Objective Urban parks, as critical nodes of public life in rapidly urbanizing contexts, play a vital role in providing recreation, social interaction, and psychological restoration. While traditional urban design research has largely emphasized the influence of physical boundaries—such as spatial morphology, green infrastructure, and facility layout—recent observations indicate that these tangible features alone cannot fully explain heterogeneous behavioral patterns such as crowd aggregation, avoidance, or uneven spatial use within open parks. This study aims to investigate how non-physical boundaries—including cultural identity, social distance, psychological safety, and behavioral norms—influence the distribution characteristics of crowd activities (DCCA) in urban parks. By developing and empirically validating a non-physical boundary perception pattern (NPBPP), the study seeks to reveal the mechanisms through which perception-driven processes influence DCCA and to support more inclusive park design .

Methods Three representative urban parks in Wuhan, China—Jiefang Park, Hankou Riverside Park, and Shahu Park—were selected as case study sites. Fifteen 50 × 50 m observation zones were delineated to capture behavioral data across varied spatial functions. First, structured questionnaires (N=277 valid responses) were designed to measure perceptions across four dimensions of non-physical boundaries: Cultural, social, psychological, and behavioral. Items were adapted from established research in environmental psychology, sociology, and proxemics, after pre-testing for contextual validity. Second, on-site behavioral observations were conducted to record path selection, stay duration, and activity type, with demographic and temporal controls. Statistical analyses proceeded in several stages: descriptive statistics summarized visitor profiles and activity preferences. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were conducted to assess the measurement model, with convergent and discriminant validity further evaluated. Reliability was confirmed with Cronbach’s α ranging from 0.79 to 0.87. A structural equation model (SEM) tested direct and indirect pathways between boundary perceptions and the distribution characteristics of crowd activities (DCCA). Bootstrap mediation analysis further examined indirect effects.

Results Findings indicate that social and psychological boundaries exert significant direct effects on the distribution characteristics of crowd activities, whereas cultural and behavioral boundaries mainly act through indirect pathways. Specifically, the SEM results show: 1) Social boundary perception (SBP) has a significant positive impact on DCCA(β=0.161, p<0.05). It also plays a mediating role in the effects of cultural and psychological factors on DCCA. 2) Psychological boundary perception (PBP) also directly influences DCCA(β=0.158, p<0.05), and is further involved in indirect pathways through social and behavioral boundaries. 3) Cultural boundary perception (CBP) shows no significant direct effect, but influences DCCA through dual mediation: via social boundary perception (CBP→SBP→DCCA, 6%) and psychological boundary perception (CBP→PBP→DCCA, 3%). 4) Behavioral boundary perception (BBP) does not independently predict DCCA, yet its effect emerges when combined with social cohesion or psychological safety, suggesting that behavioral cues are contingent upon prior perceptual conditions. Bootstrap mediation analysis further supported these findings by identifying indirect pathways, such as PBP→SBP→DCCA and SBP、PBP→BBP→DCCA. Together, these findings suggest a “perception−cognition−behavior−space” transmission chain,The results support four design principles: social gradient permeation, psychological safety anchoring, cultural symbol embedding, and collaborative behavioral cueing.

Conclusion This study contributes to research on urban public space design. Theoretically, it advances the understanding of non-physical boundaries by systematically integrating cultural, social, psychological, and behavioral dimensions into a single structural model, validated through empirical testing. The findings highlight that perceptual processes, rather than physical configurations alone, shape how individuals and groups occupy and interact within public parks. The model not only confirms the significance of direct effects (social and psychological boundaries) but also underscores the importance of indirect, chain-like mediation pathways (cultural and behavioral boundaries). Practically, these findings offer design for enhancing inclusivity and vitality in urban parks. By shifting the focus from physical segmentation to perception-oriented guidance, urban park design can more effectively accommodate diverse user groups, mitigate spatial exclusion, and foster environments that balance social interaction with psychological comfort. Future studies may extend the model across different cultural and climatic contexts, incorporate longitudinal and physiological measures of perception, and explore its applicability to other public space typologies such as plazas and waterfronts. Thus, the concept of non-physical boundaries may become a useful lens for bridging environmental psychology, sociology, and urban design in the pursuit of more human-centered and adaptive cities.