Application of high-resolution multispectral data in drought assessment on the Italian territory

Drought is a complex natural phenomenon characterized by prolonged periods of unusually low water availability, resulting in an imbalance between water supply and demand, severely affecting ecosystems, agriculture, and water resources, particularly in Mediterranean regions highly vulnerable to climate variability. The increasing frequency and intensity of drought events highlight the need for timely and spatially detailed monitoring of drought signals. This thesis explores the potential of products derived from Sentinel-2 (S2) data, mainly the fraction of Absorbed Photosynthetically Active Radiation (fAPAR), for assessing vegetation health and drought impacts. The study demonstrates that S2-derived fAPAR is consistent with other established fAPAR datasets (e.g., MODIS, CGLS) in both temporal dynamics and drought-related anomalies. Despite its relatively short temporal record, S2 fAPAR can reliably detect drought-affected periods. By combining vegetation signals with satellite-observed water extent dynamics, the analysis shows that S2 fAPAR can complement hydrological drought assessment by reducing uncertainties related to anthropogenic interventions in water reservoirs. A multi-scale analysis demonstrates vegetation stress responses vary by crop type and across fields. High spatial resolution of S2 data enables a finer characterization of vegetation responses, providing valuable insights for agricultural drought management. Overall, the thesis highlights the added value of S2 data as a high-resolution resource for monitoring vegetation stress in supporting early warning systems and localized drought mitigation. Integrating S2 fAPAR with other drought indicators can further enhance drought assessment across scales. While the inclusion of in-situ data would improve validation, the fully satellite-based methodology proposed here remains applicable in data-scarce regions.