郑州11—2月雨量时间序列变异性诊断和灰预测  

Variability Diagnosing and Grey Forecasting of Cumulative Rainfall Time Series from November to February in Zhengzhou

作　　者：王振亚[1,2] 

机构地区：[1]河南省气象台,河南郑州450003 [2]中国气象局河南省农业气象保障与应用技术重点实验室,河南郑州450003

出　　处：《人民黄河》2013年第3期24-26,29,共4页Yellow River

摘　　要：选取郑州站1952—2008年11—2月累计雨量时间序列,采用逐时段滑动分割比较序列方法,分析比较序列的差异信息测度及其增量值。结果表明:郑州站该时段雨量在1961年、1970年、1984年、1999年、2008年发生变异,并且这5年的差异信息测度增量值越来越大,表明时间序列在这5个时间点变异幅度越来越大;以1952—2001年11—2月累计雨量时间序列为样本,采用灰关联识别模式构建比较序列,预测2002—2008年11—2月累计雨量,确定灰周期参数为22,7年中有6年的预测值相对误差符合允许误差,说明灰预测用于郑州站雨量中长期预测的效果较好。The cumulative rainfall time series from November to February in the years 1952 -2008 were selected as the research objective in Zheng-zhou Station. Difference information measure and its increment were analyzed and compared by means of sliding split-by time. The results show that variability of rainfall time series occurs in the year of 1961,1970,1984,1999 and 2008. The increment of difference information measure keeps in-creasing and this shows that the time series variation increases more powerfully in the five years. The cumulative rainfall time series from November to February between 2002 and 2008 is forecasted by means of grey forecasting based on the cumulative rainfall time series from November to Febru- ary between 1952 and 2001. Grey cycle is determined as 22 and there is 6 years whose error is less than the allowable error in 7 forecasting years. Research results show that grey forecasting achieves good results in the long-term rainfall forecast in Zhengzhou Station.

关 键 词：变异诊断 灰关联分析 灰预测 降雨量 郑州 

分 类 号：TV125[水利工程—水文学及水资源] P426.623[天文地球—大气科学及气象学]