基于马尔可夫链的旱涝预测应用研究  

Study on Dryness and Wetness Prediction Based on Markov Chain

作　　者：李析男[1] 毕宏伟 胡彩虹[1] 

机构地区：[1]郑州大学环境与水利学院,郑州450001 [2]长江水利委员会水文局,武汉430010

出　　处：《气象与环境科学》2007年第4期20-23,共4页Meteorological and Environmental Sciences

基　　金：国家自然科学基金项目(40572150);水资源与水科学国家重点实验室开放基金项目(2006B023)资助

摘　　要：在阐述马尔可夫思想和预测思想及其原理的基础上,明确了采用此方法的具体步骤和参数值,同时,分析了此方法的优缺点。根据1971 2000年辽宁省本溪市市区降水量资料,取适当的α值,采用均值—方差法对年降水量进行分级,并应用加权马尔可夫链对该地区的旱涝状态进行了预测,结果表明,预测值与实际值相吻合,说明此方法有一定的实际应用意义。The process applied the Markov and the parameters are elaborated on the basis of introducing the thoughts and the forecast thought of the Markov and its principle. The good and bad points of the method are analyzed. According to the precipitation data during 1971 to 2000 in the city of Benxi in Liaoning province, the paper chooses suitable α value, and grades the annual precipitation adopting mean - variance method. And the dryness and wetness status is forecasted by applying weighted Markov chain. The results show that the forecast value of weighted Markov chain and the actual value are tallied, and the method has practical significance.

关 键 词：加权马尔可夫链 降水量 预测 

分 类 号：P456.3[天文地球—大气科学及气象学]