Impact of land-use and land-cover change on watershed hydrology: a case study of Mojo watershed, Ethiopia

The impact of land-use/-cover change on the hydrology of the Mojo watershed in Ethiopia was conducted using SWAT and partial least square regression (PLSR) model. The study is vital for the management for the water resources in the watershed. The SWAT model was used to quantify the changes in streamflow using three-time period land-use maps, while the influence of changes in individual LULC on the changes in streamflow was assessed using the PLSR model. The results of land-cover detection indicated a decline in natural vegetation due to the continued increase in cultivated land and urbanization. The results also revealed that LULC, which occurred from 1993 to 2005, resulted in increased annual streamflow by 3% and wet month streamflow by 8.1%, while the dry season flow was reduced by 8.9%. Comparing land-use maps of 2005 and 2018, the annual flow increased by 1.4%, the wet seasonal flow increased by 3.8%, and the dry season flow reduced by 8.1%. Surface runoff was found to increase and groundwater was reduced over the study period. The results of the PLSR model also indicated that annual streamflow is strongly correlated with cultivated land and waterbody with variable influence projection (VIP) and weight (w*) values (VIP = 1.15 & w* = 0.38), while the dry season streamflow is strongly correlated with a plantation (VIP = 1.2 & w* = 0.4). It is recommended that proper land-use planning and zoning should be implemented in the watershed to help reduce the rapid land-use/-cover changes.