Indication of wet and dry periods in Germany using machine learning on meteorological and remote sensing data

Abstract

This thesis investigates the calculation, comparison, and forecasting of the Standardized Precipitation-Evapotranspiration Index (SPEI) using diverse data sources and methods. The study recalculates SPEI values for Germany employing various potential evapotranspiration (PET) equations—FAO-56 Penman-Monteith, Thornthwaite, Priestley-Taylor, and Hargreaves — using weather station data from the German Weather Service and remote sensing data from ERA5-Land Reanalysis. Significant variations were observed in SPEI values across different PET equations, with the Thornthwaite and Hargreaves equations showing the closest alignment with the SPEIbase - created by the SPEI inventors. The remote sensing approach with the FAO-56 Penman-Monteith PET equation yielded smooth and seasonally consistent SPEI values.