TL;DR
This study investigates the primary drivers behind China's rapid urban expansion between 2000 and 2010, using the high-resolution GlobeLand30 dataset and a multilevel regression model. The findings highlight that market demand (economic growth) and land supply mechanisms are the most significant factors, with cultivated land area also playing a crucial role, underscoring the need for stronger farmland protection policies.
What Problem Does This Paper Address?
China has experienced unprecedented urbanization and industrialization, leading to a massive increase in impervious surfaces, accounting for a significant portion of the global total. This rapid land transformation has substantial environmental and societal impacts, including increased energy use, CO2 emissions, and altered land surface temperatures. While numerous studies have explored the drivers of land expansion, many have relied on government statistics, which can be unreliable due to issues like the exclusion of illegal land use. Furthermore, previous remote sensing studies often used datasets with limited spatial coverage, hindering global comparisons. This paper aims to bridge this gap by using a globally consistent, high-accuracy remote sensing dataset (GlobeLand30) to analyze the driving forces of impervious surface expansion across China, providing a more robust and comparable understanding of this critical phenomenon.
Methods
The researchers employed a multilevel regression model to analyze the drivers of impervious surface expansion. This approach is suitable for nested data structures, such as cities within provinces, allowing them to control for provincial-level factors while examining prefecture-level city dynamics. The model accounts for correlations within provincial groups and allows for the analysis of influences at both individual (prefecture) and contextual (provincial) levels.
Key variables were selected based on a review of existing literature, categorized into:
- Land Demand Variables: Reflecting economic activity and market forces (e.g., GDP increase, average wages).
- Land Supply Variables: Representing government control and land market mechanisms (e.g., ratio of land transfer fees to fiscal revenue, administrative levels).
- Control Variables: Including economic structure and agricultural land availability.
Additionally, spatial autocorrelation analysis (Moran's I and Getis-Ord Gi*) was used to identify hotspots and understand the spatial patterns of impervious surface expansion.
Datasets
- GlobeLand30 Dataset (2000 and 2010): A 30-m resolution global land cover dataset, noted for its high accuracy (better than 80%) and suitability for international comparisons. This dataset was crucial for identifying impervious surfaces.
- China Statistical Yearbook, China City Statistical Yearbook, and China Land and Resources Yearbook (2000-2010): These provided socioeconomic and administrative data at the prefecture and provincial levels.
Key Insights and Ideas
- Market Demand and Land Supply are Key: The study strongly confirms that both economic growth (market demand) and government-controlled land supply mechanisms are the primary drivers of impervious surface expansion in China.
- Cultivated Land as a Vulnerable Resource: The positive correlation between cultivated land area and impervious surface expansion highlights that arable land is a prime target for development, emphasizing the ongoing pressure on farmland and the need for stringent protection policies.
- Data Source Matters: Differences in findings compared to previous studies using government statistics suggest that remote sensing data, which captures actual land development rather than administrative allocations, provides a more accurate picture of impervious surface expansion.
- Administrative Levels Less Directly Influential: Contrary to some previous studies, administrative levels (like provincial capital status) showed less direct correlation with actual impervious surface expansion, possibly due to the focus on observed physical cover rather than land quota allocation.
Key Results
- GDP and Average Wages Drive Demand: An increase in GDP (coefficient 0.289) and average wages per capita (coefficient 0.215) were positively and significantly correlated with impervious surface expansion, indicating that economic development fuels demand for developed land.
- Land Transfer Fees Matter for Supply: The ratio of land transfer fees to fiscal revenue (RLTF) had a significant positive impact (coefficient 0.163), suggesting that the marketization of land supply contributes to expansion.
- Sub-Provincial City Status Negatively Correlated: Being a sub-provincial city (SUB) was negatively correlated with impervious surface expansion (coefficient -0.305), which the authors suggest might be related to illegal land use being more prevalent at lower administrative levels.
- Cultivated Land Area Promotes Expansion: The cultivated area in 2000 (CUL) showed a positive correlation (coefficient 0.18), indicating that areas with more arable land are more prone to development.
- Provincial GDP Density Matters: At the provincial level, an increase in GDP density (PGDP) positively correlated with impervious surface expansion (coefficient 0.125).
- Hotspots Identified: Spatial analysis revealed hotspots of impervious surface expansion concentrated in the Yangtze River Delta, Beijing-Tianjin-Hebei region, and the Shandong Peninsula.
Limitations and Future Work
- Data Limitations: The GlobeLand30 dataset, while accurate for impervious surfaces, does not provide detailed information on the driving forces for different land types within the broader category of "developed land."
- Future Research Directions: The authors suggest combining remote sensing data with statistical data at various scales and resolutions for more comprehensive research. They also emphasize the need for future studies to leverage the GlobeLand30 framework for cross-country comparisons and to inform specific policy implementations. Further research on the implementation periods of policies and their impact on driving forces is also recommended.
Summary updated