IHACRES model regionalization for water resource management in Sicily

Abstract

Hydrological rainfall-runoff models are important tools for understanding and predicting how watersheds respond to precipitation and other meteorological inputs, considering processes such as infiltration, surface runoff, evapotranspiration. These models are crucial in a variety of hydrological analyses, including water resources management, flood forecasting, and flood risk assessment. Regardless of their type, rainfall-runoff models rely on specific parameters that must be calibrated using historical data. Calibration involves adjusting model parameters to match observed runoff values as closely as possible, ensuring that the model can accurately simulate real-world hydrological conditions. The quality and availability of historical data play a crucial role in this process. For gauged basins with available historical data, calibration is performed using observed data. However, for ungauged basins, traditional calibration is not possible. In such cases, model parameters may be estimated based on regional data, analogies to nearby gauged basins, or other indirect methods. This approach, which typically introduces additional uncertainty, uses some regionalization techniques for models’ parameters to overcome the lack of local data. These methods usually transfer calibrated parameters from gauged to ungauged basins, based on similarities in physical and climatic characteristics, making it possible to establish statistical relationships between basins’ characteristics and models parameters.