A new hydro-meteorological precipitation and flood forecasting system for small catchments

Small, fast responding catchments are prone to flooding induced by local convective precipitation events. In such cases, the availability of high-quality information as basis for target-oriented warnings and effective protective measures in potentially affected regions is essential. More precisely, they rely on high-quality precipitation data, reliable weather and flood forecasting and last but not least information and education of persons in charge for prevention of hazards.

Within the project ‘HoWa-innovativ’ funded by the German Federal Ministry of Education and Research (BMBF) a new hydro-meteorological processing chain has been established to improve flood forecasting and warnings for small catchments. Study regions were located in the federal state of Saxony. The project consisted of three focal topics: first, the improvement of precipitation estimates by adding data from Commercial Microwave Links (CML) to the DWD gauge-adjustment of radar-based Quantitative Precipitation Estimates (QPE); second, the development of a hydrological ensemble prediction system designed for small catchments including coupling to DWD meteorological ensemble data; third, the development of a demonstrator with tailored visual information on precipitation analyses and forecasts as well as flood forecasts and its introduction to clients in disaster and flood management in specific workshops.

Small scale heavy precipitation events are difficult to detect and quantified with conventional measurements. DWD combines the areal reflectivity measurements of the radar network with quantitative gauge measurements throughout Germany to provide QPE in near real-time to clients in flood risk management. This so-called radar online adjustment (RADar-OnLine-ANeichung, RADOLAN) performs well for hourly precipitation sums. However, for small-scale, rapidly evolving convective cells the availability of corresponding gauge information is limited. In contrast to precipitation gauges, CMLs are by far more numerous and have the potential to provide near-ground precipitation information with good spatial coverage, high temporal resolution and very low latency. Within the project, the retrieval of precipitation information from CML has been optimized and prepared for automatic processing. We use CML data for the adjustment of radar-based precipitation estimates with focus on high spatial accuracy and the potential for higher temporal resolution.

This contribution gives an overview of the project results comprising the whole processing chain, while the focus will be on the new multi-sensor precipitation QPE system pyRADOLAN.