基于小波变换的兰州市近十年空气污染指数变化

Waveletanalysis of air pollution index changes in Lanzhou during the last decade

对兰州市近10年逐日空气污染指数的时间序列采用一维连续Morlet小波进行小波变换分析,研究了该市大气污染时间序列的多尺度变化特征、主周期和影响因素.结果表明:兰州市近10年空气污染指数在不同时间尺度上具有不同的"高-低"交替演化规律,且以300d左右的变化为主周期,100d左右的变化为次周期;受地形条件的影响,大气污染呈现"冬重夏轻"的格局,且受河西走廊沙尘活动的影响,往往伴随春季污染次高峰的发生,春分和秋分前后是1年中大气污染轻重状况转换的拐点;兰州市近10年大气污染状况总体趋向好转,但局部时段污染加重的现象亦时有发生.小波变换分析对于研究空气污染指数时间序列的变化规律十分有效,也适用于其他污染物时间演变规律的研究.

To demonstrate the multiscale variations,primary period and influencing factors in Lanzhou,we analyzed the time series of daily air pollution index during the last decade using the continuous Morlet wavelet transformation.The results showed that the air pollution index varied at diverse time-scales with periodic ＂high-low＂ fluctuations.The primary period of the daily variations was around 300 days and the secondary period was around 100 days.Due to topographical conditions,the air pollution was serious in winter and light in summer in most scales.A secondary peak often occurred in spring which was affected by the sand-dust of the Gansu Corridor,the inflection points of serious to light pollution index each year were the vernal and autumnal equinoxes.The general status of air pollution tended to be better during the last 10 years in Lanzhou.However,the serious air pollution sometimes worsened with the rapid economic development.Therefore,wavelet analysis is an effective method to study the variation of time series of air pollution index,and the multiscale variations of other pollutants.