Assessment of Spatial and Temporal Soil Water Storage Using a Distributed Hydrological Model

Abstract

Hydrological models are the main tools for water resources management. The Lavras Simulation of Hydrology (LASH) model was developed for watersheds with scarce data, and great results have been obtained in Brazil. This study aimed to incorporate hydrological response units (HRUs) in the LASH model to assess soil water storage (SWS) across space and time. The LASH model was calibrated and validated for the Grande River basin upstream of the Furnas Hydropower Plant, in southeastern Brazil. The Nash-Sutcliff coefficient (CNS) and its logarithmic version were analyzed in terms of both calibration and validation to appraise the model’s performance in a daily time step. The CNS for calibration and validation was 0.86 and 0.77, respectively, showing that LASH using the HRUs produced improvements in the simulations. The calibrated parameters showed a good relationship with hydrological processes in HRUs, and SWS estimates reflected the soils, topography, and land use of the watershed. LASH could describe the SWS behavior and identify the sub-watersheds with the highest and the lowest values. Therefore, the LASH model is a promising tool for SWS simulation in time and space.