HYGRID-M: A grid-based hydrological balance model for water management at River Basin District scale

Accurate water balance estimation is crucial for sustainable water management, particularly under increasing climate variability. This study presents HYGRID-M, a monthly grid-based hydrological model developed in Python for the River Basin District (RBD) scale. Its main advantages lie in (i) explicit representation of spatial heterogeneity in soil properties derived from dynamic land use, (ii) the application of a temperature-based Hargreaves evapotranspiration formulation tailored to Mediterranean climatic conditions. The model was applied to the Southern Apennines District in Italy (2000−2023), where soils and land use are highly heterogeneous and temperature-driven evapotranspiration plays a dominant role. When validating the modeled AET against estimates from GLASS, ETMonitor, and MOD16, the HYGRID-M model exhibited significant agreement with MOD16 followed by ETMonitor and GLASS. Including heterogeneous soil depths derived from dynamic land use data and a regionalized Hargreaves coefficient for southern Italy significantly improved AET accuracy . Moreover, Q estimates closely aligned with precipitation forcing and were comparable to BIGBANG outputs. The model further revealed its sensitivity to the spatial heterogeneity in soil properties in estimating Q at the RBD scale . At the basin scale, the calibrated runoff coefficient ( β ) improved the model performance with KGE and NSE reaching 0.87 and 0.89, indicating the transferability of the model from the RBD to basin scales. Overall, these results demonstrated HYGRID-M's potential as a reliable tool for water management in Mediterranean climate-sensitive regions.