Predictive Models of Tranquility in Urban Public Open Spaces Based on Audiovisual Indicators Analysis
Wei Yan,Qi Meng,Yuxin Yin,Da Yang,Mengmeng Li,Jian Kang
DOI: https://doi.org/10.1016/j.buildenv.2024.112260
IF: 7.093
2025-01-01
Building and Environment
Abstract:Tranquil areas significantly enhance residential environmental quality and social well-being. However, a structured predictive assessment mechanism has yet to be established. This study develops prediction models for diverse places by conducting tranquility analyses based on 91 sample sites, integrating 10 objective and 31 subjective audiovisual indicators. The results indicate: (1) For auditory aspects, objective indicators such as sound level and psychoacoustic parameters demonstrate higher explanatory power compared to subjective indicators. By contrast, for visual aspects, subjective indicators such as perceived intensity and evaluation demonstrate higher explanatory power than objective indicators. (2) Sensitivity to sound is higher than to visual stimuli when perceiving tranquility. Negative elements (e.g., artificial sounds (AS): r = -0.69, p ≤ 0.05, other artificial elements (OAE): r = -0.41, p ≤ 0.05) have a stronger impact than positive elements (e.g., natural sounds (NS): r = 0.62, p ≤ 0.05, natural elements (NE): r = 0.29, p ≤ 0.05). (3) Key predictive variables for potential tranquil areas include the number of noises (NN), natural sounds/artificial sounds (NS/AS), civilization level (CL), Loudness, and natural contextual elements/other artificial elements (NCE/OAE). For natural places, AS and the number of people (NP) are key predictive variables. Similarly, for historical and cultural places, LA90, NN, and OAE are key predictive variables. These findings can be applied to the prediction, identification, and evaluation of different types of urban tranquil areas, thereby guiding their creation and optimization.