Nonlinear relationship between the urban form and street vitality: a data informed approach involving twelve Chinese cities

The relationship between the urban form and urban vitality has been widely discussed over the past few decades. Nevertheless, due to the limitation of available fine-scale urban data and analytical methods, previous empirical studies have mainly focused on mesoscale urban morphological characteristics within the scope of a single city, while microscale analyses among different cities are rare. In addition, most researchers adopt traditional linear or pre-defined regression models while ignoring the nonlinear effects of urban form characteristics. To address these two gaps, this study examines multi-source big data to measure multidimensional urban form characteristics from both the top-down and human-scale perspectives. Taking the streets of 12 Chinese city old districts as samples, this study uses a gradient promotion decision tree model to examine the irregular association between the urban form and street vitality. The explanations for the model indicate that: (1) Compared with the functional and human-scale morphological characteristics, the mesoscale morphological characteristics, especially the street intersection density, block size, street density, building density, and floor area ratio are dominant contributors to street vitality; (2) There is a nonlinear relationship between some urban morphological characteristics and street vitality and their associations change drastically past some thresholds of urban form attributes.