arXiv subjects

Daniela I. V. Domeisen

Publications and source records attributed to Daniela I. V. Domeisen.

3 recordsLinked to original sources

arXiv2026

Stress-Testing Dynamical and Generative Downscaling Using Subseasonal Extreme Precipitation Forecasts

Coarse spatial resolution limits the ability of subseasonal prediction models to resolve extreme precipitation. Downscaling with either dynamical or deep generative models can overcome this issue, but the comparative performance of these models for extremes across different atmospheric regimes remains poorly understood. In this work, we evaluate the Weather Research and Forecasting (WRF) model against a diffusion-based generative model by downscaling two physically distinct, extreme precipitation events up to lead times of 3 weeks. For a fair comparison with WRF, which can downscale boundary conditions from different driving models without model-specific training, the diffusion model is trained in an unpaired fashion. Both approaches improve upon the raw European Centre for Medium-Range Weather Forecasts forecasts, in comparison to fused rain gauge-radar observations in Switzerland (CombiPrecip), but exhibit regime-dependent strengths. WRF achieves the highest probabilistic skill for a multicell, non-stationary event. Conversely, the diffusion model is more consistent across different performance metrics for the two events, outperforming WRF in a more stationary supercell event. These results demonstrate that explicit dynamical modeling can add value for specific precipitation events for subseasonal lead times, and that generative downscaling adds value more broadly in different situations.

physics.comp-ph↗

arXiv2025

Do AI models predict storm impacts as accurately as physics-based models? A case study of the February 2020 storm series over the North Atlantic

The emergence of data-driven weather forecast models provides great promise for producing faster, computationally cheaper weather forecasts, compared to physics-based numerical models. However, while the performance of artificial intelligence (AI) models have been evaluated primarily for average conditions and single extreme weather events, less is known about their capability to capture sequences of extreme events, states that are usually accompanied by multiple hazards. The storm series in February 2020 provides a prime example to evaluate the performance of AI models for storm impacts. This event was associated with high surface impacts including intense surface wind speeds and heavy precipitation, amplified regionally due to the close succession of three extratropical storms. In this study, we compare the performance of data-driven models to physics-based models in forecasting the February 2020 storm series over the United Kingdom. We show that on weekly timescales, AI models tend to outperform the numerical model in predicting mean sea level pressure (MSLP), and, to a lesser extent, surface winds. Nevertheless, certain ensemble members within the physics-based forecast system can perform as well as, or occasionally outperform, the AI models. Moreover, weaker error correlations between atmospheric variables suggest that AI models may overlook physical constraints. This analysis helps to identify gaps and limitations in the ability of data-driven models to be used for impact warnings, and emphasizes the need to integrate such models with physics-based approaches for reliable impact forecasting.

physics.ao-ph↗

arXiv2016

Nonlinear stratospheric variability: multifractal detrended fluctuation analysis and singularity spectra

Characterising the stratosphere as a turbulent system, temporal fluctuations often show different correlations for different time scales as well as intermittent behaviour that cannot be captured by a single scaling exponent. In this study, the different scaling laws in the long term stratospheric variability are studied using Multifractal de-trended Fluctuation Analysis. The analysis is performed comparing four re-analysis products and different realisations of an idealised numerical model, isolating the role of topographic forcing and seasonal variability, as well as the absence of climate teleconnections and small-scale forcing. The Northern Hemisphere (NH) shows a transition of scaling exponents for time scales shorter than about one year, for which the variability is multifractal and scales in time with a power law corresponding to a red spectrum, to longer time scales, for which the variability is monofractal and scales in time with a power law corresponding to white noise. Southern Hemisphere (SH) variability also shows a transition at annual scales. The SH also shows a narrower dynamical range in multifractality than the NH, as seen in the generalised Hurst exponent and in the singularity spectra. The numerical integrations show that the models are able to reproduce the low-frequency variability but are not able to fully capture the shorter term variability of the stratosphere.

physics.ao-ph↗