秃尾河流域降水量权马尔可夫链模型预测研究

Prediction on changes of precipitation in Tuwei River reaches based on weighted Markov Chain

针对降水过程存在着大量不确定性、不精确性的特点,采用均值—标准差分级法将降水量分为雨涝、偏涝、正常、偏旱和干旱5个状态;针对降水量为相依随机变量的特点,采取规范化的各阶自相关系数为权重,用加权马尔可夫链模型预测未来降水的变化状况;最后以黄河中游秃尾河流域为例对该方法进行了检验,分析了其未来情况下降水的可能变化趋势。结果表明,2002年秃尾河流域降水量预测状态与实际相吻合。

Mean-standard deviation classification method is applied to classify precipitation of five states,i.e.water-logging year,weak water-logging year,normal year,weak drought year,and drought year based on the fact that there are much uncertainty and imprecise characteristics in the precipitation course.Then a method called Markov chain with weights is presented to predict future precipitation state by regarding the standardized self-coefficients as weights based on the special characteristics of precipitation being a dependent stochastic variable.Finally the middle reach of the Tuwei River of the Huanghe River basin is taken as an example to carry out the method and analyze the possibility and change patterns of precipitation in the future condition. The results indicate that the precipitation prediction state of Tuwei River consistent with the practical situation in 2002.