摘要
利用有限长度的气温序列变化资料估计其无限长时序的气温过程的统计特征,对于中长期气象预报以及全球的气候变化都是十分有意义的基础工作,所以对气温过程开展遍历性(各态历经性)分析是一个值得探索的课题.为此,基于模糊粗糙聚类(fuzzy-rough C-means clustering method,FRCM)、自相关趋势图、ADF(advanced Dickey and Fuller)检验以及模糊最小二乘回归提出一种具有普适性特点的遍历性分析方法,并以上海1873—1997年的气温序列为例,进行了具体的计算和分析.结果表明:上海7月份、8月份的最高气温和最低气温变化是均值遍历和协方差遍历的,因此其气温变化过程具有遍历特征.从长时间尺度来看,上海7月份、8月份最高气温和最低气温总体上不会长期呈上升趋势,它们的变化会围绕其均值波动.最后,对上述结论做了印证性分析,对上海气温变化过程进行了挖掘性分析,从遍历性的角度佐证了前人的一些研究成果.
Estimation of statistic properties of infinite air temperature process with finite temperature materials is significant and fundamental to mid-term and long term climate forecast and global climate change studies. Therefore,ergodicity analysis of air temperature is an important research topic. In this paper, a commonly applicable ergodic property analysis model based on fuzzy-rough C-means clustering method （FRCM）,autocorrelogram,advanced Dickey and Fuller（ADF） and fuzzy least square regression （FLSR） has been proposed. The air temperature time series （1873--1997） from Shanghai hydrology station has been calculated and analyzed according to the proposed model. Results show that month-highest and month-lowest air temperatures in July and August have not only mean ergodic properties but also covariance ergodic properties. Thus the air temperature variation process is ergodic. In the long run,month-highest and month-lowest air temperatures in July and Au- gust will not keep on rising. Instead,they will fluctuate around the mean value. In the last section,the above conclusion has been confirmed by further analysis and temperature variations in Shanghai are further discussed,which corroborates some previous studies from the point of ergodicity.
出处
《天津大学学报》
EI
CAS
CSCD
北大核心
2010年第1期55-63,共9页
Journal of Tianjin University(Science and Technology)
基金
国家科技支撑计划资助项目(2006BAB04A08)
关键词
上海气温
遍历性
模糊粗糙聚类
平稳性检验
模糊最小二乘回归
air temperature of Shanghai
ergodicity
fuzzy-rough C-means clustering method （FRCM）
stationary test
fuzzy least square regression （FLSR）
