NIQKI

The comprehensive, high-resolution acquisition of highly valid meteorological data is a crucial prerequisite for reliable precipitation forecasts and the processes based on them. Currently, such data are not available in many regions (e.g., North Rhine-Westphalia) with the required quality and timeliness. Ensuring the prompt availability of valid data for hazard mitigation and similar purposes requires automated validation processes that are transferable and applicable across a wide range of systems.

The aim of the project is to develop procedures for verifying the plausibility of precipitation data by incorporating additional climate data and utilizing artificial intelligence (AI) methods for these applications. For the first time, this project employs high-density, widely available spatial data, using a specific federal state as a case study. Improvements to the procedures, user processes, and user acceptance will be analyzed and supported throughout the project.

The project aims to develop methods for validating precipitation data using artificial intelligence to improve meteorological forecasts and associated processes. Collected meteorological data are transmitted to the cloud and automatically checked for plausibility within one minute, enabling their use in near real-time. The developed procedures and validated data are made available to LANUV (State Agency for Nature, Environment and Consumer Protection) and integrated into products from HST and hydro & meteo GmbH to enhance climate and flood forecasting.

Implementation

Data collected by LANUV are transferred to the cloud, immediately validated (within < 1 minute), and fed into a wide range of applications for optimal use in near real-time. The AI ​​models are developed and trained using historical data and existing validation expertise (manual checks, semi-automated checks in NIKLAS), and the results are compared with those of previous methodologies.

Further information here: https://fue-hydromet.de/niqki/