The Study On Hydrological Simulation By AVSWAT Model In Plain Areas Of Hebei Province

Water, the carrier of life, is a strategic resource for the development of the country and society. Especially in recent decades, water resources emerge new features with the economic development. More and more attention was paid to the influence of climate change and human activities on water resources. Change in land use to some extent affected the regional runoff. Furthermore, the regional runoff can be influenced by the tendency of increasing precipitation which was attributed to the rising of country's average temperature in the future. Simulation of surface runoff is vital to the water balance calculation and calculation of non-point source pollution and so on. Hydrological model is one way to study the rule of hydrological changes. Among these models, the distributed hydrological model is important and popular. The SWAT model, as a representative of distributed hydrological model, was used extensively in many fields such as runoff simulation, calculation of non-point source pollution, agricultural management.Based on a large number of home and abroad relevant literatures, this study selected Hebei Plain as research area to simulate its runoff using SWAT model, the main results were as follows:(1) The research area's spatial database and attribute database were established under the auxiliary of the GIS software. Spatial databases included digital elevation model (DEM) map, land use map and soil type maps. Attribute databases included precipitation database, temperature database, wind speed database, attribute database of land use methods and soil properties.(2) This study analyzed the study area's DEM map, extracted watershed's digital water system, and discretized watershed. Sub-basin area's threshold is 2500hm². The study area is divided into 37 sub-basins. The threshold of land use methods and soil properties defined respectively as 20% and 10%. 37 sub-basins is divided into 73 hydrological response units (HRU).(3)The main parameters which affect the results of runoff model are runoff curve number CN₂, soil saturated water content and hydraulic conductivity( SOL-K), soil available soil water content(SOL-AWC), soil evaporation compensation coefficient( ESCO), groundwater recharge delay time(GW-DELAY) and base flow regression coefficient( ALPHA-BF). And the relationship between those parameters and runoff was analyzed in order to calibrate and verify the model.(4) After parameters which have much effect on the result of runoff model were given, the model was calibrated by the study area's runoff datum measured monthly in 1998 and 1999 by the method of manual trial and error. The results indicated that correlation coefficient(R) of runoff in calibration period is 0.98, and the model effective coefficient (Nash-Suttcliffe) and certainty coefficient (Ens) is 0.97. After parameter values determined by model calibration, then the study area's runoff datum measured monthly in 2000 were used to verify the model. The results suggested that R of validation period is 0.94, and Nash-Suttcliffe and Ens is 0.76. Both calibration and validation's R and Nash-Suttcliffe and Ens can meet the precision requirements, which verify the SWAT model's applicability in this region. Using the SWAT model to simulate runoff in this region is feasible.(5)This study demonstrates that based on the other conditions remaining unchanged, a region's runoff variation may occur when its temperature, precipitation, land use methods and other factors changed from the theoretical point of view.