After the devastating flash floods in the Ahr valley in July 2021, the insurance industry identified the need to have access to near real-time data in order to better prepare for such events. They have therefore commissioned a platform that can process any Earth Observation data for the short-term integration of real-time data from current flood and disaster events. The goal is to ensure the continuous provision of open data satellite and weather data from flood warning systems (i.e, GloFAS) and flood monitoring systems (GFM, EMSR), as well as precipitation radar data from the German Weather Service (RADOLAN). The resulting information product contains the classified data sources as well as the raw data used and is made available in near real-time (NRT) in WebGIS for further analyses.

Throughout the first phase of development in 2024, 52°North calculated statistics on the aforementioned raster time series products and estimated pixel-by-pixel extreme value distributions. Based on the distributions, we derived thresholds according to a given return period per data source. We also jointly developed the architecture and workflow concept based on Argo. The platform pattern implements data stream enrichment with extensive processing on a maximally scalable Kubernetes platform. With the help of automated Argo workflows, an information product is generated daily from forecasts, floods and suspicious areas and made available for further processing via an API. The data is delivered as decision-ready information that has been co-created with the end users of the product.