城市绿地空间是城市中重要的基础设施，在城市系统中承担多种职能。绿地空间对城 市人群的健康效益已成为城市生态学、智慧城市等领域的研究热点。当下国内外有关城市 绿地的健康效益研究较多，但研究方法较为局限。首先，先前的多数研究用传统统计的回 归模型对绿地与疾病关系进行全局归因，忽略了绿地与疾病关系的空间异质性以及地理现 象的空间分异。其次，已有研究对于绿地空间的量化方法较为粗糙，未能量化到人群在绿 地空间的暴露（Exposure）程度。本文尝试采用空间滞后模型（SLM）和地理加权回归（GWR） 等方法探究人均绿地面积、街道绿地面积、植被覆盖度（FVC）以及人口加权绿地暴露指 数（Population-weighted Greenspace Exposure，PWGE）等指标对广州市各街道的呼吸系统 疾病与心血管疾病两类疾病的住院人数与城市绿地的关系，并揭示城市绿地空间对不同疾 病健康效应的空间分异。 研究结果表明：（1）局部自相关结果发现，广州市这两类疾病的发病人数存在空间一 致性，高-高聚集区主要集中于市中心天河区、海珠、越秀的部分街道，同时，低-低值聚集 区分别位于从化、白云和南沙。（2）人均绿地面积和PWGE与两类疾病住院人数均存在负 相关，即两变量对心血管和呼吸系统疾病住院人数都存在削减作用，但该作用的空间格局 存在差异，人均绿地面积作用强度以广州市中心为核心，向外逐渐变弱；PWGE的作用强 度则由北向南逐渐递减；同时，以上两个变量对于两种疾病住院人数削减作用的显著区域 也不同，虽然这些区域存在重合，但人均绿地面积和PWGE作用与呼吸系统疾病住院人数 的显著区域更大；（3）人均绿地面积相比于PWGE，对两类疾病住院人数存在更强的削减 作用；同时人均绿地面积和PWGE对呼吸系统疾病的住院人数的削减作用比心血管疾病更 强，城市绿地空间对于呼吸系统疾病有着更强的健康效应。本研究提供了空间疾病领域中 有关绿地健康效益话题在南亚热带地区的实证研究，不仅该方法可以用于后续的相关研究， 而且也为规划部门合理调控绿地布局提供参考。

Urban Green Space (UGS) is urban green infrastructure distinct from natural vegetation, playing various roles in urban systems. The health benefits of green spaces for urban populations have become a research hotspot in fields like urban ecology and smart cities. Currently, numerous studies on the health benefits of urban green spaces have been conducted worldwide, but the research methods are somewhat limited. Most studies use traditional statistical regression models to explore the attribution of green spaces to diseases, overlooking spatial dependencies among data. The quantification methods for green space are rather macroscopic, failing to consider the three-dimensional spatial structure of green spaces or quantify the extent of population exposure in green spaces. This study attempts to utilize spatial statistical methods, Geographical Weighted Regression (GWR) and Multi Scale Geographical Weighted Regression (MGWR), this study investigates the relationship between urban green spaces and the hospitalization rates of respiratory and cardiovascular diseases across different districts of Guangzhou, focusing on indicators such as Average UGS (per-capita green space), UGS Area (street green space area), FVC, and PWGE (Population-weighted Greenspace Exposure). The aim is to reveal the health benefits associated with urban green spaces. The results indicate: (1) Local spatial autocorrelation results show that there is spatial consistency in the number of cases of these two types of diseases in Guangzhou.

High-High clustering areas are mainly concentrated in districts such as TianHe, YueXiu, HaiZhu, while Low-Low clustering areas are located in Conghua, Baiyun, and Nansha. (2) The results of Multi-Scale Geographical Weighted Regression (MGWR) show that Average UGS and PWGE are negatively correlated with the number of hospitalizations for the two types of diseases, indicating that these two variables have a reducing effect on the number of hospitalizations for diseases. However, there are differences in the spatial pattern of this effect, with the intensity of the former centered on the central Guangzhou and gradually weakening outward; the latter gradually decreasing from north to south. At the same time, the significant regions of the above two variables' effects on hospitalizations for the two types of diseases are also different, although there is overlap in these regions, the regions where the effects on hospitalizations for respiratory diseases are significant are larger. (3) Compared with PWGE, Average UGS has a stronger reducing effect on the number of hospitalizations for the two types of diseases; meanwhile, Average UGS and PWGE have a stronger reducing effect on the number of hospitalizations for respiratory diseases than for cardiovascular diseases, indicating that urban green space has a stronger health effect on respiratory diseases. This study enriches the empirical research on the topic of green space health benefits in the subtropical region of South Asia in the field of spatial diseases, which can not only be applied to future related research in terms of methods but also provide effective reference for the rational regulation of green space layout by green space planning departments.