**5. Conclusion**

The study investigated green spaces in Altstetten–Albisrieden, Zürich, at two scales. At the district scale, public, community and private green spaces, categorised by their ownership and accessibility, were used to depict green space structure. The connectivity analysis of the three green-space types suggested the role of each space in the urban fabric. Four significant green space clusters and their 400-m service areas were selected as subsite cases for further investigation concerning their potential for social integration at the neighbourhood scale.

The study probed the disparities between public green spaces' structural significance and their land areas. These gaps might indicate the misfunction of some spaces and suggest that designing public green spaces entails more than merely facilitating leftover land. Providing and designing public green spaces require more careful consideration of location, size, and roles in the urban fabric. A further investigation shall also take into account public transportation networks and other transport means, which may enlarge public green spaces' service coverage and modify their connectivity at the city scale, thus generating more accurate results.

Providing more or enlarging public green spaces might encourage space use frequency and duration and, hence, improve individual health. The relation between space sizes, space-use frequency, and duration, indicated in the subsites, is not a simple linear relationship but is also impacted by other factors.

Nevertheless, the study showed that the green space structure, represented by the four subsite cases, or, more specifically, the combination of the space types, the provision of facilities and enabling various activities to be undertaken, and the setting of boundaries could impact space-use activities and its social cohesion potential. Among the subsite cases, relatively small green spaces (demonstrated in Süsslerenanlage) interwoven with private areas can benefit the community exclusively. In contrast, green spaces with clear boundaries and various functions (as in Bachwiesen) can strengthen social diversity.

The current findings naturally have some limits. The space use analysis can be improved to reflect the actual situation more accurately by obtaining more onsite observation data during different periods of the year. The demographic data of the four subsites needs to be included. This data may explain the gender or age difference and cultural structures to further our understanding of how green spaces' support for people's outdoor activities can equally cover the different social groups.

Given the challenge that Altstetten–Albisrieden needs to house another 20% population in the coming decade, the study on public green space provision and distribution will synthesise with urban densification scenarios and building typologies as the next step. Green spaces would extend and connect well to each other horizontally on the ground and vertically in building structures. Buildings and green spaces might convert their competitive relationship into collaboration. Levels of green spaces with different accessibility might encourage further community integration with various needs. Nevertheless, a careful planning and design approach involving stakeholders is necessary to optimise development and future performance.
