Dynamics of water balance indices through a fully distributed hydrological model, ground data and earth observation

Water is a necessary condition for human existence and represents one of the most critical elements for maintaining ecosystem balance, human survival, and socioeconomic development. Water abstraction (WA) for different uses wields the most pressure on freshwater resources. Understanding the impact of WA on the spatial and temporal pattern of natural water resources is the key to developing the Water Management Plan (WMP). This work utilizes a fully distributed physically-based, raster-based hydrological model (MOBIDIC) to evaluate the Flow Duration Curve (FDC) and the Water Exploitation Index (WEI) in Tuscany (Italy) using 20 years of daily data from different sources. MOBIDIC represents watersheds using a system or reservoirs that interact through mass and energy fluxes. It shapes precipitation, runoff, evapotranspiration, and groundwater storage using ground point data, such as river discharge, precipitation, wind velocity, humidity, solar radiation, data on withdrawals and releases, and grid data containing spatial information on the soil. Remotely sensed data such as evapotranspiration (ET) and surface soil moisture (SSM) products are a good asset in getting distributed information and supporting the hydrological model's calibration and validation. MODIS data from Terra EOS AM-1 satellite have been used to calibrate the model evapotranspiration flux. Instead, Sentinel data from the Sentinel-1 C-SAR satellite have been used to calibrate the MOBIDIC simulated soil moisture. The results show that simultaneous assimilation and validation of ground and satellite data used for the model parameters estimation significantly improves the model performance in predicting river discharge and water balance indices.