MHiDROC3: A Distributed Hydrological Modeling Framework for Streamflow Simulation Across Contrasting Watersheds in Chile

Abstract

Hydrological modeling across heterogeneous watersheds remains a key challenge for water resource assessment in regions with strong hydroclimatic gradients. This study presents and evaluates the Chilean Hydrological Model for Climate Change (MHiDROC3) that integrates hydrometeorological and geospatial inputs, data imputation, sensitivity analysis, parameter calibration, distributed hydrological simulation, water-demand representation, and climate scenario analysis within a single workflow. MHiDROC3 was applied to five contrasting watersheds in south-central Chile over the 1980–2021 historical period. Across calibration, validation, and full-period simulations, KGE ranged from 0.51 to 0.77, while natural-streamflow simulations for the complete period showed KGE values of 0.52–0.74 and NSE values of 0.15 to 0.64, indicating variable model performance among basins. Mean simulated streamflow differed from observations by −32.2% to 17.7% across watersheds. Future simulations for 2022–2100 projected lower streamflow under SSP5–8.5 relative to SSP1–2.6, particularly during fall (−33.1% to −57.5%) and winter (−16.0% to −35.9%). By combining distributed process representation, demand effects, and climate-scenario testing in a Chilean framework, MHiDROC3 provides a practical basis for basin-specific water resource assessment, while its variable performance indicates that local evaluation remains necessary before operational application.