2018年中国城市PM_2.5时空分布规律及防治策略
Spatiotemporal Distribution of PM_2.5 in Chinese Cities and Prevention Strategies in 2018

作为雾霾污染的核心构成要素，探究PM_2.5的时空分布演变特征及其影响因素，对于有效遏制其浓度增长趋势具有至关重要的意义。本研究依托来自334个地级及以上城市的PM_2.5浓度数据指标，并融入GIS空间分析技术，解析PM_2.5浓度的时空分布特性。结果表明：1) 时间上，2018年，PM_2.5的月均浓度变化趋势呈现出先下降后上升的态势，呈“U”型曲线变化；在季节分布上，2018年PM_2.5的季均浓度表现为夏季降至最低，冬季则升至最高，春秋两季则居中分布。2) 空间上，2018年中国城市华北地区、西北地区PM_2.5浓度值在全国呈高值状态，其中新疆省PM_2.5浓度值最高，PM_2.5浓度最高的城市为和田地区(102.56 μg/m³)；空间尺度上，2018年大多数监测站点的PM_2.5浓度展现出了清晰的季节性波动趋势。具体而言，每年的PM_2.5浓度从冬季开始逐渐降低，直至夏季降至最低点；随后，从夏季到冬季，浓度又逐渐回升，最终在冬季达到峰值。结合2018年334个地级及以上城市的PM_2.5浓度数据及社会经济数据，建立多元线性回归模型，得出2018年中国各城市年均人口、地区生产总值、供气总量与2018年PM_2.5浓度呈相关性。并基于结果研究分析，从实际情况出发，提出大气污染防治的可行性建议。
As the core component of haze pollution, it is of great significance to explore the spatial and temporal distribution evolution characteristics and influencing factors of PM_2.5 to effectively curb its concentration growth trend. In this study, we relied on PM_2.5 concentration data from 334 prefecture-level cities and integrated GIS spatial analysis technology to analyze the spatial and temporal distribution characteristics of PM_2.5 concentration. The results showed that: 1) In 2018, the monthly average concentration of PM_2.5 showed a trend of first decreasing and then rising, showing a “U” curve. In 2018, the seasonal average concentration of PM_2.5 decreased to the lowest in summer, rose to the highest in winter, and was in the middle spring and autumn. 2) Spatially, in 2018, the PM_2.5 concentration values in North China and Northwest China were high in China, among which Xinjiang Province had the highest PM_2.5 concentration and the highest PM_2.5 concentration in Hotan (102.56 μg/m³). On a spatial scale, PM_2.5 concentrations at most of the monitored sites in 2018 showed a clear seasonal fluctuation trend. Specifically, annual PM_2.5 concentrations gradually decrease from winter to summer to their lowest point in summer, and then gradually rise from summer to winter, eventually peaking in winter. Combined with the PM_2.5 concentration data and socio-economic data of 334 cities at and above the prefecture level in 2018, a multiple linear regression model was established to show that the average annual population, gross regional