Controlling Spatial Heterogeneity of Land Transfer Dynamics Using Multiscale Regression Approach for Sustainable Resource Management

ABSTRACT Land reallocation dynamics play a crucial role in shaping urban development and environmental sustainability in rapidly urbanizing countries. This study employs Geographically Weighted Regression (GWR) and Multiscale GWR (MGWR) models to analyze the spatial heterogeneity of land reallocation drivers across 285 Chinese cities, focusing on years 2008 and 2013. Our results reveal significant spatial variations in drivers of land reallocation, with energy intensity and foreign trade openness emerging as most influential factors. MGWR model consistently outperformed the classical GWR model, as evidenced by lower Akaike Information Criterion Corrected values (664.75 vs. 742.84) and a reduced sum of squares of residuals (130.44 vs. 165.31). This demonstrates multi‐scale nature of land reallocation processes and superiority of MGWR in capturing these complex spatial relationships. We find that the impact of energy intensity on land reallocation is particularly strong in northern China, with coefficients ranging from 0.208 to 0.761 in 2008. Foreign trade openness shows a more significant influence in coastal regions, with coefficients up to 1.078 in 2008. Interestingly, urbanization rate didn't show statistically significant effects, challenging some prevailing assumptions about urban growth drivers. Temporal analysis revealed a general decrease in magnitude of effects from 2008 to 2013, suggesting evolving dynamics in China's land reallocation processes. These findings have significant implications for sustainable land management and urban planning policies in China and other rapidly urbanizing countries. Our study adds to the literature by offering a detailed, spatially explicit understanding of land reallocation dynamics, providing valuable insights for achieving SDG 15.3 on land degradation neutrality.