Analysis of Spatial Data for Hydrological Modelling of Tiga Dam Watershed, Kano, Nigeria

Abstract

Hydrological modelling depends on reliable spatial datasets that represent watershed terrain, land use/land cover (LULC), soil and drainage characteristics. This study analyzed and prepared spatial datasets for subsequent hydrological modelling of a selected 116.46 km² portion of the Tiga Dam watershed, Kano State, Nigeria, using Geographic Information System (GIS) and remote sensing techniques. A 30 m Shuttle Radar Topography Mission (SRTM) Digital Elevation Model was processed in ArcGIS to derive elevation, slope, flow direction, flow accumulation, drainage networks, and the watershed boundary. Sentinel-2 imagery with 10 m spatial resolution was classified using the Maximum Likelihood supervised classifier into vegetation, agricultural land, built-up area and water body classes. The resulting LULC distribution comprised agricultural land (75.19 km²; 64.56%), vegetation (25.41 km²; 21.82%), water body (13.32 km²; 11.44%), and built-up area (2.54 km²; 2.18%). Accuracy assessment based on 69 validation samples produced 88% overall accuracy and a Kappa coefficient of 0.84. Soil information obtained from OpenLandMap was classified into Hydrologic Soil Groups A–D, while DEM-derived flow characteristics were used to delineate the watershed and its drainage/sub-basin structure. All datasets were spatially harmonized through projection, clipping and resampling. The resulting terrain, LULC, soil and drainage layers provide a documented spatial basis for subsequent watershed and Hydrological Response Unit development. No rainfall–runoff simulation, SWAT calibration or validation was conducted. Therefore, the findings represent spatial characterization and preprocessing outputs not direct hydrological model predictions.