Objective Following the outbreak of the COVID-19 pandemic, residents are urgently seeking urban sanctuaries that provide both physical and mental solace. Urban green spaces not only offer an opportunity to escape from the urban clamor for stress relief, but also provide various recreation services, fostering the enhancement of residents’ physical and mental well-being and overall happiness. Consequently, enhancing the capacity of urban green spaces for recreation services has become a pivotal concern. It is common to see diverse types of urban green space recreation services woven into a network. However, the current research fails to clarify the characteristics of such networks and how they response to different travel modes. Understanding these gaps is essential for creating urban environments that cater to the evolving needs of residents in the post-pandemic era, thus ensuring a holistic approach to well-being and happiness.

Methods Taking the central urban area of Wuhan as the research area, this research focuses on identifying and extracting urban green spaces that provide blue and green recreation services based on the coverage characteristics of vegetation and water resources. By virtue of the real-time travel path navigation from Amap (AutoNavi), the research extracts the shortest travel paths and durations from residential areas to urban green spaces for three travel modes: walking, cycling, and driving. Using the social network analysis software Ucinet 6.0, the research constructs three types of recreation service networks for each travel mode by taking into account travel durations of 5 minutes, 10 minutes, and 15 minutes. Moreover, the research analyzes the overall characteristics of the aforesaid networks in combination with global efficiency and the proportion of isolated nodes.Additionally, a “core – periphery” analysis and a “bi-components” analysis are conducted to investigate the functional types of nodes within these networks based on travel paths and durations..

Results When the travel duration is 5 minutes, more than 85% of urban green space nodes can be efficiently incorporated into recreation service networks under cycling and driving modes, while the recreation services under walking mode exhibit scattered local clustering. As the duration of travel increases, the overall efficiency of the recreation service networks formed under the three modes of walking, cycling and driving shows a trend of rapid growth, with the growth rate under the driving mode being the highest, followed in succession by that under the cycling mode and that under the walking mode. The proportion of nodes featuring core agglomeration functionality in the recreation service networks under the three travel modes is on the rise with extended travel duration. In the 15-minute driving mode, the proportion of nodes with core agglomeration functionality reaches the peak value of 40.97%. Although core agglomeration nodes are predominantly occupied by green recreation services, both the cycling and driving modes witness a gradual rise in the proportion of core agglomeration nodes associated with blue recreation services as travel duration lengthens. In the walking mode, the proportion of nodes with crucial connectivity functionality in the recreation service network within urban green spaces demonstrates an upward trend with increasing travel duration. Conversely, the driving mode exhibits a gradual decline in the proportion of nodes with crucial connectivity functionality. Green recreation services persist as the primary type of crucial and weak connectivity nodes. However, as travel duration increases, the proportion of blue recreation services within weak connectivity nodes experiences an upward shift. Notably, in the recreation service network formed under the walking mode, the proportion of blue recreation services serving as crucial connectivity nodes reaches 2%–3%. Furthermore, in the walking, cycling, and driving modes, the proportion of edge nodes in the urban green space recreation service network is the highest, particularly at a 5-minute travel duration, where it surpasses 85%. Regarding spatial distribution, whether in the walking, cycling, or driving mode, bridge nodes tend to disperse, while hub nodes exhibit a tendency to cluster.

Conclusions Achieving an overall improvement in the efficiency of urban green space recreation service networks necessitate a nuanced consideration of residents’ diverse recreational needs under different travel modes. Simultaneously, it underscores the importance of integrating blue and green recreation spaces. These findings offer substantial insights for the development and refinement of urban green space recreation systems and the systematic amalgamation of recreation resources, laying the groundwork for sustainable urban development. Notably, this research focuses exclusively on the estimated conditions formed by the travel duration from residential areas to urban green spaces, which are taken as a basis for calculating recreation services. It does not take into consideration the actual travel behaviors of residents heading to urban green spaces and associated influencing factors. In the future, a more comprehensive investigation is needed to unveil the authentic network relationships of urban green space recreation services and the complex mechanisms that underlie them.