Anthropogenic emissions shape long-term changes in PM2.5 concentrations and health risks in China

Yiheng Wang, Guochao Chen, Y Yang, Zhaolei Zhang, Ruhan Zhang, Peng Wang, Hongliang ZhangEco-Environment 2025 · DOI

TL;DR

This study provides a comprehensive, long-term (2000-2020) analysis of PM2.5 sources and their associated health risks in China. It reveals that industrial emissions were the primary driver of both PM2.5 increases and subsequent decreases, and that anthropogenic sources disproportionately contribute to health risks due to their concentration in populated areas.

What Problem Does This Paper Address?

China has experienced significant shifts in air quality over the past two decades, with an initial rise in fine particulate matter (PM2.5) followed by a notable decrease due to stringent pollution control measures. However, understanding the specific contributions of different emission sources to these changes and their subsequent health impacts has been challenging. Existing studies often focus on shorter timeframes, lack detailed source apportionment, or don't fully account for dynamic factors like meteorology, population changes, and baseline mortality rates. This paper aims to fill this gap by providing a systematic, long-term evaluation of source-specific PM2.5 concentrations and their health risks across China, enabling more targeted and effective pollution control strategies.

Methods

The researchers employed a modified Community Multi-scale Air Quality (CMAQ) model (v5.0.2) integrated with the Weather Research and Forecasting (WRF) model. This setup allowed for sector-resolved source apportionment, meaning they could track the contribution of specific emission sectors (e.g., industry, residential, transportation) to PM2.5 concentrations. This was achieved by incorporating reactive tracers into the model, which helps account for complex atmospheric chemical reactions and nonlinear processes.

To assess health impacts, they used the Global Exposure Mortality Model (GEMM) to estimate premature deaths attributable to PM2.5 exposure, considering both non-communicable and lower respiratory infections. A key methodological contribution is the application of a piling-up decomposition approach. This technique systematically isolates the influence of different factors—namely, PM2.5 emissions (and their sources), meteorology, population dynamics (growth and aging), and baseline mortality rates—on changes in premature mortality over time. This allows for a nuanced understanding of policy effectiveness and the interplay of various drivers.

Datasets

  • Meteorological Data: Final (FNL) reanalysis data from the National Centers for Environmental Prediction (NCEP).
  • Anthropogenic Emissions: Multi-resolution Emission Inventory for China (MEIC v1.4) for China and Emissions Database for Global Atmospheric Research (EDGAR v5.1) for regions outside China.
  • Biogenic Emissions: Model of Emissions of Gases and Aerosols from Nature (MEGAN v2.1).
  • Population Data: Gridded population data from WorldPop (including age structure data).
  • Baseline Mortality Rates: Global Burden of Disease Study 2021 (GBD 2021).
  • Urban/Non-Urban Classification: Global Human Settlement Layer Settlement Model (GHS-SMOD) data.

The study does not introduce a new benchmark dataset but rather integrates and analyzes existing datasets over an extended period to provide a novel long-term perspective.

Key Insights and Ideas

  • Industrial Dominance: Industrial emissions are not just a major contributor to PM2.5 but were the primary driver of both the increase in PM2.5 from 2000-2012 and the subsequent decrease from 2012-2020. This highlights the critical role of industrial sector policies.
  • Disproportionate Health Risk: Anthropogenic sources consistently contribute more to health risks than their proportional share of PM2.5 suggests. This is because these emissions are concentrated in densely populated urban areas, leading to higher population exposure.
  • Policy Effectiveness Nuances: While China's two-phase air pollution control plans significantly reduced PM2.5-attributable premature deaths (20.4% reduction from 2012-2020), this benefit was partially offset (10.6%) by factors like population growth and aging. This underscores the need to consider demographic shifts in health impact assessments.
  • Regional Heterogeneity: Source contributions and health impacts vary significantly by region. For instance, residential sources are more dominant in Northern China (due to heating), while industrial sources are key in economically developed coastal regions.
  • Emerging Challenges: As PM2.5 levels decrease, the relative contributions of transportation and secondary organic aerosols (ASOA) to health risks are increasing in some regions (e.g., PRD), posing new challenges for targeted control strategies.

Key Results

  • PM2.5 Trends: China saw an increase in PM2.5 from 2000-2012, driven by anthropogenic emissions (51.8%), followed by a decrease from 2012-2020 (47.6%).
  • Current Source Contributions: Industrial sources are the largest contributor to PM2.5 (approx. 32%), followed by residential (approx. 22%) and transportation (approx. 13%).
  • Health Impact Reduction: The two-phase air pollution control measures led to an estimated 20.4% reduction in national premature mortality attributable to PM2.5 from 2012 to 2020.
  • Offsetting Factors: Population growth, aging, and declining baseline mortality rates offset approximately 10.6% of the health benefits gained from emission reductions between 2012 and 2020.
  • Urban vs. Non-Urban Exposure: Urban populations bear a disproportionately higher burden of anthropogenic PM2.5 exposure, with industrial and transportation sources accounting for over 70% of urban exposure.

Limitations and Future Work

The authors acknowledge several limitations:

  • Model Uncertainties: Inaccuracies in simulating nitrate and secondary organic aerosols (SOA) due to incomplete precursor emission inventories, model mechanisms, and parameterizations.
  • Emission Inventory Reliance: The study's reliance on the MEIC emission inventory means uncertainties in these data can affect source apportionment results.
  • Toxicity Assumptions: The study assumed equal toxicity of PM2.5 from different sources due to a lack of source-specific exposure-response relationships in China. Previous research suggests distinct toxicities, which could lead to under/overestimation of source-specific health risks.
  • Resolution of Exposure Analysis: The classification of urban/non-urban areas and the resampling of data onto a common grid limited the resolution of exposure analyses.

Future work should focus on:

  • Improving parameterizations for key processes (e.g., SOA formation) and strengthening constraints from local observations and emission inventories.
  • Conducting longitudinal cohort studies in China to establish source-specific toxicity and refine health risk assessments.
  • Developing higher-resolution models to better capture fine-scale exposure differences.
  • Developing region-specific emission reduction policies, particularly targeting industrial and transportation sources, to maximize future health benefits.

Summary updated

Abstract

China has experienced an initial increase and a subsequent decrease in fine particulate matter (PM<sub>2.5</sub>) concentrations since the early 21st century, with substantial heterogeneity across different source contributions. This study developed a source-oriented CMAQ model to construct a source-resolved PM<sub>2.5</sub> concentration database for China during 2000-2020. Subsequently, source-specific health risks and the contributions of key driving factors were systematically evaluated using the piling-up decomposition method. The results indicate that anthropogenic emissions, particularly from industrial sources, were the primary drivers of both the increase in PM<sub>2.5</sub> from 2000 to 2012 (51.8%) and the subsequent decrease from 2012 to 2020 (47.6%). Currently, industrial sources remain the largest contributor to PM<sub>2.5</sub> (about 32%), followed by residential (about 22%) and transportation sources (about 13%). Implementation of two-phase air pollution control measures led to a 20.4% reduction in national premature mortality attributable to PM<sub>2.5</sub> from 2012 to 2020, although 10.6% of this benefit was offset by changes in population and baseline mortality rates. Throughout 2000-2020, health risks associated with anthropogenic sources consistently exceeded their proportional contribution to PM<sub>2.5</sub>, primarily because these emissions are concentrated in densely populated areas. These findings underscore that, in addition to implementing region-specific emission reduction policies, maintaining stringent controls on anthropogenic emissions, particularly from industrial and transportation sources, is crucial to maximizing future health benefits.
BibTeX
@inproceedings{wang2025anthropogenic,
  author = {Yiheng Wang and Guochao Chen and Y Yang and Zhaolei Zhang and Ruhan Zhang and Peng Wang and Hongliang Zhang},
  title = {{Anthropogenic emissions shape long-term changes in PM2.5 concentrations and health risks in China}},
  booktitle = {Eco-Environment},
  year = {2025},
  doi = {10.1016/j.eehl.2025.100198},
}
Figure