GeoSphere Austria: A more powerful forecasting system for the Alps-Adriatic region
The C-LAEF AlpeAdria is a modernised forecasting system that can predict local weather phenomena in the Alpine region in greater detail and with higher quality. It provides forecasts updated eight times a day, including uncertainty estimates for the next 60 hours on a one-kilometre grid.
The new C-LAEF AlpeAdria forecasting system has recently come into operation and is now also available to the public as open data via the GeoSphere Austria Data Hub. C-LAEF AlpeAdria, which stands for Co-Operated Limited Area Ensemble Forecasting for the AlpeAdria region, was developed in collaboration with the partner meteorological services ARSO (Slovenia) and DHMZ (Croatia) and is operated jointly. ‘GeoSphere Austria has a long tradition of cooperating with European meteorological services in the field of numerical weather prediction. However, the joint development and, in particular, the joint operation of the forecasting system with ARSO and DHMZ represent a milestone for us in terms of international cooperation,’ says Christoph Wittmann, Head of the Analysis and Modelling Department at GeoSphere Austria.
Following its operationalisation, C-LAEF AlpeAdria is now being gradually integrated into the various services and products offered by GeoSphere Austria. Its first application over the summer was to support meteorologists in producing forecasts and in their warning and information activities. At the same time, the data was prepared for the GeoSphere Austria Data Hub, where the latest forecast data is now available as open data directly via programming interfaces and as a bulk download.
C-LAEF AlpeAdria is based on a comprehensively enhanced version of the AROME weather model and runs on a significantly finer grid than previous systems (one kilometre instead of the previous 2.5 kilometres). This allows weather phenomena in the Alpine region to be simulated in greater detail and with greater precision. Furthermore, C-LAEF AlpeAdria enables a significantly improved quantification of forecast uncertainty (eight ensemble runs per day instead of the previous two). ‘The launch of C-LAEF AlpeAdria by no means marks the end of development,’ emphasises Wittmann. Further development steps are planned. ‘In future, there will be a particular focus on linking the system with GeoSphere Austria’s AI-based model, which is currently in trial operation, in order to optimally combine the strengths of classical numerical models and modern AI approaches.’
The higher spatial resolution and more frequent ensemble runs lead to significant improvements in forecast quality – particularly for local weather phenomena. The system has been developed and tested extensively over several years. Extensive evaluations clearly demonstrate the added value compared with previous systems. This primarily concerns the forecasting of wind, precipitation (particularly convective precipitation, i.e. showers and thunderstorms), fog and high fog (especially in Alpine valleys and basins), precipitation type, and near-surface temperature and humidity.
Translated with DeepL.com (free version)