Covid-19 Pandemic Changes the Recreational Use of Urban Parks in Space and Time : Outcomes from Crowd-Sourcing and Machine Learning

The role of urban green spaces for human well-being and health has become crucial during the COVID-19 pandemic. Lockdown measures had a strong impact on the recreational use of urban green spaces. A limited access to urban parks during lockdown decreased the supply and increased the demand in recreational services. However, the effect of COVID-19 pandemic on the spatial-temporal patterns of recreational use within the parks with different landscape structure remains overlooked. In this study, the inter-relationships between landscape structure and changes in recreational use imposed by COVID-19 were investigated for the three different parks in Moscow (Russia). We used machine-learning to analyze social media photos collected for the periods prior, during and after the lockdown (2019 and 2020). Prior pandemic, different spatial (within a park) and temporal (during a year) patterns of photos were mainly explained by landscape structure and availability of infrastructure and facilities. During the pandemic, lockdown was the main factor, distinguishing temporal dynamics in the number of photos with clear minimums during the lockdown and a gradual increase within a month after the lockdown was lifted. In result, the visitation flow restored completely, and the total monthly number of photos after the lockdown was significantly higher than before it. However, the spatial patterns changed, including a higher distance between the photos’ locations and an increased number of photos related to nature observation. These findings demonstrate the crucial role of urban nature as a source of resilience in turbulent times which may have implications in decision-making processes of urban planning