IMPACT OF LAND-USE SCENARIOS ON MONTHLY RUNOFF MODELED BY SWAT

Li Zhang,Ye Liu,Lei Wu,Duo-duo Cai,Xiao-yi Ma
2016-01-01
Fresenius environmental bulletin
Abstract:We developed a modeling system, integrating the SWAT (Soil and Water Assessment Tool) and the CA-Markov (cellular automata and Markov) models, to evaluate the impacts of future land-use changes and two different land-use scenarios (Scenario1 and Scenario2) on monthly runoff in the Daxi River Watershed of Shaanxi Province, China. The CA-Markov model could predict future land uses of the watershed based on historical land-use data, with a kappa coefficient (kappa) of 0.83. The GA-PSO (genetic algorithm and particle swarm optimization) algorithm was applied to calibrate and validate the SWAT hydrological parameters. Average monthly runoff of the land -use types tended to decrease in both wet and dry seasons with the change rate of 0.3%-6.8% and 0.1%-5.5%, respectively, compared with the previous-one. Mean monthly runoff in two scenarios decreased (except December) with an increase and decrease in forest and cropland area, respectively. Average monthly runoff decreased by 9.8%-27.3% and 0.9%-11.7% in one scenario, and by 4.9%-19% and 0.4%-6.8% in the other scenario in the wet and dry seasons, respectively.
What problem does this paper attempt to address?