MODELING THE IMPACTS OF LAND USE CHANGES ON WATER QUANTITY AND QUALITY IN KIMWARER RIVER CATCHMENT

Abstract

Kimwarer River basin has an area of 138.2 km2 . The basin has experienced ecosystem degradation due to extensive farming therefore impacting on water yield and nutrient loading in the basin. The study was undertaken to assess the impacts of land use changes on the quantity and quality of river flow by application of SWAT model. SWAT is a mathematical model that has potential of predicting the impact of land management practices on water and chemical yields at catchment scale. The specific objectives of the study were; i) To determine land use changes for the past 30 years; 1985 to 2015,ii) To calibrate and validate SWAT model for prediction of stream flow and nutrient loading in various land use scenarios and iii) To simulate stream flow and nutrient losses (nitrates and phosphates) in various land uses. Input data to SWAT model were assembled from relevant sources. Current and historical flow data were collected for the purpose of model calibration and validation. The model was then used to simulate quantity and quality of stream flow for different land use and land cover scenarios by interchangeably varying forest cover and agriculture. Results showed that from the year 1986 to 2010 for which images were obtained, there were; increase in agricultural land from 62.68% in 1986 to 84.52% in 2010, decrease in forest cover from 17.33% in 1986 to 9.66 % in 2010 and the decrease of wetland from 19.99% in 1986 to 5.84% in 2010. The model was successfully calibrated and validated for stream flow. The model was capable of predicting flow during calibration with R2 and NSE values of 0.79 and 0.31 respectively. During validation, the model predicted flow with R 2 and NSE values of 0.70, 0.50 respectively. For scenario analysis to determine the effect of land use change on stream flow, it was observed that surface runoff decreased with increase in forest cover while base flow increased with increase in forest cover. Introduction of terraces as management operation option in agricultural land showed a reduction in runoff by 46%. The model was capable of predicting nitrate and phosphate losses with R2 of 0.72 and 0.79 respectively. The observed mean monthly basin loading levels were 0.10 kg/ha and 0.092 kg/ha for nitrates and phosphates respectively. Nutrient losses decreased with increase in forest cover. From the research it is evident that the current trend of land use change affects the quality and quantity of water from the catchment. The trend requires control for sustainability.