冬天极端低温回归时间的长程相关性

Long-Range Correlation of Return Time of Extreme Low Temperature in Winter

利用去趋势波动分析法研究了全国176个站点的1960—2005年冬天日最低温度序列的长程相关性和对应的极端低温回归时间序列的长程相关性.结果表明,这176个站点的回归时间序列的长程相关性标度指数在0.42~0.72之间,冬天日最低温度序列的长程相关性标度指数在0.55~0.8之间,而且回归时间序列的长程相关性比冬天日最低温度序列的长程相关性弱,在全国这2种序列的长程相关性的强弱分布大致相同,回归时间序列的长程相关性与冬天日最低温度序列的长程相关性紧密相关.

It uses the detrended fluctuation analysis to study the long-range correlation of winter daily minimum temperature series and the corresponding return time series of the extreme low temperature from 1960 to 2005 of 176 sites,and reaches this conclusion：In 176 sites,the scaling exponents of long-range correlation of return time series are between 0.42 and 0.72,and the scaling exponents of long-range correlation of winter daily minimum temperature series are between 0.55 and 0.8.The long-range correlation of return time series is weaker than correlation of winter daily minimum temperature series,and in China,both two have the same strength distribution,the long-range correlation of return time series with the correlation of winter daily minimum temperature series are closely related.