Deciphering the Nonlinear Association of Urban Park Usage with Perceived Mental Health Benefits and Satisfaction

Understanding the association between urban park usage and its effects on visitor satisfaction and mental health benefits is crucial for enhancing the visitor experience and well-being. This study aimed to explore the nonlinear relationships between urban park use and perceived mental health benefits and satisfaction using a gradient boosting decision tree (GBDT) model. The results showed that, compared to traditional linear models, the GBDT model demonstrated improved performance, increasing the R 2 values for mental health benefits and satisfaction by 40.38% and 33.7%, respectively. Regular park visits were found to maximize perceived mental health benefits, while the number of visits exhibited a U-shaped relationship with these benefits. Proximity to the park (within 10 km) and visiting with peers were associated with higher mental health benefits and satisfaction. Among all variables, visiting with peers and visit duration emerged as the most influential factors, contributing 18.3% and 14.1%, respectively, to model predictions. These findings highlight the importance of park design and management strategies that promote social interaction, encourage regular visits, and maintain park accessibility and novelty to maximize their positive impact on public health and satisfaction. This research underscores the potential of machine learning methods like GBDT in advancing our understanding of the complex interactions between urban park use and visitor experience.