The visibility of retail frontages is critical for earning profits from spontaneous traffic visits to retail shops located along a street. The urban tree canopy plays a crucial role in enhancing the street-side environment, yet more is not always better when considering the placement of retail shops behind trees with big canopies. Related evidence in the literature is rarely provided, and an unclear relationship has been reported to exist between the number of shops for a specific retail type and the quantified ratio of the canopy shade in a street view. In this study, both big data crawling and deep learning were employed to unravel this relationship for retail shops in Changchun, Northeast China. The entire study area was divided into 6037 grid cells with a side length of ~0.6 km, wherein the number of shops of five retail types (food and beverage, shopping, life services, entertainment, and hotel) were quantified by computer counting their points of interest (POIs). The canopy shade was evaluated using the green view index (GVI) quantified through the ratio of canopy pixels divided by all the pixels in a street view image obtained through an online map API. A neighboring road network was categorized into four classes: class I road density mainly reduced the number of retail shops, and the road densities of classes III and IV accounted for more retail shops. The relationship between the number of retail shops and the GVI could be fitted with positive skewness curves for class II roads, where the critical peak of the GVI was estimated to be about 3.27%. The optimization scheme indicated that more retail shops should be placed along class I and II roads. In conclusion, more retail shops for food and beverage, shopping, and life services should be placed in the landscape neighboring big canopies.

Outdoor thermal comfort (OTC) is one of the most important issues for people exposed to urban green space (UGS). It is mainly evaluated by assessing equivalent meteorological factors that may combine and result in human responses of perceptions about comfortable vs. uncomfortable environments. The assessment of these responses is challenged by limits in a database that is powerful enough to support big-data analysis and an essential parameter that can evaluate health-related responses to OTC in urban forests. In this study, a meta-analysis was conducted by synthesizing studies of physiological and psychological health responses of green space (GS) visitors to meteorological factors in host cities on a global scale. A total of 15,000 numerical records about the psychological and physiological responses of UGS visitors were obtained from 105 papers across 68 global cities in 21 countries. A novel parameterization was developed to evaluate OTC by normalizing changes in health-related parameters and detecting their regressions against multiple meteorological factors. It was revealed that OTC resulted in more benefits in psychological health parameters than in physiological ones for UGS visitors worldwide. Globally, health-promoting OTC environments were more found in temperate countries than in countries subjected to other climates. People in Asian countries were indicated to live in environments with higher comfort that benefitted health states in cities at low latitudes. Thermal environments tended to be harsher and more challenging, mostly too chilling, in northern cities than in the south globally. A deep learning model using the ResNet algorithm was found to perform to an expected level with R² as high as >70% and errors controlled generally lower than 0.01. Taking mainland China for instance, it was predicted that thermal environments in eastern cities of China highly challenged the psychological health of local UGS visitors across all seasons of the year. Northern cities of China were predicted to have satisfied thermal environments that benefited psychological and psychological health responses in spring. In summer, cities in West China were predicted to have comfortable thermal environments, and in winter, environments in the southwestern cities were more comfortable.