Forecasting convective storms remains a significant challenge in Numerical Weather Prediction (NWP). Data Assimilation (DA) plays a crucial role by improving the initial conditions for forecasts through the integration of observational data with previous model outputs (background). Among observational platforms, weather radar is particularly valuable due to its high spatial and temporal resolution, offering detailed information for storm monitoring. Therefore, assimilating radar data into Numerical Weather Prediction models has the potential to substantially enhance the accuracy of storm forecasts. However, studies, such as Fabry and Meunier (2020), have shown that short-term precipitation forecasts produced through radar-based extrapolation methods (nowcasting) often outperform model-based forecasts with Data Assimilation. This is largely because radar primarily provides information on precipitation structures and intensities within the storm-affected region but lacks direct insights into broader environmental conditions like temperature, wind fields, and humidity. These atmospheric variables, both within and outside the storm system, are critical for accurate storm evolution forecasts. A promising approach to address this limitation is the application of machine learning (ML) to develop a more advanced observation operator for Data Assimilation. Specifically, an encoder-decoder neural network can be trained to relate Numerical Weather Prediction model variables (such as temperature, wind components, and relative humidity) to corresponding radar reflectivity observations. This ML-based observation operator captures complex non-linear relationships between model variables and radar data observations, while its Jacobian can propagate radar-derived information to other atmospheric variables in the Data Assimilation system, potentially leading to improved representation of storm dynamics and enhancing convective storm forecasting capabilities.

How to cite: Stefanelli, M., Zaplotnik, Z., and Skok, G.: A neural network-based observation operator for weather radar data assimilation, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-28, https://doi.org/10.5194/ems2025-28, 2025.

How to cite: Melinc, B., Perkan, U., and Zaplotnik, Ž.: Exploring a neural network-based background-error covariance model , EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-109, https://doi.org/10.5194/ems2025-109, 2025.

Enhancing the spatial resolution of global analyses and forecasts enables weather prediction systems to more accurately capture rapidly evolving mesoscale and convective-scale atmospheric phenomena. The Integrated Forecast System (IFS) of ECMWF consists of an Earth-system model coupled with an incremental 4D-Var data assimilation (DA) system, which has recently been experimentally upgraded to higher spatial resolution.

This upgrade involved higher resolution (4.4 km) 4D-Var trajectories and higher resolution (20 km) 4D-Var minimizations using the tangent linear model (TLM) and its adjoint (ADM). This configuration requires less observation thinning and resulted in greater use of high-resolution observations, such as those from multispectral imagers aboard Himawari and Meteosat satellites. Shorter time steps in TLM and ADM further enabled the use of shorter observation time slots (400 seconds), allowing for more accurate comparison between observations and their model equivalents.

We demonstrate that this setup leads to significant improvements in the accuracy of initial conditions and enhances tropospheric forecast skill, extending medium-range predictability by 6 to 12 hours. We argue that the improved forecast skill can be attributed to a better representation of large-scale features in the stratosphere. Higher resolution DA and forecasts also lead to improved forecasting of extreme precipitation events, as well as the track and intensity of tropical cyclones. Furthermore, the improved initial conditions significantly increase the predictability machine-learning-based forecasts.

The benefits of the higher resolution DA system are demonstrated using the case of Tropical Cyclone Otis, which made landfall as a Category 5 cyclone, but was predicted by most global and regional NWP models to peak only at tropical storm intensity – a major forecasting bust. By employing higher resolution 4D-Var DA, we were able to replicate the observed rapid intensification of TC Otis. 

How to cite: Zaplotnik, Z., Schroettle, J., Bandeiras, J., Vanniere, B., Orlandi, E., Maier-Gerber, M., Holm, E., and Bonavita, M.: Towards higher resolution data assimilation in ECMWF IFS, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-691, https://doi.org/10.5194/ems2025-691, 2025.

The ICREN and NEW-ARGENT projects aim at studying the impact of GNSS data assimilation on the precipitation forecast over Italy. Specifically, the assimilation of GNSS-ZTD together with lightning was studied in the ICREN project, while NEW-ARGENT considered the assimilation of GNSS along slant paths. 

The approach used in both projects is the VSF (Very Short-term Forecast) in which a 6h data assimilation phase is followed by a 6h forecast. In addition, both projects used the WRF model and the 3DVar data assimilation system of CNR-ISAC (Federico, 2013; Torcasio et al., 2024). 

The ICREN project focuses over northern Italy, where intense convective events occur since late spring to early fall. More than 100 case studies were considered and GNSS-ZTD was assimilated with and without lightning data. Different strategy of simulations were considered and results show the important role of the VSF forecast and data assimilation in predicting convective rainfall. Specifically, the assimilation of lightning was very useful for the first hours of forecast improving substantially the prediction of intense precipitation (higher than 30 mm/3h). In any case, assimilating lightning was useful also at lower precipitation thresholds.  The GNSS-ZTD data assimilation had a lower impact on the precipitation forecast, nevertheless it improved the forecast for all precipitation thresholds. However, the combination of both data had the largest impact on the precipitation forecast, especially for the forecast period from 1h to 4h after the ending of the assimilation phase. Finally, the ICREN project showed a significant impact of the GNSS-ZTD and lightning data assimilation up to 6h forecast.

The NEW-ARGENT project considered four months (May, June and October 2023, and September 2022) and focused on the assimilation of GNSS along slant paths for three regions: Lazio, Lombardy and Sicily. This problem was firstly solved assimilating GNSS gradient that partially recover the anisotropy of GNSS observations. Specifically, we compared the precipitation forecast at the short-range in four different experiments set-up: CTRL (control), without GNSS data assimilation, GNSS-ZTD, with the assimilation of GNSS zenith delay, GNSS-GRA, in which the gradients are assimilated, and GNSS-ZTD-GRA, in which both the gradients and the zenith total delay are assimilated.Results show that the assimilation of the gradients, both alone and with the GNSS-ZTD, is beneficial for the improvement of precipitation forecast of convective events over Italy.

References

Federico, S.: doi:10.5194/amt-6-3563-2013, 2013.

Torcasio, R.C.; et al.: https://doi.org/10.3390/rs16101769, 2023.

Acknowledgments 

This work has been realized in the projects PRIN-PNRR NEW-ARGENT (MUR contract- P20228LMA2) and ICREN-PRIN project (MUR- CUP: D53D23004770006). 

How to cite: Federico, S., Realini, E., Torcasio, R. C., Transerici, C., Venuti, G., Song, X., and Crespi, M.: Assimilation of GNSS data over Italy: the results of ICREN and NEW-ARGENT projects, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-520, https://doi.org/10.5194/ems2025-520, 2025.

Within the context of the “Global-to-Regional ICON digital twin” (GLORI) project, a convection-permitting ensemble forecasting is established in order to study the predictability of high-impact weather events with high-resolution modeling (up to 500 m) and the influence of the land-surface—atmosphere coupling mechanisms.

At the DWD, ICON-D2-EPS is the limited-area high-resolution component of the ICON modeling system, running as an ensemble of 20 members at 2 km horizontal resolution over Germany and surrounding areas. The perturbed initial conditions are provided by the km-scale ensemble data assimilation system KENDA, run at the same resolution, assimilating a wide range of observations, including radar-derived radar volumes. Boundary conditions are provided by ICON-EPS, the global ensemble with a refinement at 13 km over Europe and is refreshed every 3 hours.

In this work, we employ a nested domain with horizontal resolution of 1-km in the southern region of the ICON-D2, encompassing the Alps mountains. We run a 24-hour forecast simulation starting at 00UTC on the 21^st of June 2022 and the 20^th of June 2024, with 20 ensemble members. The choice of the date are crucial as they correspond to a day when the DWD recorded instances of heavy rain and hail in southern Germany.

In our study, we perform all experiments using a two-moment microphysics scheme. Additionally, we incorporate the standard operational model perturbations and subsequently analyze the influence of various convection schemes on the predictability of processes that lead to convection development. Specifically, we examine the behavior of the convection scheme in two configurations: shallow convection only and deep convection parameterization in the so-called gray-zone-tuning version. By selectively enabling and disabling these schemes, our goal is to evaluate their individual contributions to predictability.

Following this, we implement a tailored variant of the stochastically perturbed parameterization scheme (SPP) in ICON in order to delve into the influence of some uncertain parameters within either microphysics, convection or turbulent parameterization, further advancing our understanding of its effects on the model performance.

How to cite: Parsakhoo, Z., Marsigli, C., Gebhardt, C., Seifert, A., Reinert, D., Steinert, T., and Keller, J.: Studies of Convection-Permitting Ensemble Forecasting for ICON-D2 with a 1km Nest over the Alps, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-302, https://doi.org/10.5194/ems2025-302, 2025.

In recent years, data-driven weather models have advanced rapidly, leveraging deep learning and vast reanalysis datasets to generate high-resolution forecasts with remarkable speed. Models such as Pangu-Weather, GraphCast and FourCastNet have demonstrated competitive or even superior accuracy compared to traditional numerical weather prediction (NWP) models. These AI-based models eliminate the need for explicit physical equations, relying instead on learned patterns from historical data. However, challenges remain, including limited physical interpretability, difficulty in handling extreme events, and the need for robust uncertainty quantification. Despite these challenges, the rapid progress of data-driven weather models suggests they will play an increasingly important role in future forecasting systems, potentially complementing or even transforming traditional NWP methods.

Pangu-Weather comprises four AI-driven models, each optimized for different forecast lead times (1-hour, 3-hour, 6-hour, and 24-hour). These models exhibit distinct error growth rates and capture different weather phenomena, ranging from rapidly evolving atmospheric features to large-scale global weather patterns. To leverage these characteristics, this study explores a stochastic combination approach for generating both initial perturbations and model perturbations:

• Initial perturbations are created by randomly combining differences between the 48-hour and 24-hour forecasts from the 1-hour, 3-hour, 6-hour and 24-hour models to form 15 pairs (30 members).
• Model perturbations are introduced through the stochastic combination of forecasts from the 3-hour, 6-hour and 24-hour models from 0-120 hours.

This multi-model ensemble strategy with 30 members enhances the representation of forecast uncertainties for extratropical areas of synoptic scale phenomena by incorporating the strengths of each model while mitigating their individual weaknesses. As a result, it provides a more robust and probabilistic forecasting framework with limited costs, improving forecast skills for extended range of atmospheric prediction. However, the data driven model is biased which could reduce forecast reliability; and it is still challenged for tropical uncertainty representation due to limited capability to assimilate small scale error growth.

How to cite: Tang, J., Zhu, Y., and Dai, K.: Fully Data Driven “Multi-Model” Ensemble, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-47, https://doi.org/10.5194/ems2025-47, 2025.

Modern weather forecasting is moving towards ensemble prediction by default alongside the increased use of Machine Learning (ML) models and ensembles over a range of resolutions. From the user perspective, this will mean having to deal with a very large number of forecasts from different sources, often with different biases. There could potentially be many hundreds of ensemble members made available, with increased time frequency. At the European Centre for Medium Range Weather Forecasts (ECMWF) there are already 152 members available if the medium-range and sub-seasonal ensembles are used together out to 15 days, and with the addition of a ML ensemble the number increases to over 200.

For downstream applications that have limited resources, the problem becomes one of picking out the most salient and representative information from all those forecasts to extract the most important forecast messages. There is likely to be an increasing need for forms of “storylines” or “representative forecasts” that provide context, enable simpler communication and can be used effectively in downstream models. Clearly, there is still a vital place for probabilistic outputs, but on their own they limit the full extent of what could be usefully obtained.

Here we present highlights from work that investigates combining ensembles, member extraction and verification, making use of the different ensembles available at ECMWF with a focus on the prediction of synoptic weather patterns and regimes over a range of scales. The findings make use of a new diagnostic called CURV that categorises the degree of cyclonic or anticyclonic curvature from Mean Sea Level Pressure or Geopotential Height fields.

How to cite: Roberts, N., Hewson, T., and Brookshaw, A.: Obtaining salient information from many forecasts, EMS Annual Meeting 2025, Ljubljana, Slovenia, 7–12 Sep 2025, EMS2025-510, https://doi.org/10.5194/ems2025-510, 2025.