework allows for investigation of the effect of changes in model physics parameters on the model output in isolation from any feedback to the cloud-scale dynamics.

The test case selected in this study of an idealized representation of mid-latitude squall-line convection is the same as in the prior studies. This enabled using the estimates of multi-parameter distributions from the inversions in the prior studies as the basis for setting the second-moment statistics in the SPP scheme implementation. Additionally impacts of the non-stochastic and stochastic multi-parameter representation of parameterization uncertainty on the microphysics model solution could be directly compared.

The sensitivity experiments with the SPP scheme involve ensemble simulations where each member is evolved with a different stochastic sequence of parameter perturbations, as is done in the standard practice with this scheme.  The experiments explore impacts of using different decorrelation times and different estimates of second moment statistics for the parameter perturbations.  These include uncorrelated perturbations between the parameters for several values of variance for each parameter and correlated perturbations based on multi-parameter empirical statistical distributions from the prior studies.  The selection of physical parameters for the perturbations is based on the significance of their impacts derived from the prior studies . 

The results are evaluated in terms of changes to the ensemble mean and variance of time evolving profiles of hydrometeor mass quantities, the microphysics processes within the model as well as in terms of the simulated column integral microphysics-sensitive satellite-based  observables. The latter include PR (Precipitation Rate) , LWP (Liquid Water Path), IWP (Ice water path), TOA-LW and TOA-SW (-Long and -Short Wave, respectively).  In each experiment six parameters were perturbed.

The analyses performed so far indicate a high sensitivity of the microphysics model to the SPP scheme. The ensemble simulations with the standard uncorrelated parameter perturbations exhibit a significant bias relative to the control simulation which uses the unperturbed parameters.  For the selected test case the skewness toward small parameter values in the SPP sampling based on the underlying log-normal distributions leads to less precipitating ice and more precipitating liquid and accumulated precipitation. The response is due to nonlinear relationships between the parameters and modeled microphysics output. The changes in microphysics output result in large mean changes in PR, LWP, IWP, TOA- LW and SW, suggesting a potential for using these and other microphysics sensitive satellite observations to evaluate and if needed correct properties of the underlying sampling distribution in the stochastic scheme.  Further analyses will be presented at the conference.

How to cite: Vukicevic, T., Stankovic, A., and Posselt, D.: Sensitivity of modeled microphysics to stochastically perturbed parameters, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2587, https://doi.org/10.5194/egusphere-egu21-2587, 2021.

Previous work by this team has demonstrated that assimilation of IR radiances in partially cloudy regions is beneficial to numerical weather predictions (NWPs), improving the representation of tropical cyclones (TCs) in global analyses and forecasts. The specific technique used by this team is based on the “cloud-clearing CC” methodology. Cloud-cleared hyperspectral IR radiances (CCRs), if thinned more aggressively than clear-sky radiances, have shown a strong impact on the analyzed representation and structure of TCs. However, the use of CCRs in an operational context is limited by 1) latency; and 2) external dependencies present in the original cloud-clearing algorithm. In this study, the Atmospheric InfraRed Sounder (AIRS) CC algorithm was (a) ported to NASA high end computing resources (HEC), (b) deprived of external dependencies, and (c) parallelized improving the processing by a factor of 70. The revised AIRS CC algorithm is now customizable, allowing user’s choice of channel selection, user’s model's fields as first guess, and could perform in real time. This study examines the benefits achieved when assimilating CCRs using the NASA’s Goddard Earth Observing System (GEOS) hybrid 4DEnVar system. The focus is on the 2017 Atlantic hurricane season with three infamous hurricanes (Harvey, Irma, and Maria) investigated in depth.  The impact of assimilating customized CCRs on the analyzed representation of tropical cyclone horizontal and vertical structure and on forecast skill is discussed.

How to cite: Boukachaba, N., Reale, O., L. McGrath-Spangler, E., Ganeshan, M., McCarty, W., and Gelaro, R.: Progress toward Cloud-Cleared Infrared radiance assimilation in a global modeling framework: Application to the 2017 Atlantic Tropical Cyclone Season., EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2870, https://doi.org/10.5194/egusphere-egu21-2870, 2021.

Shallow cumulus clouds play an important role in the weather in the Atlantic Tropical Convergence Zone.  Their interaction with the atmospheric environment and oceanic mixing processes has a significant impact on the convective organization and tropical dynamics.  It is still a scientific challenge for numerical weather prediction models to accurately simulate them due to deficiencies in the model’s representation of physical processes. 

In this study, we investigate how the physics parameterization schemes in NOAA’s most recent operational global forecast system (GFSv16) perform in the simulation of shallow cumulus clouds in the western Atlantic in terms of their interaction with the large-scale atmospheric dynamics.  Previous studies have indicated that the impact of physics parameterization schemes on model’s tendencies during the first few hours can provide critical information on their suitability for short- and medium-range forecasts.  Therefore, we first evaluate the GFSv16 forecasts against the observations obtained from the European field campaign called the ATOMIC/EUREC4A that occurred between 12 January and 23 February 2020.  We then diagnose the sensitivity of the GFSv16 physics tendencies to changes to the physics parameterization schemes over the first 6 hours of the forecast, which is the timescale before dynamical feedback becomes significant. Using the information from the observational evaluation and physics tendency diagnosis, we further explore possible improvement in the physical process representation that can positively affect the physics tendencies and lead to overall forecast improvement beyond 6 hours.

How to cite: Bao, J.-W., Michelson, S., and Grell, E.: Evaluation of the physics suite in NOAA’s GFSv16 using field-campaign observations and diagnosis of physics tendencies, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3387, https://doi.org/10.5194/egusphere-egu21-3387, 2021.

This study evaluates the short-to-medium range precipitation forecasts from Global Forecast System for 14 major transboundary river basins in Africa against GPM IMERG “Early”, IMERG “Final”, and CHIRPSv2 products. Daily precipitation forecasts with lead times of 1-day, 5-day, 10-day, and 15-day and accumulated precipitation forecasts with periods of 1-day, 5-day, 10-day, and 15-day are investigated. The 14 selected basins are (1) Senegal; (2) Volta; (3) Niger; (4) Chad; (5) Nile; (6) Awash; (7) Congo; (8) Omo Gibe; (9) Tana; (10) Pangani; (11) Zambezi; (12) Okavango; (13) Limpopo and (14) Orange. For each basin, several sub-basins are defined by the major dams in the basin. Our preliminary results in the Nile river basin show that in terms of temporal variability, there was a good agreement between the forecasted and observed accumulated precipitation on a 15-day basis. When compared to IMERG “Final”, the correlation coefficients of accumulated GFS forecasts scored as high as 0.75. Thus, GFS products provide relatively reliable accumulated precipitation forecasts. However, the precipitation forecasts were mostly biased: they tend to overpredict rainfall for the eastern part of the Nile river, underestimate rainfall for the northern part of the Nile river and produce almost unbiased estimates for the southern part of the river. Additionally, GFS forecasts have a general tendency to underpredict the area of precipitation across the Nile basin. Although the performance of GFS varies at different locations, the GFS precipitation forecasts can be a good reference to dam operators in Africa. 

How to cite: Yue, H. and Gebremichael, M.: Evaluation of Precipitation Forecast from Global Forecast System Over Transboundary Rivers in Africa, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3652, https://doi.org/10.5194/egusphere-egu21-3652, 2021.

The Japan Meteorological Agency (JMA) is conducting the third Japanese global atmospheric reanalysis named Japanese Reanalysis for Three Quarters of a Century (JRA-3Q) using the JMA operational data assimilation system that has been upgraded and improved since the Japanese 55-year Reanalysis (JRA-55) was conducted. Main points of improvement in the specifications of the data assimilation system are as follows (specifications of the JRA-55 data assimilation system are shown in parentheses for comparison): Vertical levels are increased up to 100 (60) layers; The top level of the system is 0.01 (0.1) hPa; The inner model resolution for 4D-Var is also increased up to TL319 (T106); Various parameterization schemes have been improved and several new schemes have been implemented. In addition, we use observations newly rescued and digitized by the ERA-CLIM and other projects as well as newly reprocessed and improved satellite observations. As for GNSS radio occultation, bending angle is assimilated up to 60 km (refractivity up to 30 km).

The early results show that both overestimation of precipitation in the tropics and dry bias in the middle troposphere are diminished compared with those in JRA-55, and the representation of diabatic heating rate is also improved. In addition, biases of surface heat fluxes and radiation fluxes at the top of the atmosphere are also reduced.

How to cite: Harada, Y., Kobayashi, S., Kosaka, Y., Chiba, J., and Tokuhiro, T.: Early results of the evaluation of the JRA-3Q reanalysis, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3762, https://doi.org/10.5194/egusphere-egu21-3762, 2021.

The use of numerical modeling for atmospheric research is complicated by the problem of verification by a limited set of measurement data. Comparison with radar measurements is widely used for assessing the quality of the simulation. The probabilistic nature of the development of convective phenomena determines the complexity of the verification process: the reproduction of the pattern of the convective event is prior to the quantitative agreement of the values at a particular point at a particular moment.

We propose a method for verifying the simulation results based on comparing areas with the same reflectivity. The method is applied for verification of WRF-modeling of convective events in the Aragats highland massif in Armenia. It is shown that numerical simulation demonstrates approximately the same form of distribution of areas of equal reflectivity as for radar-measured reflectivity. In this case, the model tends to overestimate on average reflectivity, while enabling us to obtain the qualitatively correct description of the convective phenomenon.

The proposed technique can be used to verify the simulation results using data on reflectivity obtained by a satellite or a meteoradar. The technique allows one to avoid subjectivity in the interpretation of simulation results and estimate the quality of reproducing the “general pattern” of the convective event.

How to cite: Svechnikova, E., Ilin, N., and Mareev, E.: Verification of modeling of convective events based on radar reflectivity, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5224, https://doi.org/10.5194/egusphere-egu21-5224, 2021.

 It is an essential problem for forecasting Mesoscale Convection Systems to understand the mechanism of interaction between atmospheric flow and vortices with the development of cumulonimbus clouds using a numerical weather model. In this research, potential temperature gradient based vorticity which is the expression of baroclinic is obtained to analyze the energy structure of the vorticity field in developing cumulonimbus. First, applying the variational method enables us to obtain a diagnostic equation in which the equation of motion, conservation law of mass, and entropy are considered as constraints. Second, Fourier analysis was performed on the vorticity field in the cross-section of the convective core in the isolated cumulonimbus simulation. The temporal change of the spectrum of the vorticity field indicates that the rotational intensity of potential temperature gradient based vorticity increases at the same time as the degree of baroclinicity increases. It was also found that the same tendency can be seen in the analysis of the vorticity field of developing clouds using the environment of the heavy rainfall event in the Kuma River basin that occurred on July 4, 2020. We are planning to analyze the vorticity field in the cluster of cumulonimbus clouds and consider the difference in the energy structure of the vorticity field due to the difference in model resolution. Third, we conducted the data assimilation experiment assuming the use of vertical vorticity estimated by doppler radar observation. As a result, the change in the potential temperature and vertical wind through the error covariance matrix generates coherent convection in the computations.

How to cite: Ono, A., Yamaguchi, K., and Nakakita, E.: Energy Structure Analysis of Vorticity Driven by Thermal Gradient in Developing Cumulonimbus Clouds and Application to Data Assimilation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-5570, https://doi.org/10.5194/egusphere-egu21-5570, 2021.

This study evaluates the latest Japanese Reanalysis for Three Quarters of a Century (JRA-3Q) conducted by the Japan Meteorological Agency (JMA), focusing on a semi-period of pre-satellite era (1960s and 1970s). The reanalysis is the third Japanese global atmospheric reanalysis covering the period from late 1940s onward, which is produced with the JMA's operational system as of December 2018. The atmospheric model has a TL479 horizontal resolution and 100 vertical layers up to 0.01 hPa, and the core component of the JRA-3Q data assimilation system is the 6-hourly 4D-Var of the atmospheric state with a T319-resolution inner model. Because there are only few global-covered observational datasets during the pre-satellite era, evaluation of the JRA-3Q is mainly to conduct an intercomparison of other reanalysis datasets such as representation Japanese 55-year Reanalysis (JRA-55), a JRA-55's subset of atmospheric reanalysis assimilating conventional observations only (JRA-55C), and version 3 of the Twentieth Century Reanalysis (20CRv3), and also an intercomparison of JRA-3Q between the pre-satellite and satellite eras. Emphasis of this evaluation during the non-satellite era is placed on the representation of tropical circulation, the consistency in time of the reanalysed fields, detection of tropical cyclones, and the quality of the stratospheric water vapor and ozone. For example, the surface circulation over the tropical Africa is improved by means of reducing spurious anticyclonic circulation anomalies that were found in JRA-55. Although the atmospheric model can produce self-generated quasi-biennial oscillation (QBO) by introducing non-orographic gravity wave drag, the evaluation reveals that JRA-3Q has a shorter period of around one year in the middle stratosphere and diminished QBO amplitude in the lower stratosphere, indicating that representation of the QBO in JRA-3Q is not as good as that in JRA-55.

How to cite: Naoe, H., Kobayashi, S., Kosaka, Y., Chiba, J., Tokuhiro, T., and Harada, Y.: Evaluation of a new Japanese reanalysis (JRA-3Q) in a pre-satellite era, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6881, https://doi.org/10.5194/egusphere-egu21-6881, 2021.

One important aspect of successfully implementing an ensemble Kalman filter (EnKF) in a high dimensional geophysical application is covariance localization. But for satellite radiances whose vertical locations are not well defined, covariance localization is not straightforward. The global group filter (GGF) is an adaptive localization algorithm, which can provide adaptively estimated localization parameters including the localization width and vertical location of observations for each channel and every satellite platform of radiance data, and for different regions and times. This adaptive method is based on sample correlations between ensemble priors of observations and state variables, aiming to minimize sampling errors of estimated sample correlations. The adaptively estimated localization parameters are examined here for typhoon Yutu (2018), using the regional model WRF and a cycling EnKF system. The benefits of differentiating the localization parameters for TC and non-TC regions and varying the localization parameters with time are investigated. Results from the 6-h priors verified relative to the conventional and radiance observations show that the adaptively estimated localization parameters generally produce smaller errors than the default Gaspari and Cohn (GC) localization. The adaptively estimated localization parameters better capture the onset of RI and yield improved intensity and structure forecasts for typhoon Yutu (2018) compared to the default GC localization. The time-varying localization parameters have slightly advantages over the time-constant localization parameters. Further improvements are achieved by differentiating the localization parameters for TC and non-TC regions.

How to cite: Wang, C., Lei, L., Tan, Z.-M., and Chu, K.: Adaptive Localization for Tropical Cyclones With Satellite Radiances in an Ensemble Kalman Filter, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8774, https://doi.org/10.5194/egusphere-egu21-8774, 2021.

The western Mediterranean region is frequently disrupted by heavy precipitation and flash flood episodes. Designing convection-permitting ensembles capable of accurately forecasting socially relevant aspects of these natural hazards such as timing, location, and intensity at basin scales of the order of a few hundred of squared kilometers is an extremely challenging effort. The usual forecast underdispersion prevailing at these scales motivates the research of sampling methodologies which are able to provide an adequate representation of the uncertainties in the initial atmospheric state and its time-integration by means of numerical models. This work investigates the skill of multiple techniques to sample model uncertainty in the context of heavy precipitation in the Mediterranean. The performance of multiple stochastic schemes is analyzed for a singular event occurred on 12 and 13 September 2019 in València, Murcia, and Almería (eastern Spain). This remarkable and enlightening episode caused seven casualties, the flooding of hundreds of homes and economic exceeding 425 million EUR.

Stochastic methods are compared to the popular multiphysics strategy in terms of both diversity and skill. The considered techniques include stochastic parameterization perturbation tendencies of state variables and perturbations to specific and influential parameters within the microphysics scheme (cloud condensation nuclei, fall speed factors, saturation percentage for cloud formation). The introduction of stochastic perturbations to the microphysics parameters results in an increased ensemble spread throughout the entire simulation. A conclusion of special relevance for the western Mediterranean, where local topography and deep moist convection play an essential role, is that stochastic methods significantly outperform the multiphysics-based ensemble, indicating a clear potential of stochastic parameterizations for the short-range forecast of high-impact events in the region.

How to cite: Hermoso, A., Homar, V., and Plant, R.: Improving heavy precipitation forecasting over the western Mediterranean: Benefits of stochastic techniques for model error sampling, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-6225, https://doi.org/10.5194/egusphere-egu21-6225, 2021.

In the context of Numerical Weather Prediction (NWP), continuous improvement from one version to another is made possible by the improvement of individual parts of the models. Thus the evaluation of those parts is crucial. Within a time step, we see the sequence of the resolved dynamic part and physical parametrizations. Similarly, within a data assimilation cycle, we see the sequence of forecast and analysis. These cyclical behaviours are responsible for a high coupling between the different parts of a NWP system. This means that, when evaluating an individual physical parametrization, a forecast only approach is not enough and simulations of the whole system with data assimilation over a long period are required.

In this work, we focus on the evaluation of the physical parametrization of deep convection in the French model ARPEGE. We evaluate the direct impact of this parametrization in a forecast only study as well as the indirect impact with a 4D-Var and the study of the analysis. We have replaced the previous parametrization by the one used in the Integrated Forecast System (IFS) developed at the ECMWF. We seize the opportunity of using an other model parametrization to rearrange physical tendencies in the same way as in the IFS. This diagnostic is new for the ARPEGE environment and it leads to an intecomparison between the two model physics. To evaluate the coupling, we use several ARPEGE 4D-Var to compare the change in analysis with an estimate of the analysis error. Those studies show a significant impact of the new scheme both in the tendencies and in the analysis.

How to cite: Hubans, A., Berre, L., Bouteloup, Y., Loo, C., and Marquet, P.: Impacts of a change in deep convection scheme on the ARPEGE data assimilation system, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7184, https://doi.org/10.5194/egusphere-egu21-7184, 2021.

Forecasting high impact weather events is a major challenge for numerical weather prediction. Initial condition uncertainty plays a major role but so potentially do uncertainties arising from the representation of physical processes, e.g. cloud microphysics. In this project, we investigate the impact of these uncertainties for the forecast of cloud properties, precipitation and hail of a selected severe convective storm over South-Eastern Germany.
To investigate the joint impact of initial condition and parametric uncertainty a large ensemble including perturbed initial conditions and systematic variations in several cloud microphysical parameters is conducted with the ICON model (at 1 km grid-spacing). The comparison of the baseline, unperturbed simulation to satellite, radiosonde, and radar data shows that the model reproduces the key features of the storm and its evolution. In particular also substantial hail precipitation at the surface is predicted. Here, we will present first results including the simulation set-up, the evaluation of the baseline simulation, and the variability of hail forecasts from the ensemble simulation.
In a later stage of the project we aim to assess the relative contribution of the introduced model variations to changes in the microphysical evolution of the storm and to the fore- cast uncertainty in larger-scale meteorological conditions.

How to cite: Kuntze, P., Miltenberger, A., Hoose, C., and Kunz, M.: Role of initial condition and parametric uncertainty in a severe hailstorm forecast, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7406, https://doi.org/10.5194/egusphere-egu21-7406, 2021.

Dynamic sea ice and ocean have long been recognised as an important components in the Earth System Models used to generate climate change projections and more recently seasonal forecasts. However, the benefit of forecasts on the timescales of days to weeks has received less attention. Until recently it was assumed that sea-ice-ocean fields change so slowly that it is acceptable to keep them fixed in short and medium-range forecasts. However, at the ice edge the presence of sea ice dramatically influences surface fluxes, particularly when the overlying atmosphere is much colder than the open ocean so errors in the position of the sea ice, caused by simply persisting this field, have the potential to degrade atmospheric skill. To address this and similar issues, the European Centre for Medium-range Weather Forecasts (ECMWF) recently took the pioneering step of coupling a dynamic–thermodynamic sea ice-ocean model to the Integrated Forecast System, developing the first coupled medium-range forecasting system. This was a major step towards making ECMWF’s forecasts seamless across all timescales.

In this study we assess the benefits of including coupled sea-ice ocean processes in the medium-range by comparing set of ten-day forecasts with and without dynamic ice-ocean coupling, focussing on forecast performance at the edge of the sea ice and in the surrounding region. We demonstrate that dynamic coupling improves forecasts of the sea ice edge at all leadtimes. Further, the skill gained is larger during periods when the ice edge is advancing or retreating rapidly. We will also explore whether dynamic coupling has an impact on forecast skill in atmospheric parameters downstream of the ice edge.  

How to cite: Day, J., Keeley, S., Mogensen, K., Tietsche, S., and Magnusson, L.: Benefits of ice-ocean coupling for medium-range forecasts in polar and sub-polar regions, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8054, https://doi.org/10.5194/egusphere-egu21-8054, 2021.

Since their first operational application in the 1950s, atmospheric numerical models have become essential tools in weather and climate prediction. As such, they are a constant subject to changes, thanks to advances in computer systems, numerical methods, and the ever increasing knowledge about the atmosphere of Earth. Many of the changes in today's models relate to seemingly unsuspicious modifications, associated with minor code rearrangements, changes in hardware infrastructure, or software upgrades. Such changes are meant to preserve the model formulation, yet the verification of such changes is challenged by the chaotic nature of our atmosphere - any small change, even rounding errors, can have a big impact on individual simulations. Overall this represents a serious challenge to a consistent model development and maintenance framework.

Here we propose a new methodology for quantifying and verifying the impacts of minor atmospheric model changes, or its underlying hardware/software system, by using ensemble simulations in combination with a statistical hypothesis test. The methodology can assess effects of model changes on almost any output variable over time, and can also be used with different hypothesis tests.

We present first applications of the methodology with the regional weather and climate model COSMO. The changes considered include a major system upgrade of the supercomputer used, the change from double to single precision floating-point representation, changes in the update frequency of the lateral boundary conditions, and tiny changes to selected model parameters. While providing very robust results, the methodology also shows a large sensitivity to more significant model changes, making it a good candidate for an automated tool to guarantee model consistency in the development cycle.

How to cite: Zeman, C. and Schär, C.: A new ensemble-based statistical methodology to verify changes in weather and climate models, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-4255, https://doi.org/10.5194/egusphere-egu21-4255, 2021.

Most of the precipitation formation in extratropical cyclones occurs in the warm sector along an elongated air stream ahead of the cold front - the so-called warm conveyor belt (WCB). The WCB ascends slantwise from the planetary boundary layer into the upper troposphere, where its outflow interacts with the upper-level jet and modifies the Rossby wave structure. The ascent of WCBs is strongly driven by cloud-condensational processes, which are parametrized in numerical weather prediction models, and is therefore associated with forecast uncertainty. In the European Centre for Medium-Range Weather Forecasts (ECMWF) ensemble prediction system (EPS), model uncertainty related to parametrizations is represented by the so-called stochastically perturbed parametrization tendencies (SPPT)-scheme, which introduces multiplicative noise to the physics tendencies.

In this study, we investigate the systematic effect of the SPPT-scheme on rapidly ascending air streams in the extratropics (i.e. WCBs) and on tropical convection by conducting sensitivity experiments with the ECMWF EPS based on the Integrated Forecasting System (IFS) model. The comparison of an experiment with an operational setup (initial condition and model physics perturbations) to one where model physics perturbations are switched off demonstrates that the SPPT-scheme systematically influences the activity of WCBs and tropical convection.

Globally, rapidly ascending air streams, which are detected by applying trajectory analysis in each ensemble member, are enhanced by about 37% when SPPT is activated. Also the dynamical and physical characteristics of the trajectories are systematically modified: the latent heat release and the ascent speed are increased, while the outflow latitude is decreased. This systematic modulation is stronger in the tropics and weaker in the extratropics. A detailed investigation of vertical velocities indicates that SPPT increases the frequency of relatively strong upward motion related to WCBs and tropical convection, while slower upward motion is suppressed compared to the unperturbed experiment. Despite the symmetric, zero-mean nature of the perturbations, the response of rapidly ascending air streams to the SPPT-scheme is systematically unidirectional, pointing towards non-linearities in the underlying processes.

This study shows that process-oriented diagnostics of weather systems help to advance the understanding of upscale impacts of the ensemble configuration on the representation of the large-scale circulation in numerical models.

How to cite: Pickl, M., Grams, C. M., Lang, S. T. K., and Leutbecher, M.: The effect of stochastically perturbed parametrization tendencies on rapidly ascending air streams, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9397, https://doi.org/10.5194/egusphere-egu21-9397, 2021.

For any ensemble-based data assimilation system sampling errors are introduced as a consequence of limited ensemble size, generating spurious backgound error covariances and leading to erroneous adjustments to the analysis. As a way to reduce the impact of these sampling errors, as well as improve rank deficiency, covariance localization is applied, which artifically reduces the weighting of error covariances beyond a defined physical distance between the background and observations deemed to be false.

In this study we perform sensitivity tests to find the appropriate horizontal localization scale for the SCALE-LETKF, a numerical weather prediction model that combines the SCALE regional model with the local ensemble transform Kalman filter. The system has been in development since 2013 to provide very high resolution modelling of convective weather systems and is unique in its ability to perform near real-time NWP operation at 500-m resolution refreshed every 30 seconds with observations from Phased Array Weather Radar (PAWR).  Here, we perform sensitivity tests at 500-m resolution with 30-second update cycling of PAWR data for several testcases of heavy convective rainfall over Tokyo metropolitan area from August/September 2019. Test scores showed horizontal localization scale of 2-km generally provided optimal forecast skill for lead times up to 30 minutes, although there were variations on this dependent upon lead time and case study. We show that by reducing localization scale, systematic errors leading to over-intensification of convective activity in forecasts were reduced, resulting in improved consistency with observations. This was a conseqence of generating more convectively stable, less dynamically active environment with smaller localization scale.

How to cite: Taylor, J., Honda, T., Amemiya, A., and Miyoshi, T.: Optimizing the localization scale for a convective-scale ensemble radar data assimilation , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9533, https://doi.org/10.5194/egusphere-egu21-9533, 2021.

Stochastic parameterisations are an important way to represent uncertainty in the deterministic forecasting models underlying ensemble prediction systems. In many of the currently used stochastic parameterisation approaches, random generators produce correlation patterns that induce spatially and temporally coherent perturbations to the parameterisation parameters or tendencies. The patterns that are currently used in the Harmonie ensemble prediction system are therefore unrelated to the atmospheric flow or weather situation. Here we investigate the potential of replacing such random patterns by accumulated tendency fields from parameterized physical processes in the model. The rationale hereby is that by perturbing the parameterisations with a field that reflects where parameterisations are most active, rather than a random pattern, the model obtains a more targeted increase in the degrees-of-freedom to represent forecasting uncertainty.

As an initial test case, we consider a large cold-air outbreak during 23-25 Dec 2015 that affected large parts of Scandinavia. During that time period, strong heat fluxes persisted near the ice edge, while widespread shallow convection dominated in the center of the model domain. For diagnosing the perturbation fields, we utilise an implementation of individual tendency diagnostics implemented in AROME-Arctic within the ALERTNESS project. Total physical tendencies for the horizontal wind components, for air temperature and humidity are accumulated with a time filtering throughout the 66 h forecast period.

The accumulated tendencies from all parameterisations for the different variables show overlapping and differing centers of activity. Wind parameterisations are active near the ice edge, and with smaller scale variability over land areas, in particular at lower model levels. Temperature tendency patterns show activity that is more confined to the ice edge, and a narrow coastal stripe along Northern Scandinavia. These first results show that the approach provides spatially coherent patterns of parameterisation activity, which are meaningfully related to the dominating weather situation. Based on sensitivity tests of cloud parameterisation parameters in a single-column version, we outline the next steps in the path towards diagnostic perturbation patterns for stochastically perturbed perturbations in the Harmonie EPS system.

How to cite: Sodemann, H., Kähnert, M., Valkonen, T. M., Ekrem, P., and Frogner, I.-L.: Potential of accumulated AROME-Arctic parameterisation tendency for stochastic parameterisation perturbation patterns, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9878, https://doi.org/10.5194/egusphere-egu21-9878, 2021.

Spain, having a complex topography, has many climate and weather particularities, acting in many aspects like a mini continent. This is shown in many aspects, such as supercells, which count for more than 1000 in the last 10 years. This indicates that severe weather happens yearly, and supercell thunderstorms are one of the biggest threats, producing damage to population and economical assets, which makes reliable supercell forecast for risk management and mitigation a priority.

This research evaluates supercell forecasts from the Weather Research and Forecasting (WRF-ARW) model over Spain. This first iteration analyzes 2018 supercells, trying to predict this events using three nested domains (15-3-1 km), feeded with GFS operational datasets. The configuration chosen for the model has been used in the past for a master's thesis, with great results, and thus this work aims to evalute the operational usage of this configuration for prediction with 12-36 hours of anticipation. Results so far show that around 80% of supercells could be perfectly forecasted, and another 15% could have medium forecasting skill. This results show that risk alarms could have been issued if this forecasts had being operative at the moment.

How to cite: Granda-Maestre, R., Calvo-Sancho, C., and Martín, Y.: Supercell predictability on Iberian Peninsula using WRF-ARW model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10376, https://doi.org/10.5194/egusphere-egu21-10376, 2021.

Floods and puddles are incidents that occur every year in Mexico City. The surface runoff that occurs in areas of hills and mountains, such as torrential rains where precipitation is greater than the drainage capacity, are the main factors that give rise to floods in the city. The measures that have been implemented to control floods have focused more on reactive planning instead of implementing prevention measures; so the city is completely dependent on its drainage system to mitigate flooding. For these reasons, the forecast has become essential to respond to the demand for better risk management due to the exposure of infrastructure and people to flood events; and coupled with the uncertainty of future events in Mexico City.

Rainfall is the main source of uncertainty in flood prediction; That is why, in recent years, the Numerical Climate Prediction Models (NWP) have focused on the generation of Ensemble Prediction Systems (EPS); which constitute a feasible method to predict the probability distribution function of atmospheric evolution.

The objective of this work is to evaluate the Operational Ensemble Prediction System issued by the European Centre for Medium-Range Weather Forecasts (ECMWF) to open the doors to the development of a Flood Forecasting System in Mexico City. The EPS was evaluated against observed rainfall for two study zones: Mexico Valley Basin and Mexico City, where for the latter, the forecasts were compared against information of real time observed rainfall. To carry out an objective analysis of the quality of the forecast, metrics were applied for the scalar attributes: precision, reliability, resolution, discrimination and performance. The probabilities given by the ensembles were estimated using a predictive model.

The results show the EPS do represent the probability distribution of the observed events. The first 36 hours of forecasting are the most reliable, after which uncertainty increases. Finally, the predictive model shows good performance in estimating probabilities according to the area under the receiver operating characteristic curve.

How to cite: López López, M. R. and Pedrozo Acuña, A.: Evaluation of the ECMWF Operational Forecasting System for Probabilistic Flood Prediction in Mexico City, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13602, https://doi.org/10.5194/egusphere-egu21-13602, 2021.

The new Copernicus European Regional Re-Analysis (CERRA) is a 5.5km reanalysis, starting in 1984 and ending “near-real-time”, 2021. The reanalysis was delivered using the ALADIN model under the HARMONIE scripting garb. The upper-air is analysed using a 3DVAR technique cycled 3-hourly, while the surface analysis is achieved through a conventional OI technique (MESCAN). Analyses produced by CERRA at 5.5km are assisted through an accompanying 10-member Ensemble Data Assimilation (EDA) system with 11km horizontal resolution cycled 6-hourly. The EDA system is used mainly for serially updated background error covariance estimation (B-matrix) used in the deterministic upper-air 3DVAR minimisation to produce the upper-air analysis.

The B-matrix comprises 2 principal EDA-derived components. The first component is estimated from same-resolution (5.5km) forecast differences, run in the winter and the summer periods, to represent seasonal climatology. This component also varies in time, such that a linearly appropriated proportion of summer or winter differences is taken, based on the current time of year of the reanalysis. The second component comes from the lower-resolution (11km) set of forecast differences, which represents ‘errors of the day’. This second component is a 2.5 day moving average ingested into a new B-matrix every 2 days. The B-matrix is thus comprised of 80% forecast differences coming from the first component and 20% coming from the second component. 

We show results from our study on the primacy of varying the weighting on the 2 components of forecast differences mentioned above, and how it has the potential, given a suitable observation network, to provide better B-matrix statistics.

How to cite: El-Said, A., Brousseau, P., and Ridal, M.: Investigating the primacy of B-matrix EDA flow dependence within the Copernicus Regional Re-Analysis (CERRA), EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15543, https://doi.org/10.5194/egusphere-egu21-15543, 2021.

The German Weather Service (DWD) introduces the regional NWP model ICON-LAM (ICON Limited Area Mode) in 2021 to replace the COSMO model. For the ICON-LAM data assimilation, a novel EnVAR (Ensemble VARiational data assimilation) setup is currently evaluated in comparison to the operational deterministic run of KENDA-LETKF (Local Ensemble Transform Kalman Filter). This requires special care as the observation handling differs for the global assimilation (via EnVAR) and the regional assimilation (KENDA). Furthermore, the variational quality control for the regional EnVAR may require a setup differing from the global setup. We will give an introduction to the observation processing in DWD's data assimilation framework (DACE).

For future development, we give an outlook on how a regional EnVAR can be used for a regional deterministic analysis by using a global ICON ensemble in combination with a regional deterministic ICON-LAM run. This is potentially of interest for DWD's partners with smaller computational capacities, because a regional EnVAR analysis is computationally less expensive than running a full KENDA ensemble assimilation cycle.

How to cite: Burba, M., Ulbrich, S., Hollborn, S., Potthast, R., and Knippertz, P.: EnVAR for ICON-LAM: observations and quality control, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-15603, https://doi.org/10.5194/egusphere-egu21-15603, 2021.

Short or mid-term rainfall forecasting is a major task with several environmental applications such as agricultural management or flood risks monitoring. Existing data-driven approaches, especially deep learning models, have shown significant skill at this task, using only rainfall radar images as inputs. In order to determine whether using other meteorological parameters such as wind would improve forecasts, we trained a deep learning model on a fusion of rainfall radar images and wind velocity produced by a weather forecast model. The network was compared to a similar architecture trained only on radar data, to a basic persistence model and to an approach based on optical flow. Our network outperforms by 8% the F1-score calculated for the optical flow on moderate and higher rain events for forecasts at a horizon time of 30 minutes. Furthermore, it outperforms by 7% the same architecture trained using only rainfall radar images. Merging rain and wind data has also proven to stabilize the training process and enabled significant improvement especially on the difficult-to-predict high precipitation rainfalls. These results can also be found in Bouget, V., Béréziat, D., Brajard, J., Charantonis, A., & Filoche, A. (2020). Fusion of rain radar images and wind forecasts in a deep learning model applied to rain nowcasting. arXiv preprint arXiv:2012.05015

How to cite: Charantonis, A., Bouget, V., Béréziat, D., Brajard, J., and Filoche, A.: Fusion of rain radar images and wind forecasts in adeep learning model applied to rain nowcasting, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11990, https://doi.org/10.5194/egusphere-egu21-11990, 2021.

It is important to utilize various hydrological and weather information and accurate real-time forecasts to understand the hydrological conditions of the dam in order to make decisions of dam operation. In particular, due to rainfall concentrated in a short period of time during the flood season, it is necessary to plan the exact amount of dam discharge using real-time rainfall forecasting information. Compared to the ground rain gauge network, the radar has a high resolution of time and space, which enables the continuous expression of rainfall, which is very advantageous for very short-term prediction. Especially, In particular, the radar is capable of three-dimensional observation of the atmosphere, which has an advantage in understanding the vertical development and structure of clouds and rainfall, which can be used to observe torrential rain in the dam basin and to anticipate future rainfall intensity changes, rainfall movement and duration time. This study aims to develop a suitable radar-based very short-term rainfall prediction technique and to produce rainfall prediction information of the dam basin for stable dam operation and water disaster prevention. The radar-based rainfall prediction in this study is to be performed using a convolutional deep neural network with the 8 years weather radar data of the Korea Meteorological Administration. And, we select rainfall cases with high rainfall intensity and train the deep neural network to ensure the accuracy of flood season rainfall prediction. In addition, we intend to perform the accuracy evaluation with extrapolation-based rainfall prediction results for the dam basin.

This work was supported by KOREA HYDRO & NUCLEAR POWER CO., LTD (No. 2018-Tech-20)

How to cite: Yoon, S. and Shin, H.: Very short-term radar rainfall prediction using deep neural network for hydropower dam operation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1054, https://doi.org/10.5194/egusphere-egu21-1054, 2021.

Forecasting valley winds over complex terrain using a coarse horizontal resolution mesoscale model is a challenging task. Mesoscale models such as
the Weather Research and Forecasting (WRF) model tend to perform poorly over such regions. In this study, we assess the added value of downscaling
WRF wind forecasts using artificial neural networks (ANN) over the Cadarache Valley which is located in southeast France. Wind forecasts over the Cadarache valley are generated using WRF with a horizontal resolution of 3km on a daily basis. We used performance metrics such as Directional ACCuracy (DACC) and mean absolute error (MAE) for the evaluation of the WRF and ANN. WRF horizontal wind components at 110m and the near surface vertical potential temperature gradient were used as input data and observed horizontal wind components at 10m within the valley as targets during ANN training. We found an increase of DACC from 56% to 79% after post-processing WRF forecasts with ANN. Further analysis show that the ANN performed well during day and night, but poorly during morning and afternoon transition. The performance of WRF has a huge influence on ANN performance with bad WRF forecasts affecting ANN performance. However, the ANN improves the poor WRF forecasts to a DACC exceeding 60%. A change in lead time and domain resolution showed negligible impact suggesting that 3km resolution and a lead time of 24-47h is effective and relatively cheap to apply. Additionally, WRF performs well in near-neutral conditions and poorly in other atmospheric stability conditions. However ANN showed a consistent improvement in wind forecast during all stability classes with a DACC of nearly 80%. The study clearly demonstrates the ability to improve Cadarache valley wind forecasts using ANN from WRF simulations on a daily basis.

How to cite: Chinyoka, S., Hedde, T., and Steeneveld, G.-J.: Using an Artificial Neural Network to improve operational wind prediction in a small unresolved valley, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3800, https://doi.org/10.5194/egusphere-egu21-3800, 2021.

Potential vorticity streamers (PVSs) are elongated quasi-horizontal filaments of stratospheric air in the upper troposphere related to, for example, Rossby wave breaking events. They are known to be related to partly extreme weather events in the midlatitudes and subtropics and can also be involved in (sub-)tropical cyclogenesis. While several algorithms have been developed to identify and track PVSs on planar isentropic surfaces, less is known about the evolution of these streamers in 3D, both climatologically but also for a better understanding of individual weather events. Furthermore, characteristics of their 3D shape have barely been considered as a predictor for high impact weather events like (sub-)tropical cyclones.

We introduce a novel algorithm for detection and identification of PVSs based on image processing techniques which can be applied to 2D and 3D gridded datasets. The potential vorticity was taken from high resolution isentropic analyses based on the ERA5 dataset. The algorithm uses the 2 PVU (Potential Vorticity Unit) threshold to identify and extract anomalies in the PV field using signed distance functions. This is accomplished by using a stereographic projection to eliminate singularities and keeping track of the reduced distortions by storing precomputed distance maps. This approach is computationally efficient and detects more interesting structures that exhibit the general behavior of PVSs compared to existing 2D techniques.

For each identified object a feature vector is computed, containing the individual characteristics of the streamers. In the 3D case, the algorithm looks at the structure en bloc instead of operating individually on multiple 2D levels. This also makes the identification stable regarding the seasonal cycle. Feature vectors contain parameters about quality, intensity and shape. In the case of 2D datasets, best-fitting ellipses computed from the statistical moments are regarded as a description of their shape. For 3D datasets, recent visualizations show that the boundary of these structures could be approximated by quadric surfaces . The feature vectors are also amended by tracking information, for example splitting and merging events. This low-dimensional representation serves as base for ERA5 climatologies. The data will be correlated with (sub-)tropical cyclone occurrence to spot useful and novel predictors for cyclone activity and preceding Rossby Wave Breaking events.

Overall, this new type of PVS identification algorithm, applicable in 2D or 3D, allows to diagnose the role of PVS in extreme weather events, including their predictability in ensemble forecasts.

How to cite: Fischer, C., Schömer, E., Fink, A. H., Riemer, M., and Maier-Gerber, M.: A novel identification and tracking method of weather-relevant 3D Potential vorticity streamers, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9115, https://doi.org/10.5194/egusphere-egu21-9115, 2021.

Six adaptive post-processing methods for correcting systematic biases in forecasts of near-surface air temperatures, using local meteorological observations, are assessed and compared. The methods tested are based on the simple moving average and the more advanced Kalman filter - constructed to remove the longer-term bias, the very short-term errors or a combination of the two. Forecasts from a coarser-resolution global model and a regional high-resolution model are post-processed and the results are evaluated for one hundred private weather stations in Denmark. Overall, the postprocessing method for which a moving average is combined with a Kalman filter, constructed to remove the very short-term errors, performs the best. The biases of both the global coarserresolution forecasts and the regional high-resolution forecasts are reduced close to zero for all forecast lead times. The standard deviation is reduced for all forecast lead times for the coarser resolution model, whereas for the high-resolution model the most significant reduction is seen for the first six forecast lead hours. This shows that the application of a relatively simple postprocessing method, based on a short training period, can give good results.

How to cite: Kaas, E. and Alerskans, E.: Local temperature forecasts based on post-processing of Numerical Weather Prediction data, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11270, https://doi.org/10.5194/egusphere-egu21-11270, 2021.

Numerical weather prediction (NWP) models are known to exhibit systematic errors, especially for near-surface variables such as air temperature. This is partly due to deficiencies in the physical formulation of the model dynamics and the inability of these models to successfully handle sub-grid phenomena. Forecasts that better match the locally observed weather can be obtained by post-processing NWP model output using local meteorological observations. Here, we have implemented a non-linear post-processing model based on machine learning techniques with the aim of post-processing near-surface air temperature forecasts from a global coarse-resolution model in order to produce localized forecasts. The model is trained on observational from a network of private weather stations and forecast data from the global coarse-resolution NWP model. Independent data is used to assess the performance of the model and the results are compared with the performance of the raw NWP model output. Overall, the non-linear machine learning post-processing method reduces the bias and the standard deviation compared to the raw NWP forecast and produces a forecast that better match the locally observed weather.

How to cite: Alerskans, E., Nyborg, J., Birk, M., and Kaas, E.: Prediction of near-surface temperatures using a non-linear machine learning post-processing model, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11378, https://doi.org/10.5194/egusphere-egu21-11378, 2021.

Probability distributions based on ensemble forecasts are commonly used to assess uncertainty in weather prediction. However, interpreting these distributions is not trivial, especially in the case of multimodality with distinct likely outcomes. The conventional summary employs mean and standard deviation across ensemble members, which works well for unimodal, Gaussian-like distributions. In the case of multimodality this misleads, discarding crucial information. 

We aim at combining previously developed clustering algorithms in machine learning and topological data analysis to extract useful information such as the number of clusters in an ensemble. Given the chaotic behaviour of the atmosphere, machine learning techniques can provide relevant results even if no, or very little, a priori information about the data is available. In addition, topological methods that analyse the shape of the data can make results explainable.

Given an ensemble of univariate time series, a graph is generated whose edges and vertices represent clusters of members, including additional information for each cluster such as the members belonging to them, their uncertainty, and their relevance according to the graph. In the case of multimodality, this approach provides relevant and quantitative information beyond the commonly used mean and standard deviation approach that helps to further characterise the predictability.

How to cite: Galmiche, N., Blaser, N., Brun, M., Hauser, H., Spengler, T., and Spensberger, C.: Exploring multi-modalities in weather prediction using a univariate graph based on machine learning techniques, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-11747, https://doi.org/10.5194/egusphere-egu21-11747, 2021.

Forecasting extreme rainfall in the tropics is a major challenge for numerical weather prediction. Convection-permitting (CP) models are intended to enable forecasts of high-impact weather events. Development and operation of these models in the tropics has only just been realised. This study describes and evaluates recently developed Met Office Unified Model CP ensemble forecasts of varying resolutions over three domains in Southeast Asia, covering Malaysia, Indonesia and the Philippines.

Fractions Skill Score is used to assess the spatial scale-dependence of skill in forecasts of precipitation during October 2018 - March 2019. CP forecasts are skilful for 3-hour precipitation accumulations at spatial scales greater than 200 km in all domains during the first day of forecasts but all ensembles have low spread relative to forecast skill. Skill decreases with lead time and is highly dependent on the diurnal cycle over Malaysia and Indonesia. Skill is largest during daytime when precipitation is over land and is constrained by orography, but is lower at night when precipitation is over the ocean. Comparisons of CP ensembles using 2.2, 4.5 and 8.8 km grid spacing and an 8.8km ensemble with parameterised convection are made to examine the role of resolution and convection parameterisation on forecast skill for the three domains.

How to cite: Ferrett, S., Frame, T., Methven, J., Holloway, C., Webster, S., Stein, T., and Cafaro, C.: Evaluating convection-permitting ensemble forecasts of precipitation over Southeast Asia , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-12375, https://doi.org/10.5194/egusphere-egu21-12375, 2021.

Convection-permitting ensemble prediction systems (CP-ENS) have been implemented in the
mid-latitudes for weather forecasting timescales over the past decade, enabled by the increase in
computational resources. Recently, efforts are being made to study the benefits of CP-ENS for
tropical regions. This study examines CP-ENS forecasts produced by the UK Met Office over
tropical East Africa, for 24 cases in the period April-May 2019. The CP-ENS, an ensemble with
parametrized convection (Glob-ENS), and their deterministic counterparts are evaluated against
rainfall estimates derived from satellite observations (GPM-IMERG). The CP configurations have
the best representation of the diurnal cycle, although heavy rainfall amounts are overestimated
compared to observations. Pairwise comparisons between the different configurations reveal that
the CP-ENS is generally the most skilful forecast for both 3-h and 24-h accumulations of heavy
rainfall (97th percentile), followed by the CP deterministic forecast. More precisely, probabilistic
forecasts of heavy rainfall, verified using a neighbourhood approach, show that the CP-ENS is
skilful at scales greater than 100 km, significantly better than the Glob-ENS, although not as good
as found in the mid-latitudes. Skill decreases with lead time and varies diurnally, especially for
CP forecasts. The CP-ENS is under-spread both in terms of forecasting the locations of heavy
rainfall and in terms of domain-averaged rainfall. This study demonstrates potential benefits in
using CP-ENS for operational forecasting of heavy rainfall over tropical Africa and gives specific
suggestions for further research and development, including probabilistic forecast guidance.

How to cite: Cafaro, C., Woodhams, B. J., Stein, T. H. M., Birch, C. E., Webster, S., Bain, C. L., Hartley, A., Clarke, S., Ferrett, S., and Hill, P.: Do convection-permitting ensembles lead to more skilful short-range probabilistic rainfall forecasts over tropical East Africa ?, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-13689, https://doi.org/10.5194/egusphere-egu21-13689, 2021.

Heatwaves can have devastating impact on society and reliable early warnings at several weeks lead time are needed. Heatwaves are often associated with quasi-stationary Rossby waves, which interact with sea surface temperature (SST). Previous studies showed that north-Pacific SST can provide long-lead predictability for eastern U.S. temperature, moderated by an atmospheric Rossby wave. The exact mechanisms, however, are not well understood. Here we analyze Rossby waves associated with heatwaves in western and eastern US. Causal inference analyses reveal that both waves are characterized by positive ocean-atmosphere feedbacks at synoptic timescales, amplifying the waves. However, this positive feedback on short timescales is not the causal mechanism that leads to a long-lead SST signal. Only the eastern US shows a long-lead causal link from SSTs to the Rossby wave. We show that the long-lead SST signal derives from low-frequency PDO variability, providing the source of eastern US temperature predictability. We use this improved physical understanding to identify more reliable long-lead predictions. When, at the onset of summer, the Pacific is in a pronounced PDO phase, the SST signal is expected to persist throughout summer. These summers are characterized by a stronger ocean-boundary forcing, thereby more than doubling the eastern US temperature forecast skill, providing a temporary window of enhanced predictability.

How to cite: Vijverberg, S. and Coumou, D.: The role of the Pacific Decadal Oscillation and ocean-atmosphere interactions in driving United States heatwaves, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2488, https://doi.org/10.5194/egusphere-egu21-2488, 2021.

Sub-seasonal heatwave-driven concurrent hot and dry extreme events (HDEs) can cause substantial damage to crops, and hence to lives and livelihoods. However, the physical processes that lead to these devastating events are not well-understood.

Based on observations and reanalysis data for 1979-2016 over China, we show that HDEs occur preferentially over central and eastern China (CEC) and southern China (SC), with a maximum of 3 events year^-1 along the Yangtze Valley. The probability of longer-lived and potentially more damaging HDEs is larger in SC than in CEC. Over SC the key factors of HDEs—positive anomalies of surface air temperature and evapotranspiration, and negative anomalies of soil moisture—begin two pentads before maximising at the peak of the HDEs. These anomalies occur south of a positive height anomaly at 200 hPa, associated with a large-scale subsidence anomaly. The processes over CEC are similar to SC, but the anomalies begin one pentad before the peak. HDE frequency is strongly related to the Silk Road Pattern and the Boreal Summer Intraseasonal Oscillation. Positive phases of the Silk Road Pattern and suppressed phases of the Boreal Summer Intraseasonal Oscillation are associated with positive height anomalies over CEC and SC, increasing HDE frequency by about 35-54% relative to the climatological mean.  Understanding the effects of sub-seasonal and seasonal atmospheric circulation variability, such as the Silk Road Pattern and Boreal Summer Intraseasonal Oscillation, on HDEs is important to improve HDE predictions over China.

How to cite: Tian, F., Klingaman, N., and Dong, B.: The driving processes of concurrent hot and dry extreme events in China, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-2643, https://doi.org/10.5194/egusphere-egu21-2643, 2021.

How to cite: Howard, E., Gonzalez, P., Thomas, S., Frame, T., Martinez-Alvarado, O., Methven, J., and Woolnough, S.:  Weather regimes in South East Asia: connections with synoptic phenomena and high impact weather, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7472, https://doi.org/10.5194/egusphere-egu21-7472, 2021.

How to cite: Gonzalez, P., Howard, E., Thomas, S., Frame, T., Martinez-Alvarado, O., Methven, J., and Woolnough, S.: Weather regimes in South East Asia: Sub-seasonal predictability of the regimes and the associated high impact weather  , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7411, https://doi.org/10.5194/egusphere-egu21-7411, 2021.

Central Southwest Asia (CSWA) is a region with the largest number of glaciers, outside the polar regions in its northeast and the Arabian desert to its southwest. The region receives precipitation from November to April period also known as the wet season, which contributes to the regional freshwater resources. Mainly, El Niño–Southern Oscillation (ENSO) modulates the wet season precipitation over CSWA, with a positive relationship. However, the intraseasonal characteristics of ENSO influence are largely unknown, which may be important to understand the regional sub-seasonal to seasonal hydroclimate variability. We noted that the ENSO‐CSWA teleconnection varies intraseasonally and is a combination of direct and indirect positive influences. The ENSO direct influence is through a Rossby wave‐like pattern in the tail months of the wet season, while the indirect influence is noted through an ENSO‐forced atmospheric dipole of diabatic heating anomalies in the tropical Indian Ocean (TIO), which also generates a Rossby wave‐like forcing and persists throughout the wet season. The stronger ENSO influence is found when both direct and indirect modes are in phase, while the relationship breaks down when the two modes are out of phase. Moreover, the idealized numerical simulations confirm and reproduce the observed circulation patterns. This suggests that improvements in sub-seasonal to seasonal scale predictability requires the better representation of intraseasonal variability of ENSO teleconnection, as well as the role of interbasin interactions in its propagation.

How to cite: Abid, M. A., Ashfaq, M., Kucharski, F., Evans, K. J., and Almazroui, M.: Indian Ocean mediates the ENSO teleconnections to the Central Southwest Asia during the wet season , EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-10315, https://doi.org/10.5194/egusphere-egu21-10315, 2021.

We investigated the spatial structure of the intraseasonal variation (15-30 day) in cloud cover in the mid-latitudes during winter. We attempted to interpret the spatial pattern of clouds in　the context of Rossby waves.

We used the total cloud cover in H-series dataset (1984-2016) by the International Satellite Cloud Climatology Project (ISCCP) based on the satellite observations, and ERA-Interim re-analysis data (1980-2016) including high, medium, and low cloud covers defined by σ coordinate.

We calculated correlation coefficients between the geopotential height at 300hPa (Z300) at a certain position and the cloud covers, meridional wind, and vertical velocity in the surrounding area. The positions of the maximum of high (0.45≧σ) and medium cloud cover (0.8≧σ＞0.45) relative to Z300 are longitudinally constant for all longitudes except the region from east Asia to western part of the Pacific. The position of the maximum of the high cloud cover is located just west of the ridge and just east of the maximum positions of the upward motions of re-analysis vertical velocity and its adiabatic component. These results suggest that the adiabatic upward motion in the southerly wind region west of the ridge contributes to the generation of high cloud cover. In contrast, the position of the maximum of medium cloud cover is located just east of the trough. The position of the maximum of diabatic upward motion, which is consider to be due to condensation process is located near the maximum of medium cloud cover. These results suggest that Rossby waves modulate activity of short-period disturbances with precipitation. Apart from high and medium cloud covers, the position of the maximum of low cloud cover (σ＞0.8) has large longitudinal dependency. While the position of the maximum is located at almost the same as that of medium cloud cover mainly over the continent, the position of the maximum is located just east of the ridge mainly over the ocean.

The correlation coefficients between ISCCP total cloud cover and Z300 are statistically significant only over the continent, where the positions of the maximum of high, medium, and low cloud covers are all located east of the trough and west of the ridge.

How to cite: Satoh, R., Nishi, N., and Mukougawa, H.:  Intraseasonal Variability of Cloud Cover in Midlatitudes during Boreal Winter, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3804, https://doi.org/10.5194/egusphere-egu21-3804, 2021.

Using the ensemble empirical mode decomposition (EEMD) method, this study systematically investigates the multiple timescales of the Southern Annular Mode (SAM) and identifies their relative contributions to the low-frequency persistence of SAM. Analyses show that the subseasonal sustaining of SAM mainly depends on the contribution of longer-timescale variabilities, especially the cross-seasonal variability. When subtracting the cross-seasonal variability from the SAM, the positive covariance between the eddy and zonal flow, which is suggested the positive eddy feedback in SAM, disappears. Composite analysis shows that only with strong cross-seasonal variability, the meridional shift of zonal wind, eddy momentum forcing and baroclinicity anomalies can be maintained for more than 20 days, mainly resulting from the longer-timescale (especially the cross-seasonal timescale) eddy-zonal flow interactions. This study further suggests that the dipolar sea surface temperature (SST) anomalies in the mid latitude of Southern Hemisphere (SH) is a possible cause for the cross-seasonal variability. Analysis shows that about half of the strong cross-seasonal timescale events are accompanied by evident dipolar SST anomalies, which mostly occur in austral summer. The cross-seasonal dependence of the eddy-zonal flow interactions suggests the longer-timescale (especially the cross-seasonal timescale) contribution cannot be neglected in subseasonal prediction of SAM.

How to cite: Zhang, Q., Zhang, Y., and Wu, Z.: Multiple Timescales of the Southern Annular Mode, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3667, https://doi.org/10.5194/egusphere-egu21-3667, 2021.

The strength of the stratospheric polar vortex is a key contributor to subseasonal prediction during boreal winter. Anomalously weak polar vortex events can be induced by enhanced vertically propagating Rossby waves from the troposphere, driven by blocking and wave breaking. Here, we analyse a tropospheric pattern—the Scandinavia–Greenland (S–G) pattern—associated with both processes. The S–G pattern is defined as the second empirical orthogonal function (EOF) of mean sea‐level pressure in the northeast Atlantic. The first EOF is a zonal pattern resembling the North Atlantic Oscillation. We show that the S–G pattern is associated with a transient amplification of planetary wavenumber 2 and meridional eddy heat flux, followed by the onset of a persistently weakened polar vortex. We then analyse 10 different models from the S2S database, finding that, while all models represent the structure of the S–G pattern well, some models have a zonal bias with more than the observed variability in their first EOF, and accordingly less in their second EOF. This bias is largest in the models with the lowest resolution, and consistent with biases in blocking and Rossby wave breaking in these models. Skill in predicting the S–G pattern is not high beyond week 2 in any model, in contrast to the zonal pattern. We find that the relationship between the S–G pattern, enhanced eddy heat flux in the stratosphere, and a weakened polar vortex is initially well represented, but decays significantly with lead time in most S2S models. Our results motivate improved representation of the S–G pattern and its stratospheric response at longer lead times for improved subseasonal prediction of the stratospheric polar vortex.

How to cite: Lee, S., Charlton-Perez, A., Furtado, J., and Woolnough, S.: Representation of the Scandinavia-Greenland Pattern and its Relationship with the Polar Vortex in S2S Models, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1370, https://doi.org/10.5194/egusphere-egu21-1370, 2021.

How to cite: Kolstad, E. W., Wulff, C. O., Domeisen, D., and Woollings, T.: Tracing North Atlantic Oscillation Forecast Errors to Stratospheric Origins, with a new analysis of the 2021 winter, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14216, https://doi.org/10.5194/egusphere-egu21-14216, 2021.

The ECMWF sub-seasonal forecast model includes dynamic representations of the atmosphere, ocean, sea-ice, and ocean-waves. Coupling to a dynamic ocean model allows more a realistic representation of air-sea interaction, but also introduces the potential for systematic errors in sea surface temperatures (SST). Here, we show that North Atlantic SST biases associated with errors in the position of the Gulf Stream have a significant impact on initialized forecasts at the sub‐seasonal time range. Correcting these errors with an online SST bias‐correction scheme improves the mean state of the North Atlantic region and has a significant positive impact on forecasts of atmospheric circulation anomalies. Improvements to forecast skill extend beyond the North Atlantic into Europe and along the northern hemisphere subtropical waveguide. These impacts provide important evidence for the potential benefits to initialized coupled forecast systems of higher‐resolution ocean models that can better resolve the position of the Gulf Stream.

Reference: Roberts, C. D., Vitart, F., & Balmaseda, M. A. Hemispheric impact of North Atlantic SSTs in sub‐seasonal forecasts. Geophysical Research Letters, e2020GL091446.

How to cite: Roberts, C., Vitart, F., and Balmaseda, M.: Remote impact of North Atlantic sea surface temperature errors in sub‐seasonal forecasts, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-846, https://doi.org/10.5194/egusphere-egu21-846, 2021.

The Tibetan Plateau (TP), referred to as the “Asian water tower”, contains one of the largest land ice masses on Earth. The local glacier shrinkage and frozen-water storage are strongly affected by variations in surface air temperature over the TP (TPSAT), especially in springtime. This study reveals a distinct out-of-phase connection between the February North Atlantic Oscillation (NAO) and March TPSAT, which is non-stationary and regulated by the warm phase of the Atlantic Multidecadal Variability (AMV+). The results show that during the AMV+, the negative phase of the NAO persists from February to March, and is accompanied by a quasi-stationary Rossby wave train trapped along a northward-shifted subtropical westerly jet stream across Eurasia, inducing an anomalous adiabatic descent that warms the TP. However, during the cold phase of the AMV, the negative NAO does not persist into March. The Rossby wave train propagates along the well-separated polar and subtropical westerly jets, and the NAO−TPSAT connection is broken. Further investigation suggests that the enhanced synoptic eddy and low-frequency flow (SELF) interaction over the North Atlantic in February and March during the AMV+, caused by the enhanced and southward-shifted storm track, help maintain the NAO anomaly pattern via positive eddy feedback. This study provides a new detailed perspective on the decadal variability of the North Atlantic−TP connections in late winter−early spring.

How to cite: Li, J., Li, F., He, S., Wang, H., and Orsolini, Y. J.: The AMV phase-dependence of the connection between February NAO and March surface air temperature over the Tibetan Plateau, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9879, https://doi.org/10.5194/egusphere-egu21-9879, 2021.

Issuing skillful forecasts beyond the typical horizon of weather predictability remains a challenge actively addressed by the scientific community. This study evaluates winter subseasonal reforecasts delivered by the CNRM and ECMWF dynamical systems and identifies that the level of skill for predicting temperature in Europe varies fairly consistently in both systems. In particular, forecasts initialized during positive NAO phases tend to be more skillful over Europe at week three in both systems. Composite analyses performed in an atmospheric reanalysis, a long-term climate simulation and both forecast systems unveil very similar temperature and sea-level pressure patterns 3 weeks after NAO+ conditions. Furthermore, regressing these fields onto the 3-week previous NAO index in a reanalysis shows consistent patterns over Europe but also eastern North America, thereby revealing a lagged teleconnection, either related to the persistence or recurrence of the NAO+ weather regime. Since this feature is well captured by forecast systems, this is a key mechanism for determining a priori confidence in the skill of wintertime subseasonal forecasts over Europe and North America.

How to cite: Ardilouze, C., Specq, D., Batté, L., and Cassou, C.: Flow dependence of wintertime subseasonal prediction skill over Europe, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-8666, https://doi.org/10.5194/egusphere-egu21-8666, 2021.

The accuracy and reliability in predicting winter anomalies, particularly high-impact events, is crucial for economic sectors like energy production and trade. Sub-seasonal predictions can provide a useful tool for early detection of these events. In this context, this study aims to target atmospheric patterns leading to skillful winter predictions at S2S lead times.

With a focus on Europe, we explore extended range predictability (up to 35 days) in the ECMWF hindcast dataset (1999-2018). This dataset provides a sizable sample for assessing the winter months predictive skill of the model and can be considered as a preparatory step to the use of the more comprehensive real-time ensemble forecasts.

The verification is performed on geopotential and temperature fields against the ERA5 reanalysis. We first identify the most skillful predictions both in terms of lead-time and period of initialization. Later we assess whether these skillful predictions correspond to high-impact events, especially cold spells. Finally, in the attempt to identify the potential drivers of improved predictability, we track back to the dominant Euro-Atlantic modes of variability present in the initial atmospheric states of the well predicted events.

How to cite: Mastrangelo, D., Giuntoli, I., and Malguzzi, P.: Exploring winter predictability in Europe using the ECMWF hindcasts, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9538, https://doi.org/10.5194/egusphere-egu21-9538, 2021.

The skill of predicted wind speed at 100 m and temperature at 2 m has been assessed in extended-range forecasts and hindcasts of the European Center for Medium-Range Weather Forecasts, starting from December 2015 to November 2019. The assessment was carried out over Europe grid-point wise and also by considering several spatially averaged country-sized domains, using standard scores such as the Continuous Ranked Probability Score and Anomaly Correlation Coefficient. The (re-)forecasts showed skill over climatology in predicting weekly mean wind speed and temperature well beyond two weeks. Even at a lead time of 6 weeks, the probability of the (re-)forecasts being skillful is greater than 50%, encouraging the use of operational subseasonal forecasts in the decision making value chain. The analysis also exhibited​ significant differences in skill in the predictability of different variables, with temperature being more skillful than wind speed, and for different seasons, with winter allowing more skillful forecasts. The predictability also displayed a clear spatial pattern with forecasts for temperature having more skill for Eastern than for Western Europe, and wind speed forecasts having more skill in Northern than Southern Europe.

How to cite: Goutham, N., Plougonven, R., Omrani, H., Parey, S., Tankov, P., Tantet, A., and Drobinski, P.: Verification of the European subseasonal forecasts of wind speed and temperature, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-14728, https://doi.org/10.5194/egusphere-egu21-14728, 2021.

How to cite: Newman, M. and Albers, J.: Subseasonal Predictability of the North Atlantic Oscillation, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1427, https://doi.org/10.5194/egusphere-egu21-1427, 2021.

Midlatitude prediction on subseasonal timescales is difficult due to the chaotic nature of the atmosphere and often requires the identification of favorable atmospheric conditions that may lead to enhanced skill ("forecasts of opportunity"). Here, we demonstrate that an artificial neural network can identify such opportunities for tropical-extratropical teleconnections to the North Atlantic circulation at a lead of 22 days using the network's confidence in a given prediction. Furthermore, layer-wise relevance propagation, an ANN interpretability technique, pinpoints the relevant tropical features the ANN uses to make accurate predictions. We find that layer-wise relevance propagation identifies tropical hot spots that correspond to known favorable regions for midlatitude teleconnections and reveals a potential new pattern for prediction over the North Atlantic on subseasonal timescales.

How to cite: Mayer, K. and Barnes, E.: Subseasonal Forecasts of Opportunity Identified by an Explainable Neural Network, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-9015, https://doi.org/10.5194/egusphere-egu21-9015, 2021.

Although there is an increasing interest in precipitation information at the subseasonal timescales in a wide range of sectors, the use of subseasonal precipitation forecasts from general circulation models is often impaired by poor reliability and low forecast skill. One crucial step to improve forecast quality is statistical correction and post-processing, which is particularly important for a parameterized variable like rainfall. This study introduces and evaluates a statistical-dynamical post-processing scheme, based on a Bayesian framework, that aims at providing more skillful and more reliable subseasonal forecasts of weekly precipitation. On the one hand, this method relies on the statistical relationship between observed and dynamically-forecast precipitation, that is determined in a set of reforecasts and depends on the lead time. On the other hand, it also takes advantage of the climatological impacts of large-scale drivers affecting rainfall, that are generally better represented by numerical models than rainfall itself. These two aspects of the method are respectively called calibration and bridging.

This statistical-dynamical prediction scheme is illustrated with an application to the austral summer precipitation in the southwest tropical Pacific, using the Météo-France and ECMWF reforecasts in the Subseasonal-to-seasonal (S2S) database. Indices representing El Niño Southern Oscillation and the Madden-Julian Oscillation – the major sources of predictability in the area – are used for bridging. Probabilistic forecasts of heavy rainfall spells are evaluated in terms of discrimination (ROC skill score) and reliability, which are both improved by the Bayesian method at all lead times (from week 1 to week 4). Additional results show that the calibration part of the method, using forecast precipitation as a predictor, is necessary to enhance forecast skill. The bridging part also provides additional discrimination skill, that is mostly due to the ENSO-related information.

How to cite: Specq, D. and Batté, L.: A statistical-dynamical approach to improve subseasonal precipitation forecasts: application to the southwest tropical Pacific, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-1292, https://doi.org/10.5194/egusphere-egu21-1292, 2021.

Besides the ocean, the land surface is a crucial component for predictability at (sub-)seasonal time scales. While the prediction of 2m temperature up to several months is possible for some maritime regions, continental regions lack predictive skill. Improved representation of the land surface in seasonal forecasting systems could help to close this gap. Snow cover fraction and snow water equivalent (SWE) are essential properties of the land surface. A snow-covered land surface leads to local temperature decreases in the overlying air (snow-albedo effect and high emissivity) and melting snow cools the surface air and contributes to soil moisture. First, we analyse the dynamical relationships between snow, 2m temperature and sensible/latent heat fluxes in reanalysis data in the northern hemisphere. Then we investigate whether these relationships are also present in operational seasonal forecast models provided by Copernicus Climate Change Service (C3S). First results show that the quality of the 2m temperature forecast over continental regions drops sharply after the first forecasted month, whereas anomalies in snow water equivalent can be predicted up to several months. Forecasted anomalies in sensible and latent heat fluxes of continental land surfaces show predictive skill during winter and spring only locally in some places, which reduces potential interactions between snow/land surface and the atmosphere in the models. The goal of this ongoing work is to assess the importance of snow initialisation and parameterisation for seasonal forecasting.

How to cite: Risto, D., Ahrens, B., and Fröhlich, K.: Snow as a source of predictability in seasonal forecasts?, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-7331, https://doi.org/10.5194/egusphere-egu21-7331, 2021.

The quality of weather forecasts is continuously improving for decades. However, increases in forecast skills have slowed down in recent years. This highlights the importance of exploring new avenues towards future forecast system improvements. Until now, (near) real-time information on vegetation anomalies is not used in most forecasting models. Addressing this gap, we explore the potential of the vegetation state for explaining the spatial and temporal variation in forecast accuracy globally across climate regions, seasons, and vegetation types. For this purpose, we employ re-forecasts from the European Centre of Medium-Range Weather Forecasting (ECMWF) and infer the vegetation status through the Enhanced Vegetation Index derived from the Moderate-Resolution Imaging Spectroradiometer (MODIS) satellite observations during the 2000-2019 period. In particular, we focus on land surface variables such as evaporation and temperature to study the relationship between forecast errors and vegetation anomalies.

The results show a stronger correlation between forecast errors and vegetation anomalies in semi-arid and sub-humid regions during the growing season, which highlights that vegetation information has the potential to help advance weather forecast performance. To put these results into perspective, we will further perform a multivariate analysis to determine the relative roles of vegetation, hydrology and climate in explaining weather forecast errors. Thereby, our results can inform the future development of weather forecast models and underlying data assimilation schemes.

How to cite: Ruiz, M., Oh, S., Orth, R., and Balsamo, G.: Exploring the potential of vegetation information for improving weather forecast performance, EGU General Assembly 2021, online, 19–30 Apr 2021, EGU21-3105, https://doi.org/10.5194/egusphere-egu21-3105, 2021.