摘要
依据济南市1960—2014年的降水量资料,采用均方差分级法对年降水量进行分级,将其分为枯水年、偏枯年、平水年、偏丰年、丰水年5个状态,同时验证此序列满足马尔科夫要求。在此基础上,以规范化的各阶自相关系数为权重,建立了适用于济南市年降水量的加权马尔科夫链预测模型,并根据模糊集理论中的级别特征值对降水量进行了预测。结果表明:12015年、2016年济南市的降水量状态均为平水年,年降水量分别为691.21、645.28 mm;2济南市年降水量处于偏枯年及平水年状态的可能性比较大,其重现期分别为5.38、2.34 a。
According to the rainfall data of Jinan City from 1960 to 2014, the paper used mean square error deviation method to classify the rainfall and divided it into five states of diy year, ＂weak dry year, normal year, ＂weak w,et year and w,et year and verified this sequence tosatisfy the requirements of Markov. On this basis, by using standardized order autocorrelation coefficients as weights, the prediction model was established that was applicable to the weighted Markov chain of Jinan City annual rainfall. The results show that the rainfall in 2015 and2016 in Jinan are normal years and the annual precipitation is 691.21 mm and 645.28 mm respectively. The possibility of normal year andweak dry year is larger in Jinan and the return period is 5.38 years and 2.34 years respectively.
出处
《人民黄河》
CAS
北大核心
2016年第9期13-16,共4页
Yellow River
基金
高等学校学科创新引智计划项目(B08039)
中国工程院咨询研究项目(2014-07-XZ-002)
