SWAT模型在清江流域上游径流模拟中的应用

Application of SWAT Model to Runoff Simulation in the Upper Reaches of Qingjiang River Basin

利用GIS软件与SWAT模型对清江流域上游的恩施水文站进行月径流过程模拟,结合SUFI-2算法对参数敏感性进行分析以及参数率定与验证.结果表明:选取的19个与径流相关的参数中对研究区径流模拟影响较大的5个参数分别为SCS径流曲线数(CN2)、河岸调蓄的基流α因子(ALPHA_BNK)、植物吸收补偿因子(EPCO)、主河道水力传导系数(CH_K2)和浅层含水层快速蓄水的阈值水深(REVAPMN).模型率定期和验证期的相关系数R^(2)、Nash效率系数E_(ns)均达到0.90以上,率定期和验证期的百分比偏差PBIAS分别为0.48%和10%,模型模拟效果理想,表明SWAT模型在清江上游流域月径流模拟中具有很好的适应性.

The monthly runoff process of Enshi Hydrological Station in the upper reaches of Qingjiang River Basin is simulated by using GIS software and SWAT model,and the sensitivity of parameters is analyzed,calibrated and verified by SUFI-2 algorithm.The results show that among the 19 runoff-related parameters selected,five parameters which have great influence on the runoff simulation in the study area are SCS runoff curve number(CN2),basic flowαfactor of riverbank regulation and storage(ALPHA_BNK),plant absorption compensation factor(EPCO),hydraulic conductivity coefficient of main channel(CH_K2)and threshold water depth(REVAPMN)of rapid impoundment in shallow aquifer.The correlation coefficient R^(2) and Nash efficiency coefficient E_(ns) of model rate periodicity and verification period are more than 0.90,and the percentage deviation PBIAS of rate periodicity and verification period are 0.48% and 10%respectively.The simulation effect of the model is ideal,which shows that SWAT model has good adaptability in monthly runoff simulation in the upper reaches of Qingjiang River Basin.