Dynamic agricultural weather indicators for extreme weather prediction in agriculture

The project “DynAWI – dynamische Agararwetterindikatoren” (dynamic agriculture weather indices) aims to develop a process chain for data integration and real-time analysis for extreme weather. Extreme weather events have a major impact on agriculture and horticulture and cause significant economic costs. The damage depends not only on the type of extreme weather event (e.g. heat wave, drought stress or heavy precipitation), but also on the ontogenetic development of the crops. Previously, farmers calculated their risk with fixed weather indicators and because of the multi-dimensionality of the source data and it was difficult to calculate using traditional relational databases in an acceptable time.

We have developed a web application for real-time calculation of dynamic weather indicators by linking a back-end infrastructure of Datacube servers and a Vue front-end infrastructure with a machine learning model in an R environment. The web application can perform real-time analyses based on multi-dimensional spatio-temporal data. Future plans include enriching the web application with additional agricultural weather indicators and linking it to weather forecasts to provide an in-season risk assessment for crop losses.