Spatio–Temporal Relationship and Evolvement of Socioeconomic Factors and PM2.5 in China During 1998–2016

Yi Yang, Jie Li, Guobin Zhu, Qiangqiang YuanInternational Journal of Environmental Research and Public Health 2019 · DOI

TL;DR

This study reveals a complex, evolving relationship between China's rapid economic development and its fine particulate matter (PM2.5) pollution from 1998-2016. While economic growth generally correlates with increased PM2.5, the study highlights regional differences and the growing importance of spatial spillover effects, suggesting that localized environmental policies need to consider inter-provincial impacts for effective sustainable development.

What Problem Does This Paper Address?

Air pollution, particularly from fine particulate matter (PM2.5), is a significant environmental and health concern in China, largely driven by its rapid economic growth over the past few decades. While the link between economic development and pollution is intuitive, understanding the nuanced spatio-temporal dynamics and the specific contributions of various socioeconomic factors is crucial for effective environmental management and sustainable development strategies. This paper aims to disentangle these complex relationships across China's provinces over a substantial time period, moving beyond simple correlations to explore spatial dependencies and spillover effects.

Methods

The researchers employed a multi-faceted approach to analyze the spatio-temporal relationships:

  • Unary Linear Regression: Used to determine the temporal trend (slope) of PM2.5 and socioeconomic factors over time.
  • Spearman's Rank Correlation Analysis: Applied to assess the traditional statistical correlation between PM2.5 and socioeconomic factors.
  • Univariate Spatial Autocorrelation (Global and Local Moran's I): Employed to understand how PM2.5 concentrations are clustered or dispersed spatially within provinces and across neighboring regions.
  • Bivariate Spatial Correlation Analysis (Global and Local Bivariate Moran's I): Utilized to examine the spatial correlation between PM2.5 and individual socioeconomic factors.
  • Spatial Regression Models (Spatial Lag Model - SLM and Spatial Error Model - SEM): Used to quantify the impact of socioeconomic factors on PM2.5 while accounting for spatial dependencies and spillover effects. The study selected SLM based on statistical tests.

Datasets

  • Global Surface PM2.5 Concentration Dataset: Estimated by the GEOS-Chem chemical transport model, combined with aerosol optical depth (AOD) data from NASA MODIS, MISR, and SeaWIFS satellites. This dataset was calibrated using ground PM2.5 observations and provided at a 0.1° × 0.1° resolution for 31 Chinese provinces from 1998 to 2016.
  • Socioeconomic Data: Obtained from the National Bureau of Statistics of the People's Republic of China for the same 31 provinces and the period 1998–2016. The four key socioeconomic factors analyzed were:
  • GDP per capita (GDPP)
    • Industrial added values (IVA)
    • Urban population density (UPD)
    • Private car ownership (PCO)

All socioeconomic factors were standardized using z-scores to account for dimensional differences.

Key Insights and Ideas

  • Economic Growth Fuels PM2.5, But Not Uniformly: The study confirms that generally, as China's economy grew, so did PM2.5 levels. However, this relationship is not a simple linear one and varies significantly by region.
  • Regional Divergence in PM2.5 Trends: While most provinces saw PM2.5 increase with economic development, some regions like Southwest China experienced slower growth, and Northwest China even showed a descending trend in PM2.5 despite economic progress. This suggests that other factors (like natural conditions or specific policy interventions) play a crucial role.
  • Shifting Hotspots: PM2.5 pollution hotspots have shifted from Northwest China to North and East China over the study period, indicating a concentration of pollution in more economically developed and industrialized areas.
  • Urbanization's Changing Role: Urban population density (UPD) initially showed a positive correlation with PM2.5, but this relationship turned negative after 2006. This could be attributed to effective population control policies in some major cities.
  • Spatial Spillover is Key: The study strongly emphasizes that PM2.5 pollution doesn't respect provincial borders. Spatial spillover effects (where pollution in one province affects its neighbors) are significant and have been increasing over time, particularly in economically developed regions. This means isolated policy interventions are insufficient.
  • GDP and Industry Lead the Pack: GDP per capita and industrial added values were identified as the most significant drivers of PM2.5 increases, followed by private car ownership.

Key Results

  • General Increase: PM2.5 concentration in most Chinese provinces increased rapidly alongside socioeconomic factors from 1998 to 2016.
  • Regional Variations: Southwest China showed a slow PM2.5 growth trend, while Northwest China exhibited a descending trend.
  • Significant Correlations: GDP per capita, industrial added value, and private car ownership showed significant positive spatial correlations with PM2.5. Urban population density showed a negative spatial correlation after 2006.
  • Dominant Factors: GDP per capita and industrial added values were the most influential factors contributing to PM2.5 increases.
  • Spatial Autocorrelation: Global Moran's I values for PM2.5 were consistently positive and increased over time, indicating strong spatial clustering and homogeneity.
  • Spatial Lag Model: The spatial lag model confirmed significant spatial spillover effects of PM2.5 between adjacent provinces throughout the study period.

Limitations and Future Work

  • Data Granularity: While the study uses provincial-level data, finer-resolution data (e.g., city-level) could provide more granular insights.
  • Causality vs. Correlation: While the study identifies strong correlations and uses spatial regression to infer impacts, establishing definitive causality remains challenging in complex environmental systems.
  • Meteorological Factors: While mentioned as external causes, a more integrated analysis incorporating detailed meteorological data could further refine the understanding of PM2.5 dynamics.
  • Policy Implications: The authors suggest that their findings provide a theoretical basis for sustainable development and environmental protection, implying a need for coordinated inter-provincial policies to address the identified spatial spillover effects. Future work could focus on evaluating the effectiveness of specific policy interventions in mitigating PM2.5 pollution.

Summary updated

Abstract

A comprehensive understanding of the relationships between PM<sub>2.5</sub> concentration and socioeconomic factors provides new insight into environmental management decision-making for sustainable development. In order to identify the contributions of socioeconomic development to PM<sub>2.5</sub>, their spatial interaction and temporal variation of long time series are analyzed in this paper. Unary linear regression method, Spearman's rank and bivariate Moran's I methods were used to investigate spatio⁻temporal variations and relationships of socioeconomic factors and PM<sub>2.5</sub> concentration in 31 provinces of China during the period of 1998⁻2016. Spatial spillover effect of PM<sub>2.5</sub> concentration and the impact of socioeconomic factors on PM<sub>2.5</sub> concentration were analyzed by spatial lag model. Results demonstrated that PM<sub>2.5</sub> concentration in most provinces of China increased rapidly along with the increase of socioeconomic factors, while PM<sub>2.5</sub> presented a slow growth trend in Southwest China and a descending trend in Northwest China along with the increase of socioeconomic factors. Long time series analysis revealed the relationships between PM<sub>2.5</sub> concentration and four socioeconomic factors. PM<sub>2.5</sub> concentration was significantly positive spatial correlated with GDP per capita, industrial added value and private car ownership, while urban population density appeared a negative spatial correlation since 2006. GDP per capita and industrial added values were the most important factors to increase PM<sub>2.5</sub>, followed by private car ownership and urban population density. The findings of the study revealed spatial spillover effects of PM<sub>2.5</sub> between different provinces, and can provide a theoretical basis for sustainable development and environmental protection.
BibTeX
@article{yang2019spatiotemporal,
  author = {Yi Yang and Jie Li and Guobin Zhu and Qiangqiang Yuan},
  title = {{Spatio–Temporal Relationship and Evolvement of Socioeconomic Factors and PM2.5 in China During 1998–2016}},
  journal = {International Journal of Environmental Research and Public Health},
  year = {2019},
  doi = {10.3390/ijerph16071149},
}
Figure