Sustained PM2.5 decline in Shenzhen confronts emerging challenges: Strengthening regional governance and secondary aerosol mitigation

Abstract

Long-term observation at different environmental sites is crucial for accurately tracking PM_2.5 trends and formulating effective pollution control strategies. This study conducted PM_2.5 sampling and component analysis at urban and suburban sites in Shenzhen in 2014, 2019, and 2024, using machine learning models to assess the impact of meteorological and source emission changes on PM_2.5 trends and receptor model to identify PM_2.5 sources. The results show that Shenzhen's PM_2.5 levels declined by 47 % from 2014 to 2024, mostly due to long-term emission cuts, with meteorological impacts accounting for a minor 11.9 % of the change. Urban PM_2.5 concentrations consistently surpass suburban ones, with both experiencing approximately 30 % declines in Phase I (2014–2019), whereas in Phase II (2019–2024), the urban site shows a greater 29 % drop compared with a 17 % drop at the suburban site, narrowing their gap. The source apportionment results show that in Phase I, the urban site's PM_2.5 decline stemmed mainly from reduced vehicle emissions and secondary sulfate, while the suburban site's drop resulted from decreased secondary sulfate, ship emissions, and secondary organic aerosols. In Phase II, fugitive dust reduction is the primary driver of PM_2.5 declines at both sites, though other sources also play a contributing role. Emission reductions from sources in Phase II are smaller than those in Phase I, coupled with a narrowed concentration gap between the two sites and a rising relative contribution from secondary sources, highlighting Shenzhen's need for stronger regional coordination and secondary pollutant control to lower PM_2.5 levels.