This dataset serves as training data for modeling the temperature field emanating from several open loop groundwater heat pumps (six heat pumps, randomly placed). It is simulated with Pflotran and stored in h5 format. The dataset contains 1000 data points, each consisting of one simulation run until a near steady state is reached. Each data point measures 3.2 x 3.2 km with 640 x 640 cells. The varying parameter of the dataset is the heterogeneous permeability with a fixed minimum and maximum. Other parameters that define the dataset, such as porosity, are chosen to be as close as possible to reality. Source: "Die hydraulischen Grundwasserverhältnisse des quartären und des oberflächennahen tertiären Grundwasserleiters im Großraum München", Geologica Bavarica Volume 122. Generated with scripts from Dataset generation with Pflotran (commit 78e8095) with arguments given in inputs/args.yaml.

The dataset is prepared with Heat Plume Prediction into reasonable format, reduced set of in/outputs, 2D, normalization. Last relevant git commit: cae87d68faf96b2bd8dab935.