摘要
运用计算多分形谱的多分形非趋势波动分析法,研究全球、北半球和中国月平均温度距平的多分形特征。全球气温、北半球气温和中国气温的变化是自相似的多分形;它们都表现出一定的正长程相关性,全球气温的长程相关性最强,北半球次之,中国气温相对最弱;中国气温的最强涨落较北半球气温大,而全球气温的最强涨落最小。
The multifractal detrended fluctuation analysis method for calculating multifractal singularity spectra is introduced in detail and applied to the study of the multifractal characteristics of the global temperature as well as the temperature anomalies of Northern Hemisphere and China. Temperatures of the global, Northern Hemisphere and China are self-similar multifractals; The positive long-range dependence and behaviors between 1/f noise and brown noise are observed in the three series. And the long-range dependence of the global series is the strongest and the China series is weaker than Northern Hemisphere. In general, the temperature variation of China presents the strongest singularity and the most intensive fluctuation among the three temperature series, and the global temperature shows the strongest regularity and smallest intensive fluctuations.
出处
《热带气象学报》
CSCD
北大核心
2004年第6期673-678,共6页
Journal of Tropical Meteorology
基金
"十五"国家科技攻关计划"奥运气象保障技术研究"(AY2003A-01-01021)
国家气象中心自建项目(ZK2002ASZ-07)资助
关键词
多分形
非趋势波动分析
长程相关性
广义HURST指数
multifractal
detrended fluctuation analysis
long range dependence
temperature series
generalized Hurst exponent