Understanding the spatial non-stationarity of industrial pollution's impact on cancer prevalence is crucial for targeted surveillance. This study examines the spatial non-stationarity of localized industrial point source emissions on regional colorectal cancer (CRC) patterns, utilizing a novel spatial coupling framework that integrates an exposure population-weighted assessment model (EPAM) with multiscale geographically weighted regression (MGWR). The key findings are as follows: First, we demonstrate that the association between metal surface treatment industry (MSTI) emissions and CRC is most accurately captured at a fine, localized scale of population exposure, a dimension obscured by conventional regional-aggregate or proximity-based exposure proxies. Further, our analysis reveals significant spatial non-stationarity, wherein the influence of MSTI emissions on CRC is concentrated in specific high-risk clusters, which primarily industrialized cities along China's southeastern coast. This spatial non-stationarity arises from the convergence of large-scale industrial pollution emissions, terrain favorable to pollutant dispersion, and high population density. Crucially, this EPAM-MGWR coupled framework quantifies localized exposure with a small-scale bandwidth, outperforming conventional medium-to-large-scale exposure proxies by enhancing the explained variance in CRC spatial patterns by 22 %–83 % compared to traditional Geographically Weighted Regression. In sum, these findings indicate that the carcinogenic impact of industrial pollution is a localized process, whose accurate detection at the regional level requires an analytical framework that reconciles the fine-grained emission dispersion with the multiscale health determinants. The coupling framework developed in this study offers a broadly applicable technical approach for examining the spatial associations between industrial point source pollution and various cancer types.
扫码关注我们
求助内容：
应助结果提醒方式：
京公网安备 11010802042870号