论文部分内容阅读
为在江西修水流域高沙水文站以上汇水区域构建SWAT分布式水文模型,采用SUFI-2算法进行模型参数敏感性分析、率定、验证及不确定性分析,对修水流域1984~2003年月径流量过程进行模拟。结果表明,敏感度最强的前三个参数依次为土壤饱和导水率、土壤蒸发补偿系数和径流曲线数;率定期与验证期的纳什系数ENS值和相关系数R2值均高于0.90,月径流量模拟值与实测流量过程线拟合程度较好;95PPU置信区间的模拟值和实测值较为一致;不确定性区间较小,验证期小于率定期;实证了采用SUFI-2算法的SWAT模型在修水流域月径流模拟中具有一定程度的适宜性。
In order to construct the SWAT distributed hydrological model above the catchment area of Gaosha Hydrographic Station in Xiushui River Basin, the parameter sensitivity analysis, calibration, verification and uncertainty analysis of the model parameters were carried out using the SUFI-2 algorithm. Annual runoff process simulation. The results showed that the first three parameters with the highest sensitivity were soil saturated hydraulic conductivity, soil evaporative compensation coefficient and runoff curve number, respectively. The values of ENS and correlation coefficient R2 of Nash coefficient in periodical and verification period were all higher than 0.90, The simulated values of runoff and measured flow were better fitted. The simulated and measured values of 95PPU confidence interval were consistent, the uncertainty interval was small, and the verification period was less than the rate. The SWAT model with SUFI-2 There is a certain degree of suitability in simulating monthly runoff in Xiu-Shui Valley.