Global aging is getting worse, especially in China, a country with a large population. It is urgently needed to plan the site of new urban elderly care facilities for an aging society. Based on point of interest data and machine learning algorithms, we established a site selection model of urban elderly care facilities for Wuhan in China and selected potential optimal sites for new urban elderly care facilities. We found that 2059 of the 31,390 grids with a resolution of 500 m × 500 m of Wuhan are priority layout grids for new urban elderly care facilities. A total of 635 priority grids were further selected based on the agglomeration degree of the aging population in each street. They are mainly distributed in the areas with a concentrated aging population within the Second Ring Road around the urban centers. Additionally, some outer suburban streets with a relatively high aging degree also require immediate facility construction. The point of interest data and machine learning algorithms to select the location of urban elderly care facilities can optimize their overall configuration and avoid the subjectivity of site selection to some degree, provide empirical support for how to achieve a good configuration of “population–facilities” in space, and continuously improve the science of the spatial allocation of elderly care facilities.

Within the context of the “30·60 dual carbon” goal, China’s low-carbon sustainable development is affected by a series of environmental problems caused by rapid urbanization. Revealing the impacts of urbanization on carbon emissions (CEs) is conducive to low-carbon city construction and green transformation, attracting the attention of scholars worldwide. The research is rich concerning the impacts of urbanization on CEs but lacking in studies on their spatial dependence and heterogeneity at multiple different scales, especially in areas with important ecological statuses, such as the Han River Ecological Economic Belt (HREEB) in China. To address these gaps, this study first constructed an urbanization level (UL) measurement method. Then, using a bivariate spatial autocorrelation analysis and geographically weighted regression model, the spatial relationships between UL and CEs from 2000 to 2020 were investigated from a multiscale perspective. The results were shown as follows. The total CEs in the HREEB witnessed an upsurge in the past two decades, which was mainly dispersed in the central urban areas of the HREEB. The ULs in different regions of the HREEB varied evidently, with high levels in the east and low levels in the central and western regions, while the overall UL in 2020 was higher than that in 2000, regardless of the research scale. During the study period, there was a significant, positive spatial autocorrelation between UL and CEs, and similar spatial distribution characteristics of the bivariate spatial autocorrelation between CEs and UL at different times, and different scales were observed. UL impacted CEs positively, but the impacts varied at different grid scales during the study period. The regression coefficients in 2020 were higher than those in 2000, but the spatial distribution was more scattered, and more detailed information was provided at the 5 km grid scale than at the 10 km grid scale. The findings of this research can advance policy enlightenment for low-carbon city construction and green transformation in HREEB and provide a reference for CE reduction in other similar regions of the world.

No abstract