华中地区PM_2.5时空分布特征及气象驱动因子分析

Spatial and Temporal Distribution Characteristics of PM_2.5 in Central China and Analysis of Meteorological Driving Factors

摘要: 在我国城市化进程提速、大气污染问题愈发突出的背景下，PM_2.5作为核心大气污染物，其污染防控已成为生态环境治理的关键任务。从PM_2.5浓度CHAP栅格数据提取华中地区2019~2023年PM_2.5年均浓度数据，利用空间自相关、普通最小二乘法回归、地理探测器等方法研究PM_2.5时空分布特征及气象驱动因子特征。结果表明：1) 2019~2023年华中地区PM_2.5年均浓度整体呈现“先持续下降、后小幅反弹”，总体下降的特征，月均浓度呈现出显著的周期性波动特征，峰值浓度多集中在1~2月，6~8月为全年浓度低谷，四季浓度始终保持“冬季最高、夏季最低”的特征。2) 2019~2023年华中地区PM_2.5浓度存在稳定的空间自相关性。局部空间呈现出清晰的“北高南低”分异格局，河南省北部大部分地区在五年间均呈现高–高集聚区域，低–低集聚区域主要分布于湖南南部和湖北西部。3) 风速和气温对PM_2.5浓度具有正向驱动作用，降水量和相对湿度对PM_2.5浓度具有抑制作用，2019~2023年间气象因子对PM_2.5浓度平均解释力为相对湿度(0.726) > 降水量(0.552) > 风速(0.443) > 气温(0.417)。2020年前降水量为主导驱动因子，2020年后相对湿度为主导驱动因子，其中气温和风速的交互作用对PM_2.5浓度的影响最强。

Abstract: Against the backdrop of accelerated urbanization and increasingly prominent air pollution issues in China, PM_2.5, as a key air pollutant, its pollution prevention and control has become a critical task in ecological and environmental governance. Based on the CHAP raster data of PM_2.5 concentration, annual average PM_2.5 concentration data in Central China from 2019 to 2023 were extracted. Spatial autocorrelation, ordinary least squares regression, geographic detector, and other methods were used to investigate the spatiotemporal distribution characteristics of PM_2.5 and its meteorological driving factors. The results show that: 1) From 2019 to 2023, the annual average PM_2.5 concentration in Central China generally presented a trend of “continuous decline first, followed by a slight rebound” with an overall downward tendency. Monthly average concentration showed significant periodic fluctuations, with peak concentrations mostly concentrated in January and February, and the lowest concentrations from June to August. Seasonal concentrations consistently maintained the characteristic of “highest in winter and lowest in summer”. 2) Stable spatial autocorrelation of PM_2.5 concentration existed in Central China from 2019 to 2023. Local spatial differentiation featured a clear “high in the north and low in the south” pattern. High-high (H-H) clusters were mainly distributed in most areas of northern Henan Province during the five years, while low-low (L-L) clusters were primarily located in southern Hunan and western Hubei. 3) Wind speed and air temperature exerted positive driving effects on PM_2.5 concentration, whereas precipitation and relative humidity had inhibitory effects. During 2019~2023, the average explanatory power of meteorological factors on PM_2.5 concentration ranked as: relative humidity (0.726) > precipitation (0.552) > wind speed (0.443) > air temperature (0.417). Precipitation was the dominant driving factor before 2020, while relative humidity took the leading role after 2020. Among interactive effects, the combination of air temperature and wind speed imposed the strongest impact on PM_2.5 concentration.