Effect of land use land cover changes on runoff using hydrological model: a case study in Hiranyakeshi watershed

The assessment of the effects of land use land cover changes is crucial to know the hydrological status of the river basins. Various applications like remote sensing, geographic information system (GIS), and hydrological modeling approach are required to visualize and understand these effects on the watershed hydrological system. The research focuses on assessing the effects of land use land cover (LU/LC) changes and its impacts on the runoff in the Hiranyakeshi watershed using a semi-distributed hydrological model viz. soil and water assessment tool (SWAT). Therefore, this research was initially carried out by analyzing changes in LULC by classifying Landsat 7 and Landsat 8 satellite images for the year 2000 and 2015. The LULC classification is prepared with different classes such as forest, agriculture, fallow land, plantation, water bodies, and urban area. The accuracy assessment of the classified images was performed by error matrix; the values were found to be 65% (for two classified images); the kappa coefficient value found to be 0.51 and 0.56 for the year 2000 and 2015. The classification results obtained are within satisfactory limits; hence, the classified image is used to assess the changes in LULC over the period. Results show an increase in agricultural activities (11.45%), urban area (1.15%), water body (0.15%), and a decrease in barren land (11.46%) and forest (1.42%). To simulate the changes in the runoff, the classified LULC image for the year 2000 is used in the SWAT model during the calibration (from 1995 to 1999) and validation (from 2000 to 2005). The performance results of the model are evaluated with statistical indicators R2 and NSE, and the value 0.73 and 0.83 during the calibration and 0.70 and 0.84 during the validation period was achieved. Further, the model is simulated using classified LULC image of the year 2015 to assess changes in annual runoff; it is observed a decrease in annual runoff in the study area with respect to changes in LULC.