Integration of multi-sensor nowcasting systems to improve the early detection of severe convective storms in the Alpine area

In case of severe thunderstorms MeteoSwiss alerts authorities and population by means of flash-news warnings with a lead-time of 30-60 min. For these very short lead times, in addition to usual short-range forecasting tools, the two specific real-time, multi-sensor nowcasting systems TRT (Thunderstorms Radar Tracking) and COALITION (Context and Scale Oriented Thunderstorm Satellite Predictors Development) are used operationally with a refresh time of 5 min. TRT is a mainly radar-based system for the automatic detection, tracking, analysis and trajectory extrapolation of intense convective cells. It also runs a heuristic cell severity ranking algorithm that integrates the most significant severity attributes from the 3D storm structure in a single numerical parameter in order to assess the potential dangerousness of individual cells. COALITION is a newly developed system that combines the best available information on convective processes provided by different data sources. This algorithm integrates evolving thunderstorm attributes (predictants, e.g. Cloud Top Temperature, Radar based VIL,...) with severe convection predictors (environmental parameters, e.g. Meteosat Second Generation Convection Initiation, terrain slopes, CAPE,...) into a heuristic model. The purpose of COALITION is the automatic detection of severe thunderstorms early in their development phase and thus an anticipation of severe convective cells available in form of probability maps.

At present state thunderstorm warning is classically performed “on detection”. These developments aim at moving in the next years towards a “warning on early detection”, i.e. to help weather forecasters to improve the nowcasting and to extend the lead-time when issuing severe weather warnings. An anticipation even of a few minutes can be crucial in areas characterized by complex orography like the Alpine and pre-Alpine region, in particular when dealing with relatively small but high-impacting thunderstorms that can produce local flash-floods, intense hail and wind gust, causing severe damages.

In the pre-Alpine and Swiss plateau regions, orography can also lead to the organization of fast moving convective cells at the meso-beta scale. Although these thunderstorms are well detected and tracked by the existing systems in most cases, an improved early detection could help also in this situation to mitigate the damage potential of these cells by increasing the lead-time of the flash-news warnings. On the other hand, an enhanced nowcasting is needed for all convective storms during the whole life cycle, from early detection until dissipation, to better characterize the different phases of the thunderstorms, as well as the future evolution of already well-developed cells.

To address these issues we attempt a new approach that combines the nowcasting system TRT and the COALITION system. Both systems are object-based, run on the same spatial domain and have the same temporal resolution. The integration of information from different sources like radar nowcasting products, Meteosat Second Generation Rapid Scan Service products, lightning data, NWP models, orographic forcing and climatology should allow it to improve nowcasting during the whole thunderstorm cycle, from early detection until dissipation and to better characterize the different phases and the evolution of a convective storm.

The COALITION information about the potential of a strong thunderstorm development for the next 30-60 minutes will be ingested into the TRT chain and used by the cell severity ranking algorithm. The ultimate goal is to automatically integrate this information about potentially dangerous convective cells into the TRT-objects, which will then be used by the forecasters to issue severe thunderstorms flash-news “warnings on early detection”. When a sufficient accuracy will be reached by this system, automatic warnings could be considered. The paper will focus on this integration and show first preliminary results.