加权马尔可夫链在银川地区降雨量预测中的应用

Prediction of Precipitation Based on the Weighted Markov Chain in Yinchuan Area

依据1960年-2008年银川地区的年降雨量资料,采用均值-标准差分级法,对其进行状态分级,分为枯水年、偏枯年、平水年、偏丰年和丰水年5个状态。根据马尔可夫链预测方法,验证了该降水序列满足马氏性,并以规范化的各阶自相关系数为权,建立了适用于银川地区年降雨量的加权马尔可夫链预测模型,实例验证结果令人满意。在此基础上结合模糊集理论中的级别特征值,预测了银川地区2009年及2010年的年降雨量。预测结果表明:银川地区2009年为偏枯年,年降雨量为150.45mm;2010年为平水年,年降雨量为177.38mm。最后采用平稳分布估计年降雨量各状态的重现期,结果显示银川地区年降雨量处于平水年和偏枯年的可能性较大。

Based on the annual precipitation data from 1960 to 2008 in Yinchuan area, taking the means and standard deviation of information series as the classification standard of precipitation, the past 49 years are classified into five states： drought year, weak drought year, normal year, weak water logging year and water logging year. After the verification of the Markov property of precipitation, the weighted Markov chain model used for predicting the state of precipitation in Yinchuan area was established by taking standardized auto-correlation coefficient as weight coefficient, and the prediction results are satisfactory. The concrete value of precipitation in 2009 and 2010 in Yinchuan area is predicted with the level characteristics value of fuzzy sets. The results show that： the state of Yinchuan area is weak drought year and the precipitation is 150. 45mm in 2009; the state is normal year and the precipitation is 177. 38mm in 2010, and the possibility of normal year and weak drought year is the larger according to the stationary distribution.