Weather and climate modelling are at a transformative moment, driven by rapid advances in machine learning (ML) and artificial intelligence (AI), access to extreme-scale supercomputing, and the development of very high-resolution Earth system models. These advances offer new opportunities to improve predictions of extreme events, explore future climate scenarios, and provide more tailored information to users. At the same time, they bring new scientific, technical, and operational challenges — from model development and verification to efficient deployment on emerging computing architectures.

The European Commission’s Destination Earth (DestinE) initiative addresses some of these challenges and opportunities through a large-scale collaborative European effort. DestinE is implemented by ECMWF, ESA and EUMETSAT, together with many partners across Europe, under the lead of EC's DG CNECT. DestinE is building Digital Twins of the Earth system — interactive, high-resolution simulations that combine advanced modelling, Earth observations, and emerging AI methods, leveraging the EuroHPC world class supercomputers. These Digital Twins enable the exploration of “what-if” scenarios and provide detailed information to support adaptation and response to extreme weather and climate events.

In this contribution, we will present key experiences from DestinE, focusing on successes and challenges in building the Digital Twin Engine, the first two Digital Twins on weather-induced extremes and climate change adaptation, as well as efforts to develop an ML-based Earth system model. We will discuss how these developments, led by ECMWF and over 100 partner institutions across Europe, push the boundaries of modelling capabilities, while explicitly aiming to complement existing national and European capabilities and services.

An important aspect of DestinE is its effort in bringing weather and climate modelling closer together. Many of the scientific and technical challenges — related to high-resolution modelling, verification, exploitation of Europe's largest supercomputers and AI integration — are shared across timescales. DestinE provides a framework to address these challenges in a coordinated way, helping to advance Europe’s capability to respond and adapt to environmental extremes, and supporting national authorities in their role to protect lives and assets from extreme events and climate change.

How to cite: Sandu, I., Wedi, N., and Pappenberger, F.: Opportunities and Challenges in Weather and Climate Modelling: Experiences from the Destination Earth Initiative, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-280, https://doi.org/10.5194/ems2025-280, 2025.

In this presentation, we introduce a new limited-area model (LAM) that employs the spectral element method (SEM) on a cubed-sphere grid. This is the first SEM-based LAM developed for dynamical downscaling applications. As the limited-area version of the Korean Integrated Model (KIM)—Korea’s operational global weather forecasting model since 2020—the LAM shares core components with the global KIM, including the non-hydrostatic dynamical core and scale-aware physics, differing only in the simulation domain.

At the previous European Meteorological Society Annual Meeting, we demonstrated that our lateral boundary condition design for SEM maintains numerical stability beyond 10 days of integration in idealized test runs nested within global KIM simulations. Since then, additional experiments have confirmed that the LAM can stably perform 1 km resolution simulations over the Korean Peninsula using 6-hourly lateral boundary conditions derived from 8 km resolution global forecasts. The model has also demonstrated stable performance when driven by ERA5 reanalysis data from the European Centre for Medium-Range Weather Forecasts (ECMWF), underscoring its potential as a general-purpose downscaling tool. These results suggest that an SEM-based LAM can serve as an efficient and robust dynamical downscaling system, especially given that our simulations utilized minimal linear relaxation at the lateral boundaries and no spectral nudging.

With the ultimate goal of supporting high-resolution short-range forecasting, we have also developed a test operational scenario aligned with the operational forecast schedule of the global KIM model which will provide the initial and boundary conditions for the LAM. Additionally, we present the ongoing development of a dedicated data assimilation system for the LAM, designed to directly assimilate reflectivity and radial velocity observations from Korea’s ground-based radar network.

How to cite: Cho, H., Kim, J., Noh, I., and Jang, J.: Toward Operational High-Resolution Forecasting: First Application of a Spectral Element Limited-Area Model, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-346, https://doi.org/10.5194/ems2025-346, 2025.

The operational Canadian Global Deterministic Prediction System suffers from a weak-intensity bias for simulated tropical cyclones.  The presence of this bias is confirmed in progressively simplified experiments using a hierarchical system development technique.  Within a semi-idealized, simplified-physics framework, an unexpected insensitivity to the representation of relevant physical processes leads to investigation of the model's semi-Lagrangian dynamical core.  The root cause of the weak-intensity bias is identified as excessive numerical dissipation caused by substantial off-centering in the two time-level time integration scheme used to solve the governing equations.  Any (semi-)implicit semi-Lagrangian model that employs such off-centering to enhance numerical stability will be afflicted by a  misalignment of the pressure gradient force in strong vortices.  Although the associated drag is maximized in the tropical cyclone eyewall, the impact on storm intensity can be mitigated through an intercomparison-constrained adjustment of the model's temporal discretization. The revised configuration is more sensitive to changes in physical parameterizations and simulated tropical cyclone intensities are improved at each step of increasing experimental complexity. Although some rebalancing of the operational system may be required to adapt to the increased effective resolution, significant reduction of the weak-intensity bias will improve the quality of Canadian guidance for global tropical cyclone forecasting.

How to cite: McTaggart-Cowan, R., Nolan, D., Aider, R., Charron, M., Chen, J.-H., Cossette, J.-F., Gaudreault, S., Husain, S., Magnusson, L., Qaddouri, A., Separovic, L., Subich, C., and Yang, J.: Reducing a Tropical Cyclone Weak-Intensity Bias in a Global Numerical Weather Prediction System, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-52, https://doi.org/10.5194/ems2025-52, 2025.

The issue of poles has been impeding the increase of horizontal resolution of global atmospheric models on the latitude-longitude grid. This is due to the strict limitation on the time step size. Although implicit or semi-implicit methods have been designed to enhance the numerical stability, their algorithms for solving the large matrix systems are very complex and not suitable for massive parallel computers. This study aims to address this problem by developing an effective Gaussian filter. The 1D filter is applied in the zonal direction, with its kernel width decreasing smoothly according to the zonal Courant-Friedrichs-Lewy (CFL) condition from the pole to the equator. By filtering the dynamic tendencies, the zonal spatial scale is enlarged to boost the numerical stability in the polar region. Parallelization is optimized through setting different zonal halo widths in different processes and using asynchronized MPI communication. The new dynamical core solves the nonhydrostatic equations under the terrain following mass coordinate, with the horizontal momentum equations in vector-invariant form. The variables are staggered on the C-grid for discretization. Compared with other dynamical cores, the implementation here is simpler and more user-friendly, which includes the shallow water and baroclinic versions in one code base. Several standard test cases are employed to verify the efficacy of this new dynamical core, including shallow water cases, hydrostatic/nonhydrostatic baroclinic cases. The results demonstrate that this method can effectively mitigate the pole problem, allowing for a larger explicit time step size comparable with quasi-uniform cores while preserving numerical accuracy as much as possible.

How to cite: Dong, L.: Development of a new dynamical core on the latitude-longitude grid targeting high resolution simulations, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-694, https://doi.org/10.5194/ems2025-694, 2025.

High-impact weather caused by summertime convection poses a considerable threat to people and property. Southwest Germany is particularly frequently affected by hailstorms and associated damage, with an anomalously high number of hail days per year. Yet, forecasting convective events remains one of the greatest challenges in numerical weather prediction (NWP), even for regional high-resolution convective-scale NWP systems. Here, we test to what extent the assimilation of comprehensive high-resolution observations of the lower troposphere from a field campaign improves initial conditions and the quality of subsequent forecasts of convective events.

From June to August 2023, the ‘Swabian MOSES 2023’ campaign took place in the southwest German mountain ranges. During the campaign, a wide range of instruments to observe the dynamical and thermodynamical characteristics of the lower troposphere, distributed across an area of roughly 100 km x 100 km, was deployed. In this contribution, we will present a campaign re-analysis dataset that utilizes 3 months of ground-based remote sensing and in-situ campaign observations in addition to observations from the operational observation network. The observations are assimilated hourly in the regional forecasting system of the DWD, which employs the non-hydrostatic model ICON at 2.2 km grid spacing (ICON-D2) and the Kilometer Scale Ensemble Data Assimilation system (KENDA) with 40 ensemble members. In addition to the operationally available observations we assimilate 1) vertical profiles of the horizontal wind retrieved from an unprecedented network of 12 Doppler wind lidars (DWL), 2) reflectivity from an X-Band radar, 3) radiosoundings released at 2 sites during intensive observation periods, 4) ground-based zenith path-delay observations from a (not yet operationally assimilated) German-wide network of Global Navigation Satellite Systems receivers, and 5) 2-meter temperature and relative humidity observations from six campaign surface stations.

We will present the setup of the assimilation experiments and quantify which and where new information is added to the analysis. We specifically focus on the assimilation of DWL-retrieved wind profiles and their influence on the mesoscale circulation in southwest Germany. For example, we find that the mean absolute observation-to-background difference is slightly larger for the wind profiles retrieved from DWLs than for wind profiles retrieved from operational radar wind profilers. Yet, the observation-to-analysis difference is similar for all wind profiler data. In addition, we compare the 4D campaign re-analysis dataset with a quasi-operational control re-analysis that excludes the campaign observations. The comparison indicates systematic wind direction differences in southwest Germany resulting from the assimilation of campaign observations. These differences are propagated in re-forecasts, which we will illustrate exemplarily for (i) one week of deterministic re-forecasts of dry and convective days, and (ii) for a 20-member ensemble re-forecast of a small-scale high-impact convective event.

How to cite: Thomas, J., Reich, H., Geppert, G., Stephan, K., Steinert, T., Keller, J., Knippertz, P., and Oertel, A.: The ‘Swabian MOSES 2023’ re-analysis: how do high-resolution campaign observations change the analysis state of a convective-scale NWP system?, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-566, https://doi.org/10.5194/ems2025-566, 2025.

Verification of global high-resolution precipitation forecasts is challenging. Spatial verification techniques address some shortcomings of traditional verification. However most existing methods do not account for the non-planar geometry of a global domain, or their computational complexity is too large for global assessments. We present an adaptation of the recently developed Precipitation Attribution Distance (PAD) metric, designed for verifying precipitation, enabling its use on the Earth's spherical geometry. PAD estimates the magnitude of location errors in the forecasts employing the mathematical theory of Optimal Transport, as it provides a close upper bound for the Wasserstein distance. The method is fast and flexible with time complexity O(n log(n)). Its behavior is analyzed using idealized cases and 7 years of operational global deterministic 6-hourly precipitation forecasts from the Integrated Forecasting System (IFS) of the European Centre for Medium-Range Weather Forecasts. Due to the lack of availability of high-resolution, high-quality global precipitation observations, we employed the short-term forecasts by the same model as verification truth. The summary results for the whole period show how location errors in the IFS model grow steadily with increasing lead time for all analyzed regions. Moreover, by examining the time-series of the results, we can determine the trends in the score's value and identify the regions where there is a statistically significant improvement (or worsening) of the forecast performance. The results can also be analyzed separately for different intensities of precipitation. Overall, the PAD provides meaningful results for estimating location errors in global high-resolution precipitation forecasts at an affordable computational cost.

How to cite: Skok, G. and Lledó, L.: Spatial verification of global precipitation forecasts, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-42, https://doi.org/10.5194/ems2025-42, 2025.

Forecast verification is a well-established activity essential for climate prediction providers to assess and communicate the capabilities and limitations of their forecasts. It is also crucial for informing users about the reliability and usefulness of the climate information they access. However, the verification process involves several methodological choices—such as the selection of reference datasets, spatial transformations, and statistical metrics—which can significantly affect the outcome.

In this study, we use scorecards to summarise and evaluate the impact of key decisions made during forecast verification. Scorecards is a visual synthesis tool we have developed to summarise the performance of seasonal forecast systems across multiple initialisation dates, forecast times, and statistical metrics. Presented in a tabular format, scorecards enable a compact and flexible overview of verification results for user-defined variables, regions, and climate indices, helping to identify consistent patterns without replacing detailed spatial maps. Specifically, we quantify the effects of: (i) regridding forecasts to the observational reference grid versus regridding observations to the model grid, (ii) applying orographic corrections to air temperature when model and observational grids differ, (iii) the timing and implementation of cross-validated anomaly computation, and (iv) the evaluation of forecasts against blended observational products combining air and sea surface temperatures.

We also examine the influence of different approaches to spatially aggregating verification metrics and propose methodologies for assessing the statistical significance of these aggregated scores. By systematically quantifying these methodological impacts, we provide practical recommendations for improving forecast verification practices.

This framework is being tested within the context of the CERISE project, which aims to advance the next generation of C3S seasonal prediction systems through improved land-atmosphere data assimilation and land surface initialisation techniques. Since land surface initial conditions can substantially influence near-surface climate predictions for several months, especially for variables relevant to heatwaves, droughts, and water availability, robust verification is critical. The proposed framework will support fair and consistent comparisons of evolving seasonal forecast systems developed within CERISE.

How to cite: Pérez-Zanón, N., Milders, N., Delgado-Torres, C., Agudetse, V., G. Muñoz, Á., and Doblas-Reyes, F.: Comprehensive Assessment of Seasonal Forecasts Across Multiple Models and Model Versions, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-194, https://doi.org/10.5194/ems2025-194, 2025.