The potential of novel remote sensing evapotranspiration data and global soil maps for SWAT+ agro-hydrological modeling in data-scarce regions of the North Mediterranean

Mediterranean ecosystems are increasingly affected by rising temperatures, shifting precipitation patterns and water resource overexploitation, while limited ground-based data challenge effective land and water management. The Messinia region in the southwestern Greece is severely affected by soil degradation and inadequate water management practices. Covering about 70 % of olive farming, the region plays a crucial role in the regional agricultural economy. The study uses the SWAT+ model for comprehensive agro-hydrological modeling and multisite calibration of four ungauged watersheds in the Messinia region. It integrates remotely sensed evapotranspiration data as reference data, high-resolution global soil maps and agricultural practices (planting, irrigation, and harvesting). The results show that incorporating DSOLMap soil data improved simulations over a local soil map. Additionally, GLEAM, aligns better with FAO-56 method and offers a more accurate representation of AET than MODIS in the Messinia region. Our findings are in line with the literature which reports that GLEAM accounts for soil moisture and vegetation dynamics while MODIS has known limitations in capturing irrigation effects and fine-scale variability. The SWAT+ model using GLEAM, DSOLMap, and management schedules achieved the best results (NSE > 0.5; PBIAS < ±15 %), outperforming other model setups, particularly when compared to MODIS-based simulations which showed underperformance. This approach that combines SWAT+ with remote sensing and integrates management schedules, offers a more accurate representation of the water cycle and enhances water resource management in data-scarce regions. Additionally, the approach is scalable and replicable in other data-scarce regions and offers a valuable tool for site-specific management strategies. • GLEAM outperformed MODIS in AET simulation, especially in agricultural-dominated landscapes. • High-resolution soil data (DSOLMap) improved SWAT+ accuracy in AET simulation in ungauged small watersheds. • Integrating crop statistics and irrigation schedules improved AET and water balance accuracy in SWAT+. • SWAT+ model performance was sensitive to land cover type and resolution of input data in small heterogeneous watersheds. • Our modeling framework supports site-specific water management and is scalable for data-scarce regions.