Abstract:

To investigate the recent changes in NO₂ column concentrations and their natural dominant factors in the Guangdong-Hong Kong-Macao Greater Bay Area (GBA), this study utilized near real-time tropospheric NO₂ column concentration data (NRTI/L3_NO₂) provided by the Sentinel-5P atmospheric monitoring satellite, along with meteorological data, NDVI data, and land surface temperature data. Methods such as Sen’s trend analysis, Mann-Kendall tests, geographical detector models, and the Geographically and Temporally Weighted Regression (GTWR) model were employed to analyze the spatiotemporal variations and natural driving mechanisms of NO₂ column concentrations in GBA from 2019 to 2023. The main findings are as follows: 1) At the annual scale, NO₂ column concentrations peaked in 2021 and were lowest in 2020. Seasonally, the concentrations were the highest in winter and the lowest in summer. Spatially, a “high in the center and low in the periphery was observed. 2) Sen’s trend analysis indicated an increasing trend in NO₂ concentrations in areas such as the Guangzhou-Foshan border, western Shenzhen, and Zhaoqing, while decreasing trends were observed in Zhuhai, Jiangmen, Macao, and parts of Zhongshan. The Mann-Kendall test results showed that significant increases were mainly concentrated in northern Guangzhou and Zhaoqing, while other areas exhibited minimal changes, although seasonal differences were pronounced. 3) The geographical detector analysis revealed that wind speed, temperature, humidity, and air pressure were the primary explanatory factors, whereas precipitation and solar radiation had weaker effects. The interaction between humidity and wind speed, as well as between humidity and temperature, was the most significant, with nonlinear enhancement effects primarily reflected in interactions involving air pressure, precipitation, and other factors. 4) The GTWR model analysis showed that wind speed, temperature, and land surface temperature had positive impacts on NO₂ concentrations, particularly in Guangzhou-Foshan and Shenzhen. Conversely, air pressure, humidity, and NDVI have negative effects, particularly in Jiangmen and Zhuhai. The effects of precipitation and solar radiation are more complex, with significant spatial variations. This study provides insights into the natural driving mechanisms of NO₂ pollution and its spatiotemporal variations in the GBA over the past five years, offering valuable support for the formulation of targeted air quality management and pollution control strategies.

Key words: NO₂ column concentration, temporal and spatial change characteristics, Google Earth Engine (GEE Cloud platform), Influencing factors, Guangdong-Hong Kong-Macao Greater Bay Area

摘要： 基于Sentinel-5P卫星提供的二氧化氮对流层柱浓度数据（NRTI/L3_NO₂），结合气象数据、NDVI和陆表温度数据，采用Sen趋势分析、Mann-Kendall检验等方法，并辅以地理探测器与时空地理加权回归模型（GTWR），解析2019－2023年粤港澳大湾区NO₂柱浓度时空变化与自然驱动机制。结果显示：1）年际变化上，2021年NO₂柱浓度达到峰值，2020年为最低，季节性变化上冬季浓度最高，夏季最低，空间分布呈“中间高、四周低”的特点；2）Sen年趋势分析表明，广佛交界、深圳西部、肇庆等地NO₂浓度上升，珠海、江门、澳门等地下降；Mann-Kendall检验显示，广州北部与肇庆为显著增长区；3）地理探测器分析表明，风速、温度、湿度和气压是主要影响因子，降水和太阳辐射影响较弱；湿度与风速、湿度与温度的交互作用显著，非线性增强效应表现在气压、降水与其他因子的交互中；4）GTWR模型分析显示，风速、温度和陆表温度对NO₂浓度存在正向影响，广佛与深圳尤为显著；气压、湿度与植被指数对其存在负向影响，江门与珠海更为明显；降水与太阳辐射的影响复杂，空间差异较大。该研究可为理解大湾区NO₂污染的时空变化及自然驱动机制提供参考，助力空气质量管理和污染控制策略的制定。

关键词: NO₂柱浓度, 时空变化特征, 谷歌地球引擎（GEE云平台）, 影响因素, 粤港澳大湾区