Unveiling urban spatial dynamics in climate and air quality indicators: a local estimation of the environmental Kuznets curve and the impact of green infrastructure

Abstract

This study tested the environmental Kuznets curve (EKC) hypothesis for climate and air quality indicators-CO2, PM10, NO2, CO, and O3-in relation to economic development and investigated the effect of green infrastructure (GI) in the EKC model. Employing a spatial approach, we utilized microregional data from the 25 autonomous districts in Seoul, South Korea, for 2012-2020. Economic status was measured by housing prices and GI by the ratio of diverse types of green space to the total area. Spatial heterogeneity was visually confirmed and analyzed using Moran's I statistics. Recognizing the importance of spatial heterogeneity, we employed geographically weighted regression (GWR) models to verify regional variations in the EKC hypothesis testing and the effects of GI. The results indicated significant spatial heterogeneity in the distribution of pollutants and significant spatial autocorrelation with strong clustering for CO2 and PM10 as well as for economic levels and GI ratios. Although the global model did not confirm the existence of an inverted U-shaped EKC, GWR analysis identified EKC for CO2, CO, and O3 in specific districts. The impact of GI on climate and air quality indicators varied by region and pollutant, contributing to PM10 reduction but an increase in O3. Together, these findings suggest the need for localized approaches to environmental policymaking, given that both the EKC relationship and the impact of GI can vary significantly by area and pollutant. This microregional spatial heterogeneity offers an additional foundation for facilitating local policy interventions to address each district's unique climate and air quality challenges.