| The rapid development of urbanization has led to a rapid increase in the number of urban parks.However,this process has also resulted in uneven spatial distribution of urban park resources and imbalance in visitation between different parks.In this context,there is an urgent need to improve the utilization and vibrancy of urban parks.Also,the construction of "park cities" requires high-quality development of parks,which is consistent with the shift from "acceleration" to "improvement" under the background of stock planning.Thus,exploring the impact mechanism of park vibrancy and balancing the vibrancy gap between different parks becomes an effective solution to alleviate the above problems.This study aims to use multi-source big data to characterize the distribution of urban park vibrancy,explore the impact mechanism of built environment on park vibrancy,and propose optimization strategies for park vibrancy to provide reference and decision support for urban park management and urban planning.First,this study quantifies urban park vibrancy and the built environment factors that may impact it with multisource big data.Second,data statistical analysis,spatial statistical analysis,and machine learning technology based on XGBoost algorithm are used to study the impact mechanism of park vibrancy.Third,the SHAP explanation method is used to explore the non-linear relationship between park vibrancy and built environment factors and the interaction between them.Finally,the author proposed some personalized vibrancy optimization strategies for a specific park are based on the research results.The main results of this study include:(1)Urban parks with high vibrancy in Nanjing are mainly distributed in the main urban area with a trend of high-value clustering.(2)The importance of the four types of built environment factors in this study is ranked as park attributes >surrounding environmental characteristics > social and economic attributes > public transportation accessibility.The permeability of parks with the surrounding environment and community has an important impact on park vibrancy.(3)There is generally a non-linear relationship and threshold effect between park vibrancy influencing factors and park vibrancy.The park vibrancy influencing factors that show a monotonic increasing trend in their impact on park vibrancy include the construction years of parks,the length of park boundary,the number of park entrances,the water coverage rate,the roundness of park profile,and the number of surrounding bus routes.The park vibrancy influencing factors that show a monotonic decreasing trend include the pedestrian route directness between parks and surrounding residential areas,and the distance between the nearest bus stop and subway station.(4)When the value reaches a certain threshold,there may be synergistic effects or negative interactions among certain vibrancy factors.Based on all the above results,the author proposed specific optimization strategies for improving park vibrancy.By applying new machine learning techniques,this study has achieved a refined exploration of the impact mechanism of urban park vibrancy.This can help planners and designers make better decisions and provide targeted recommendations for urban planning and park design. |
