The complete and small-scale collection of highly valid meteorological measurement data is a crucial prerequisite for reliable precipitation forecasts and the processes based on them. At present, such data is not available in many places (e.g. NRW) in the required quality and timeliness. The prompt availability of valid data for hazard prevention, etc. requires automated testing processes that are transferable and applicable to a wide variety of systems.
The aim of the project is to develop methods to check the plausibility of precipitation data, taking into account other climate data and using artificial intelligence (AI) methods for the aforementioned applications. In this project, data will be used for the first time in high density and availability for the area using the example of a federal state. The improvement of procedures, user processes and user acceptance will be analysed and monitored.
The project aims to develop methods to check the plausibility of precipitation data using artificial intelligence to improve meteorological forecasts and associated processes. The collected meteorological data is transferred to a cloud and automatically checked for plausibility within one minute, enabling near-real-time use. The developed processes and plausibilised data are made available to the LANUV (State Agency for Nature, Environment and Consumer Protection) and integrated into the products of HST and hydro & meteo GmbH in order to improve climate and flood forecasts.
Realisation
The data collected by the LANUV is transferred to a cloud, immediately plausibilised (< 1 minute) and fed to a large number of applications for optimal use in quasi-real time. The AI models are developed and trained on the basis of historical data and existing plausibility checks (manual checks, semi-automatic checks in NIKLAS), and the results are compared with those of previous methods.
The research here is tough.
News about the inwa and its events.
We look forward to receiving your enquiry.
© 2024
EN 
DE