Project Summary 

"We will use multiple sources of data, including empirical data on park visits
(cellphone big data), landscape features data (GIS and remote sensing data), and socioeconomic status
data (Census data). Using a combination of (big) data sources may address the limitations associated with traditional methods (e.g., surveys, observations, and interviews) that analyze park use and park benefits, which become resource-intensive and time-consuming when the number of park samples increases (Chen et al. 2020). Further, few studies have evaluated a city's park system performance or have explored ways to elevate the park system's performance from spatial planning and design.

We will analyze the spatial distributions of visits in more than 150 parks in Tucson using data from
cellphone geolocations, high-resolution landscape features, and Census population. Specifically, we will
conduct negative binomial-based logistic regression analyses to model the relationships between park
visits and design and socioeconomic status variables. We will also perform field observations at selected
parks to address the potential biases associated with cellphone big data due to contextual land-use
conditions. In summary, we expect that a creative combination of different (big) data sources can provide
a comprehensive evaluation of a city's park system performance and shed light on improving performance
through targeted planning and design interventions. The assessment framework presented in this study
can be instrumental in guiding park management decisions, including resource allocation and making
connections to public health and economic development goals and initiatives in Tucson and other cities in
the US."