Hydrological modeling and simulation of water balance components in northern Algeria's dam watersheds using GIS and SWAT

ABSTRACT

The graphical abstract presents the workflow and main findings of the study. Hydro-meteorological data (precipitation, temperature, humidity, wind, and radiation) and GIS spatial data (DEM, land use, soil, and river network) are integrated into the SWAT model using ArcSWAT. The model is applied to five reservoir watersheds in northern Algeria and includes warm-up, monthly calibration, and monthly validation. Parameter sensitivity analysis using SUFI-2 identifies CN2, GW_REVAP, and SOL_AWC as the most influential parameters controlling streamflow simulation. Model performance is evaluated using monthly discharge observations and independent evapotranspiration validation with GLEAM v4.2a data, achieving satisfactory Nash–Sutcliffe Efficiency (NSE) values of 0.70–0.88 during calibration and 0.67–0.80 during validation. The model quantifies major water balance components and annual reservoir inflows. The results demonstrate that SWAT provides a reliable basis for hydrological assessment, reservoir inflow prediction, and sustainable water resources management, supporting climate change adaptation and decision-making for Algerian water authorities

In this study, the Soil and Water Assessment Tool (SWAT), integrated with a Geographic Information System (GIS), was e mployed to simulate streamflow and quantify water balance components across five strategic watersheds supplying reservoirs in northern Algeria under increasing water scarcity. The Sequential Uncertainty Fitting-2 (SUFI-2) approach was applied to assess parameter sensitivity and model un certainty. Sensitivity analysis indicated that the SCS Curve Number (CN2), groundwater ‘revap’ coefficient (GW_REVAP), and soil available wat er capacity (SOL_AWC) were the most influential parameters for streamflow simulation. The study used monthly discharge time series for rigorou s calibration and validation, complemented by independent evapotranspiration validation using satellite-based GLEAM4 (v4.2a) estimates. Model performance showed good agreement between observed and simulated flows, with Nash–Sutcliffe Efficiency (NSE) values ranging from 0.70 to 0.8 8 during calibration and from 0.67 to 0.80 during validation. The results highlighted spatial variability in hydrological processes: lateral flow (LAT_ Q) dominated in the El Hamiz and K'sob watersheds, whereas surface runoff (SURQ) prevailed in the Boukerdane, Hammam Debagh, and Ouizert basins. This study confirms the capability of SWAT to reproduce the hydrological behavior of northern Algerian watersheds and its reliability as a decision-support tool for reservoir planning and sustainable water resource management.