摘要
为了解PM2.5浓度随时间演变的动力学特征,采用多重消除趋势波动分析法(MFDFA),对上海市春夏秋冬四季的PM2.5日浓度数据时间序列进行了分析,结果表明:上海市四季的PM2.5日时间序列在整个时间尺度上均表现出正长程相关性,且存在多重分形特征,且夏季多重分性特征最强,冬季最弱;上海市四季的PM2.5时间序列的多重分形特征均是波动的胖尾分布和自身的长程相关性共同作用的结果:对于春夏两季,序列本身的长程相关性对多重分形特征的影响比较大;对于秋冬两季,波动的胖尾分布对多重分形特征的贡献比较大。MF-DFA可以有效地分析PM2.5时间序列的标度不变性和多重分形特征,对于描述时间序列的动力学演变特征具有非常重要的意义,是一种有效的大气污染物时间序列的非线性研究方法。
It analyzed the seasonal PM2.5 detrended fluctuation analysis method. The show positive long-range correlation multifractal natures of the seasonal time series of Shanghai by using mumlractm ana result show that： 1） the seasonal PM2.5 time series and multifractal nature on the whole time scale; 2） The PM2.5 time series of Shanghai are due to both broad probability density function and long-range correlation： Broad probability density plays a more important role in the multifractal nature of the daily PM2.5 time series of autumn and winter, while muhifractal natures of the daily PM2.5 time series of spring and summer are mainly attributed to the long-range correlation. MF-DFA can identify the scaling invariance and multifractal characteristics of time series, which has practical significance for describing the dynamics of the air pollutant time series and provides an effective way for nonlinear study of atmospheric pollutants.
作者
卢凯丽
沈可
王博
LU Kai-li Shen Ke Wang bo(School of Mathematics and Physics, China University of Geosciences, Wuhan 430074, Chin)
出处
《能源环境保护》
2017年第5期1-7,共7页
Energy Environmental Protection
