新安江模型敏感参数动态变化规律研究

Study on dynamic variations of sensitive parameters in Xin’anjiang model

为了提高新安江模型在变化环境下的模拟精度,揭示模型参数动态变化规律,以淮河上游大坡岭流域为研究区构建分布式新安江模型,采用Spearman相关系数对比分析了模型敏感参数(KC、SM和CS)动态变化与环境因子(降水、气温和归一化植被指数)的变化关系,并建立了回归方程用于未来参数计算。结果表明:参数KC和SM的动态变化反映了研究区域的降水、气温和植被条件变化,而参数CS的动态变化仅反映了年降水总量变化;相较于静态参数,动态参数的模拟结果精度显著提高,径流深相对误差均值从21.3%降低到9.1%,纳什效率系数均值从0.75提高到0.81,动态参数比静态参数更适用于当前变化的环境。

In order to investigate the dynamic variations of parameters in the Xin’anjiang model and improve the model accuracy under changing conditions,a distributed Xin’anjiang model was implemented for the Dapoling Basin in the upper Huai River.The impact of changes in precipitation,temperature,and NDVI(normalized difference vegetation index)on the dynamic patterns of the sensitive parameters KC,SM,and CS was analyzed using the Spearman correlation coefficient.The results indicate that the dynamic variations of parameters KC and SM reflect changes in precipitation,temperature,and vegetation coverage conditions in the study area,while the CS parameter only reflects changes in the total annual precipitation.Compared to static parameters,the simulation results with dynamic parameters exhibit a significant improvement in accuracy,with the mean relative error of runoff depth decreasing from 21.3% to 9.1% and mean Nash efficiency coefficient increasing from 0.75 to 0.81.This suggests that dynamic parameters are more suitable for the current changing climate than static parameters.