基于小波变换的湖南省PM_(2.5)污染特征分析

Analysis of PM_(2.5) Pollution Characteristics in Hunan Province Based on Wavelet Transform

该文基于湖南省2017-2021年PM_(2.5)与其他5种大气污染物浓度监测数据以及气象资料,分析了PM_(2.5)浓度时空分布特征,利用连续小波变换和小波相干谱分别研究了PM_(2.5)浓度周期、PM_(2.5)与其他大气污染物及气象因子的相关性,确定了PM_(2.5)浓度的主要影响因子,结果表明:(1)湖南省PM_(2.5)浓度年平均值总体呈下降趋势,年均下降率达4.88%,PM_(2.5)浓度月平均值呈“U”型分布,2017年1月月均值浓度最高(91.9μg/m^(3)),2020年6月月均值浓度最低(15.8μg/m^(3));PM_(2.5)浓度年平均值的空间分布则呈现东北至西南高、西北及东南低的特点,PM_(2.5)浓度季平均值呈现冬季>秋季>春季>夏季的特征。(2)14个市(州)PM_(2.5)浓度都存在显著周期性变化,主要包括272 d左右的长周期和36 d左右的短周期。(3)PM_(2.5)浓度与其他大气污染物的相关性在不同时频尺度上存在差异。PM_(2.5)浓度与PM_(10)浓度在所有时频尺度上均存在显著的正相关。在小时频尺度(0~64 d)上,PM_(2.5)浓度与O_(3)、CO、NO_(2)和SO_(2)浓度都存在正相关性;在中时频尺度(64~256 d)上,PM_(2.5)浓度与O_(3)浓度的相关性不明显,与CO、NO_(2)和SO_(2)浓度存在正相关;在大时频尺度(256~512 d)上,PM_(2.5)浓度与O_(3)浓度呈现负相关,与NO_(2)和SO_(2)浓度呈现正相关,PM_(2.5)浓度与CO浓度在长沙和衡阳呈现正相关,在怀化呈现负相关。(4)PM_(2.5)浓度与气象因子的相关性在不同时频尺度上也存在差异。在小时频尺度上,PM_(2.5)与降水量、相对湿度和气压存在负相关,与温度和日照时数存在正相关,与风速存在正相关也存在负相关;在中时频尺度上,PM_(2.5)与各气象因子的相关性均不明显;在大时频尺度上,PM_(2.5)与降水、温度和日照时数呈现负相关,PM_(2.5)与气压呈现正相关,PM_(2.5)与相对湿度和风速的相关性不明显。

Based on the monitoring data of PM_(2.5)and other five atmospheric pollutants and meteorological data in Hunan Province from 2017 to 2021,the spatial and temporal distribution characteristics of PM_(2.5)concentration were analyzed.The continuous wavelet transform and wavelet coherence spectrum were used to study the PM_(2.5)concentration cycle and the corre⁃lation between PM_(2.5)and other atmospheric pollutants and meteorological factors.The main influencing factors of PM_(2.5)concentration were determined.The results showed that the annual average PM_(2.5)concentration in Hunan Province showed a downward trend,with an average annual decline rate of 4.88%.The monthly average PM_(2.5)concentration showed a“U”type distribution.The highest concentration was in January 2017(91.9μg/m^(3)),and the lowest concentration was in June 2020(15.8μg/m^(3)).The spatial distribution of the annual average of PM_(2.5)concentration shows the characteristics of high from northeast to southwest,low from northwest to southeast,and the seasonal average of PM_(2.5)concentration shows the characteris⁃tics of winter>autumn>spring>summer.There are significant periodic changes in PM_(2.5)concentration in 14 cities(states),mainly including a long period of about 272 days and a short period of about 36 days.The correlation between PM_(2.5)concentra⁃tion and other air pollutants is different at different time-frequency scales.There is a significant positive correlation between PM_(2.5)concentration and PM10 concentration at all time-frequency scales.On the hourly scale(0~64 d),PM_(2.5)concentration was positively correlated with O_(3),CO,NO_(2) and SO_(2) concentrations.On the medium time-frequency scale(64~256 d),the correla⁃tion between PM_(2.5)concentration and O_(3) concentration is not obvious,and there is a positive correlation with CO,NO_(2) and SO_(2) concentration.On the large time-frequency scale(256~512 d),PM_(2.5)concentration was negatively correlated with O_(3) concentration,and positively correlated with NO_(2) and SO_(2) concentration.PM_(2.5)concentration was positively correlated with CO concentration in Changsha and Hengyang,and negatively correlated in Huaihua.The correlation between PM_(2.5)concentra⁃tion and meteorological factors is also different at different time-frequency scales.On the hourly frequency scale,PM_(2.5)is negatively correlated with precipitation,relative humidity and air pressure,positively correlated with temperature and sunshine hours,and positively correlated with wind speed.On the medium-time-frequency scale,the correlation between PM_(2.5)and various meteorological factors is not obvious;on the large time-frequency scale,PM_(2.5)is negatively correlated with precipita⁃tion,temperature and sunshine hours,PM_(2.5)is positively correlated with air pressure,and the correlation between PM_(2.5)and rel⁃ative humidity and wind speed is not obvious.