Exploring the Morphology Characteristics of Health Risk Space in the Post-Epidemic Era: The Case of Guangzhou

In the post-epidemic era, the notion of “healthy city” has received widespread attention. Due to the continuously increasing population mobility and complicated environmental system, some urban areas have become the potential health risk space. In existing studies, the identification of health risk space mainly concentrates on socio-demographic factors, while the corresponding space strategies are less touched. This study put forward 12 spatial indicators such as road environment, development density, building form, and ventilation potential to construct an indicator system. This study used random forest algorithm for in-depth learning of existing epidemic risk data, then optimized the spatial morphology based on the model results on the parametric platform, and proposed improvement strategies. The study found that the training model based on the spatial morphology indicators had a better identification accuracy (80.5%) than socio-demographic indicators. Results indicated that building patterns with large building intervals and excellent ventilation, multi-scale and scattered green open spaces, and small-scale and multi-functional communities could facilitate the decline of the health risk level.