基于权马尔可夫链的宝鸡市年降水量状态预测

为了全面分析宝鸡市自然降水的时间分布规律和变化趋势，为该区域水资源预测、旱涝灾害防治、水量调度和水资源管理提供依据，对宝鸡市年降水量状态进行了权马尔可夫链分析。针对降水过程中的不确定和不精确性，采用均值-标准差分级法，将宝鸡市1956-2009年降水量分为干旱、偏旱、正常、偏涝、雨涝5个状态。以规范化的各阶自相关系数为权重，采用权马尔可夫链模型对宝鸡市未来年份的降水量状态进行预测。平稳分析结果表明，宝鸡市多年降水过程中干旱、偏旱、正常、偏涝、雨涝等5种状态出现的概率分别为0.1305、0.2767、0.2751、0.2075、0.1101。宝鸡市未来多年降水过程中偏旱和正常状态占优势，而雨涝出现概率较低，反映了该区域气候暖干化的长期变化。

Prediction of Annual Precipitation Grade in Baoji City Based on Weighted Markov Chain

In order to make a comprehensive analysis on the temporal distribution regularity and variation trend,providing scientific basis for water resource prediction,prevention and mitigation of flood drought disaster,water regulation,as well as water resource management,the annual precipitation grade of Baoji were analyzed by using weighted Markov chain.Considering the uncertainty and imprecision of precipitation process,the precipitation numbers of Baoji from 1956 to 2009 were classified into 5 grades：drought,partial drought,normal,partial abundance and abundance,by using mean-standard deviation classification method.Regarding standardized self-coefficients as weights,the precipitation grades of the future years in Baoji were predicted by using weighted Markov chain.Analysis on stationary distribution showed that,the probabilities of 5 grades above were respectively 0.1305,0.2767,0.2751,0.2075,0.1101.The partial drought and normal grades dominated in the long-term precipitation process,reflecting warming-drying trend of climate changing in Baoji.