Abstract: Urban green spaces are of great significance to public health and sustainable development. The shape of park green spaces plays a critical role in evaluating their service capacity and ensuring equity. Traditional assessment methods rely primarily on accessibility indicators and tend to overlook the deep effects of morphological characteristics. Taking Xiamen as a case study, we incorporated a park shape index into an improved gravity model, quantified park service capacity by the size of the potentially served population, and assessed spatial equity using the Gini coefficient and Lorenz curve. The results showed that: (1) park shape complexity was positively correlated with service capacity (r=0.71), with a particularly strong correlation for linear parks (r=0.86); (2) For parks of similar size, those with higher shape indices exhibited stronger service capacity, indicating that the morphological effect was independent of park area; (3) Replacing the traditional centroid-to-centroid distance with the distance from the community centroid to the park boundary optimized the distance algorithm, reduced the average travel distance by 1684 m, and significantly improved overall service capacity, with the service capacity of linear parks increasing by up to 122.07%; and (4) Park service capacity differed significantly among park types and administrative districts, while the overall level of spatial equity remained low (Gini coefficient=0.49). Moreover, flexible park layouts adapted to natural topography improved the equity pattern. These findings elucidate the mechanism by which park shape influences service capacity and provide a scientific basis and practical pathway for precise green-space planning, the improvement of spatial equity, and sustainable urban development.

Key words: park service capacity, improved gravitational model, linear park, accessibility