一种新的径流过程随机模拟方法

利用Copula函数构造边缘分布为PⅢ型分布的两变量联合分布,用以描述径流过程的年径流量与最大月径流量。根据该联合分布进行随机抽样,再优选实测径流过程,采用典型解集方法分解转化为模拟径流过程线,从而建立一种新的径流过程随机模拟方法。基于Copula函数的解集随机模型与现有的径流过程随机模拟模型如季节性一阶自回归模型、典型解集模型相比,其所模拟的年径流量统计特征的相对均方差等指标均较优,能较好地模拟原序列的统计特性,为水文水资源随机模拟研究提供了一种新的途径。

New stochastic simulation method for streamflow hydrographs

The Copula function was employed to construct a bivariate joint probability distribution with the marginal distribution being the Pearson type Ⅲ for describing the total annual streamflow volume and the maximum monthly streamflow volume.According to the joint probability distribution,samples were randomly drawn,and then the observed streamflow hydrographs were selected and transformed into those for simulations by means of the typical disaggregation model.Accordingly,a new stochastic simulation method for streamflow hydrographs was established.Compared with the existing models for streamflow simulations, such as the seasonal automatic regression first-order model and the typical disaggregation model,the proposed method based on the Copula function is superior in terms of the root mean square deviation of the statistical characteristics of the simulated annual streamflow volume,and it can be employed to satisfactorily simulate the statistical characteristics of the original series.It provides a new way for the studies on simulation of hydrology and water resources.