High concentrations of particulate matter (PM2.5), particularly carbonaceous aerosols like black carbon (BC) and organic carbon (OC), pose significant threats to human health and environmental sustainability in East Asia, contributing substantially to climate warming. China's implementation of clean air actions between 2013 and 2017 led to a considerable decrease in annual mean PM2.5 concentrations. However, the impact on the source contributions of carbonaceous aerosols remained unclear, especially concerning the persistent winter haze events. Shanghai, a megacity with a population of approximately 25 million, experiences frequent severe pollution episodes during winter, making it a crucial case study to understand the evolving sources of air pollution and inform effective mitigation strategies. This research aims to quantify the sources and geographical origins of BC in Shanghai, focusing on the changes in source contributions after the implementation of clean air policies. Understanding these sources is critical for achieving the United Nations' Sustainable Development Goals (SDGs) related to air quality, climate change, and ecological sustainability, particularly in light of Shanghai's aim to become a model for sustainable city development. The study utilizes a combination of dual-isotope analysis (δ¹³C and Δ¹⁴C) to identify the sources of BC and a chemical transport model (FLEXPART) to determine the geographical origin of the aerosols. This combined approach provides a comprehensive understanding of the changing nature of air pollution in Shanghai.

Previous studies have documented the detrimental health impacts of high PM2.5 concentrations in Chinese megacities and the overall decline in PM2.5 levels following the implementation of clean air actions. However, the specific changes in the source contributions of carbonaceous aerosols, particularly BC, remained unclear. Existing research has utilized various methods to characterize BC sources in the region, including chemical transport models and isotopic analysis of BC. Isotopic techniques, especially the use of ¹⁴C and ¹³C, allow for a more precise quantification of the relative contributions of fossil fuel and biomass burning sources to BC. However, such studies have often been limited in scope or temporal coverage, making it difficult to fully understand the long-term trends and seasonal variations in BC sources. Prior work has highlighted the importance of residential biomass burning as a significant source of air pollution in many parts of China, but its specific contribution to megacity haze, particularly after recent emission control measures, was not fully understood. This research addresses this gap by combining advanced observational techniques with detailed modeling to comprehensively quantify the shifting contributions of different sources to BC pollution in Shanghai.

The study employed a multi-faceted approach combining field observations and model simulations. A year-round PM2.5 sampling campaign was conducted at two locations in Shanghai: a suburban site on Chongming Island (YRE) and an urban site. Samples were collected on quartz-fiber filters using high-volume samplers equipped with PM2.5 impactors. PM2.5 concentrations were determined via gravimetric analysis. A subset of samples was analyzed for organic carbon (OC) and BC (elemental carbon, EC) concentrations using a thermal/optical carbon analyzer, following the IMPROVE_A protocol. Water-soluble organic carbon (WSOC) was also extracted and analyzed. A significant fraction of selected samples was further analyzed for δ¹³C and Δ¹⁴C isotopes using accelerator mass spectrometry (AMS) and isotope ratio mass spectrometry (IRMS). The δ¹³C and Δ¹⁴C values provided source apportionment information, distinguishing between biomass burning and fossil fuel sources. A Monte Carlo (MC) approach was used to incorporate the variability in isotopic measurements and source endmembers, providing a more robust quantification of the relative contribution from different fuel types (biomass, coal, liquid fossil). The Lagrangian atmospheric transport model FLEXPART (version 10.4) was coupled with emission inventories from ECLIPSE (version 6B) and GFED (version 4.1) to simulate BC concentrations and source contributions at the YRE site. The FLEXPART model included meteorological data from ECMWF and accounted for various atmospheric processes such as dry and wet deposition. Model simulations were compared to observational data to evaluate the model's performance and to further constrain the source apportionment.

The study revealed a strong seasonal variation in BC concentrations, with significantly higher levels observed during winter compared to summer. The highest BC concentrations were observed in winter, which coincided with the most pronounced haze events. Isotopic analysis showed that the fraction of BC originating from biomass burning (fbb) was substantially higher in winter (around 45%) compared to summer (around 31%). This finding represents a notable increase in the biomass burning contribution to winter haze in Shanghai relative to earlier studies (2013-2014), indicating a shift in the dominant pollution sources. The Monte Carlo (MC) simulations further quantified the sources, showing a higher contribution of biomass burning during winter compared to other seasons and a consistent underestimation of biomass burning in earlier studies. The model results (FLEXPART) showed good agreement with observations, capturing the overall seasonal trend and magnitude of BC concentrations. The model also identified central-east China as the primary geographical source region for wintertime BC pollution in Shanghai, with significant contributions from provinces like Jiangsu, Anhui, Henan, and Shandong. Sectoral analysis indicated a much larger contribution of residential emissions during winter compared to summer. These residential emissions were identified as originating predominantly from central-east China. The residential sector was also found to be the most dominant BC source during winter in Shanghai, comprising roughly 50% of total emissions. In comparison, other sectors like transport and industry were also significant contributors, but the influence of the residential sector was considerably greater, especially during peak pollution events.

The findings of this study demonstrate a significant shift in the relative contribution of biomass burning to wintertime haze in Shanghai since the implementation of China's clean air actions. While previous efforts focused on reducing emissions from industry and transportation, the results highlight the crucial and escalating role of residential biomass burning, primarily from the central-east region of China. This underscores the need for policies to tackle small-scale, scattered residential emissions, which are currently underestimated in many emission inventories. The increased reliance on biomass burning may be a consequence of the prioritization of industrial and transportation emission controls, leading to a relative increase in the share of residential sources. The close agreement between the observational and modeling results provides strong support for the findings, suggesting an accurate quantification of the evolving source contributions. The trans-provincial transport of pollutants from central-east China to Shanghai underscores the need for regional cooperation in addressing air pollution issues.

This research provides compelling evidence of the growing contribution of residential biomass burning to wintertime haze in Shanghai. The findings suggest a need for more comprehensive emission inventories, incorporating more accurate estimates of small-scale residential emissions. Effective mitigation strategies require a combined approach, focusing on both local and regional emission reductions. Future research should explore targeted strategies to reduce residential biomass burning, such as promoting cleaner cooking and heating technologies, while also continuing efforts to reduce emissions from transportation and industry.

The study primarily focused on Shanghai, limiting the generalizability of the findings to other regions. The model used (FLEXPART) has inherent uncertainties related to emission inventories and meteorological data, which could influence the accuracy of the source apportionment. The sampling period was limited to one year, which might not capture long-term trends completely. Finally, the isotope analysis was conducted on a selected subset of the samples, potentially not representing the full range of BC pollution events.