Accurate evaporation estimates are essential for applications such as irrigation planning, drought prediction, ecosystem health monitoring, and water availability assessment. Traditional in situ measurements are spatially and temporally limited, which has led to the development of satellite-based algorithms in recent years. However, these evaporation models often fail to capture the influence of soil moisture limitations or overlook plant access to groundwater, resulting in significant uncertainties under dry conditions. This project investigates novel machine learning approaches to incorporate plant access to groundwater, represent root depth dynamics, and assimilate satellite water storage observations into a global satellite-based evaporation model. Enhanced evaporation monitoring will facilitate a better understanding of the global water cycle and water resource management. This project aims at PhD research that will be conducted in collaboration with Prof. Niko Verhoest (BW20).