Sci Total Environ. 2023 Oct 29:168185. doi: 10.1016/j.scitotenv.2023.168185. Online ahead of print.
Being one of the most serious biomass burning regions in the world, the air pollution caused by spring combustion in the Indo-China Peninsula (ICP) has already had an impact on Yunnan Province’s beautiful environment and excellent air quality to some extent. In this study, considering the differences in geographical location and topography of Yunnan, we used the K-Means algorithm to divide it into five clustering zones according to the spatiotemporal variation characteristics of PM2.5. Then this study explored the spatial and temporal characteristics of pollution in Yunnan Province and biomass combustion in ICP based on the multi-source data such as MOD14A1, GDAS1, and ground-based PM2.5 data, and used HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) pollution tracer analysis and other data statistical methods. The results show that the spatiotemporal variation characteristics of PM2.5 in Yunnan Province show large differences within each clustering zone (CZ). Spatially, CZ 2 has better air quality throughout the year, and the areas with higher PM2.5 are mainly in CZ 1 and CZ 3. Temporally, the months with higher concentration values were mainly from February to April, and also this period owed high biomass burning activities in the ICP, which resulted in pollution values exceeding 60 μg/m3 within certain CZs. Finally, the results of the pollution tracer analysis showed that within CZs other than CZ 2, the contribution due to the burning in the ICP was variable, and that the countries with a high contribution of pollution to Yunnan Province were Myanmar, and the other sources of pollution are mainly caused by local and neighbouring anthropogenic activities. Therefore, based on overall improvement of air quality, Yunnan Province is necessary to prevent and control not only the pollutants from the ICP from February to April, but also the pollution caused by the emissions from rapid economic development.