TL;DR
This study investigates how urban sprawl in China's Yangtze River Delta has evolved over two decades and its impact on PM2.5 air pollution. It reveals that while sprawl initially worsened air quality, policy interventions have shifted this relationship, with optimized sprawl now showing pollution-mitigating effects, highlighting the critical role of planning and regional cooperation.
What Problem Does This Paper Address?
China's rapid urbanization, particularly in the Yangtze River Delta (YRD), has led to significant urban sprawl. This sprawl has raised concerns about its impact on regional sustainable development and, crucially, air quality, with PM2.5 pollution being a major environmental challenge. While previous studies have examined urban sprawl and air pollution separately or in simpler models, there's a need for a comprehensive spatio-temporal analysis that accounts for the complex interactions and policy influences within a highly interconnected region like the YRD. This paper aims to fill this gap by analyzing the dynamic evolution of urban sprawl and its nuanced relationship with PM2.5 concentrations, considering both local and inter-regional effects, and the impact of policy changes.
Methods
The study employs a multi-faceted approach:
- Urban Sprawl Index Construction: A novel urban sprawl index (USI) was developed using nighttime light (NTL) data from DMSP/OLS and NPP/VIIRS satellites. This index captures both vertical (internal intensity) and horizontal (spatial extent) dimensions of sprawl.
- Spatio-Temporal Analysis:
- Exploratory Spatio-Temporal Data Analysis (ESTDA): Global and Local Moran's I indices were used to assess spatial autocorrelation and identify clustering patterns of urban sprawl over time. LISA (Local Indicators of Spatial Association) temporal pathways were analyzed to understand the dynamic migration trajectories of spatial association types.
- Slope Method: Used to classify sprawl into distinct growth patterns (slow, moderate, fast, rapid).
- Econometric Modeling:
- Multiple Linear Regression (MLR): To examine the relationship between urban sprawl, economic variables, and PM2.5. The analysis was divided into two phases: pre- and post-2013 policy implementation.
- Spatial Durbin Model (SDM): A more advanced spatial econometric model was used to capture spatial dependencies and spillover effects, accounting for both direct (within-city) and indirect (neighboring-city) impacts of urban sprawl and other factors on PM2.5.
Datasets
- Nighttime Light (NTL) Data: DMSP/OLS (2000-2013) and NPP/VIIRS (2012-2020) data were used to construct the urban sprawl index. These datasets were calibrated and corrected for continuity.
- PM2.5 Concentration Data: High-resolution (1 km) monthly average PM2.5 concentration data from 2000-2020, sourced from the China Atmospheric Composition Near-Real-Time Tracking (TAP) Data Resource Network.
- Other Variable Data: GDP, population, built-up area, green coverage rate, industrial structure, and foreign direct investment data were sourced from national and provincial statistical yearbooks.
- Geographic Data: Vector data for provincial and municipal boundaries were obtained from the National Geographic.
Key Insights and Ideas
- Policy as a Turning Point: The most striking insight is the dramatic shift in the urban sprawl-PM2.5 relationship around 2013, coinciding with China's "Ten Key Measures for Air Pollution Prevention and Control." This highlights the profound impact of targeted environmental policies on urban development trajectories.
- Sprawl's Dual Nature: Urban sprawl is not inherently good or bad for air quality; its impact is contingent on planning and policy. Disorderly sprawl exacerbates pollution, while optimized, compact sprawl can mitigate it.
- Interconnectedness Matters: The YRD is a highly interconnected region, and urban sprawl in one city significantly influences its neighbors. This necessitates a regional approach to environmental governance, moving beyond isolated city-level solutions.
- Spatial Inertia and Dynamics: Urban sprawl exhibits spatial dependence and inertia, meaning past patterns strongly influence future ones. However, policy interventions can disrupt these patterns and drive significant spatio-temporal transitions.
Key Results
- Sprawl Evolution: Urban sprawl in the YRD has shown a sustained growth trend from 2000-2020, with a slight northwestward shift in the center of sprawl. The dominant pattern is rapid expansion with significant spatial correlation.
- Spatial Clustering Decline: While urban sprawl exhibits spatial clustering, the global Moran's I index shows a gradual decline, indicating a weakening global clustering effect. Local high-high agglomeration zones decreased, while low-low zones increased.
- Pre-2013: Urban sprawl significantly intensified PM2.5 pollution, with positive direct, indirect, and total effects, driven by disorderly expansion and lack of planning.
- Post-2013: The relationship reversed, with urban sprawl showing pollution-mitigating effects, particularly through indirect pathways, due to policy guidance and optimized spatial layouts.
- Economic Factors: GDP and FDI positively correlate with PM2.5, while industrial structure and urban scale show negative correlations.
Limitations and Future Work
- NTL Data Limitations: The study acknowledges that NTL data, while useful for large-scale analysis, may not fully capture fine-grained variations in land use intensity or intra-urban functional differentiation.
- PM2.5 Data Validation: The PM2.5 data is based on gridded estimates, and future work could benefit from ground-based monitoring station validation.
- Causal Pathways: While the SDM captures spatial spillovers, it has limitations in disentangling specific causal pathways (e.g., urban form, industrial layout) influencing air quality.
- Future Research: The authors suggest comparative studies across different megaregions, using higher-resolution datasets (e.g., mobile sensor data), and employing more advanced modeling approaches like nonlinear regression, machine learning, and causal inference frameworks to further understand the complex interplay between urban sprawl and environmental outcomes.
Summary updated