Simulation of impacts of land-use and land cover changes on water balance in Oued Fez basin (Morocco)

Abstract

Land-use and land cover changes have great impacts on both the hydrology and water balance in a watershed and therefore evaluation of these impacts can greatly help in efficient water resources management. The Oued Fez basin has undergone severe land-use changes in the past 30 years (1988-2018). The objective of this research was to assess the impacts of land use changes in the Oued Fez basin, water balance using the Soil and Water Assessment Tool (SWAT) model, remote sensing and GIS techniques. The changes in the basin's water balance are attribute to the changes in land-use and land cover. The assessment was conducted using two land cover maps, 1988 and 2018 respectively and climatic data for period (1988-2018). The calibration was done using 17 random points of observed stream flow data collected over the period 2009-2011, while no data was available for validation. Over the 30 year period, rangelands drastically reduced by (333.62%), big increases were observed for urban (72.71%), arboriculture (90.43), water (36.85%), and irrigated agriculture (33. 78%). The CN2 and ESCO were adjusted to obtain results of long-term catchment water balance and monthly streamflow that matched the observations. The impacts of land-use change on the long term annual average water balance indicated an increase in the; total water yield by 1.20%, and surface runoff by 5.10%, reductions were also noted for: ET by 0.44%, lateral flow by 1.49%, ground water in shallow aquifer by 10.28%, and groundwater deep aquifer by 10.37%. Statistical tools including NSE, PBIAS and R2 were used to evaluate the model observed performance. NSE value of -2.45, PBIAS of 9.70 and R2 of 0.42 were obtained during the calibration period, and the performance evaluation indicated a model underestimation during the calibration period. The simulation results also indicate an influence of land-use changes, on the hydrological water balance and streamflow. Generally, the model performance was unsatisfactory and this is due to insufficient data for calibration. If all uncertainties are addressed the model would give reasonable hydrologic simulations.