A human-scale investigation into economic benefits of urban green and blue infrastructure based on big data and machine learning: A case study of Wuhan

Abstract

Rapid urbanization not only helps the urban economy achieve rapid growth but also imposes great pressure on resources and the environment. The green and blue infrastructure (GBI) in urban areas appears to be an important measure to enhance the economic benefits of the urban environment, but related research is still limited. In this article, we combine big data and machine learning methods to investigate the economic benefits of GBI in the city of Wuhan in Hubei Province. Specifically, the daily travel routes of citizens are analyzed at two scale levels (the housing and neighborhood scales); in our analysis, we start with the structural attributes of citizens’ own houses, move on to the city street views that citizens walk through, pass by the locational amenities where citizens spend their leisure time, and arrive at open spaces located throughout the city. The above trajectory dataset includes 30 variables, which were used as the input into ordinary least squares (OLS), geographically weighted regression (GWR), and multiscale GWR (MGWR). The results show that MGWR (adj. R2 = 0.524 at the housing scale, while adj. R2 = 0.869 at the neighborhood scale) with all control variables and the GBI attribute variables have the highest goodness of fit: the closer GBI is to urban residents, the higher the economic benefits are regardless of the area; and the higher the street visible green rate is, the greater the economic benefits are. Therefore, urban areas can appropriately increase the number of well-designed small GBIs located near urban residents. Our research provides insights into how big data and machine learning can be employed in frameworks to characterize the economic benefits of GBI and can be applied in other countries and regions.