气候变化影响下极端水文事件的多变量统计模型研究

以黄河流域太原气象站和淮河流域鲁台子水文站为研究对象，利用Copula函数构建气候要素（降水）同极端水文事件（干旱和洪水）之间的多元统计模型，分析不同降水条件下不同等级干旱和洪水的发生概率变化。结果表明，Gumbel Copula能够较好地描述太原站7月份的前期累加降水量和帕默尔干旱指数（PDSI）的相关结构。随着降水量的增加，极端干旱的发生概率逐渐减小，重旱、中旱和轻旱的发生概率则先增加后减小。Clayton Copula能够较好地描述鲁台子水文站前期累加降水量和洪峰流量之间的相关结构。当前期累加降水量大于等于某一定值时，随着年最大洪峰x的增大，发生洪峰≥x的极端洪水事件的概率逐渐减小。在同一个极端洪水发生概率下，前期累加降水量越大，洪峰流量出现大值的可能性越大。

The Multi-Variable Statistical Models of Extreme Hydrological Events Under Climate Change

The probability change of extreme hydrological events under climate change is analyzed by constructing multi-variable model between climate factors and extreme events. Taiyuan meteorological station in the Yellow River basin and Lutaizi hydrological station in the Huaihe River basin are selected as study area. The results of Taiyuan station show that Gumbel Copula can better simulate the dependence structure of antecedent cumulated precipitation and monthly PDSI series. As the precipitation amount increases, the probability of extreme drought decreases while the probability of severe, moderate and slight drought firstly increases and then decreases. The results of Lutaizi station show that Clayton Copula can better simulate the dependence structure of precipitation and peak of flood. When the precipitation is over or equal to a certain value, the conditional probability of the extreme flood events that peak flow is over or equal to x gradually decreases as the annual maximum peak flow increases. Under the same conditional probability, the peak flow is more likely to get large values when the cumulative precipitation is large.