Development of reliable area information on groundwater levels through the use of Machine Learning/Deep Learning (ML/DL) methods

The aim is to develop and evaluate a contemporary, scientifically sound and at the same time practice-oriented approach to the derivation of groundwater level information at the federal level on a scale of 1:250,000. To this end, based on a comprehensive literature review, through the systematic collection and integration of data, to the development and evaluation of deep learning approaches, the methodological foundations are to be worked out that enable integration into an existing specialist information system (groundwater level prediction at individual measuring points) at the BGR.