(C) PLOS One
This story was originally published by PLOS One and is unaltered.
 . . . . . . . . . .



Using smartphone-GPS data to quantify human activity in green spaces [1]

['Alessandro Filazzola', 'Centre For Urban Environments', 'University Of Toronto Mississauga', 'Mississauga', 'Ontario', 'Apex Resource Management Solutions', 'Ottawa', 'Garland Xie', 'Department Of Biological Sciences', 'University Of Toronto Scarborough']

Date: 2023-01 

Cities are growing in density and coverage globally, increasing the value of green spaces for human health and well-being. Understanding the interactions between people and green spaces is also critical for biological conservation and sustainable development. However, quantifying green space use is particularly challenging. We used an activity index of anonymized GPS data from smart devices provided by Mapbox ( www.mapbox.com ) to characterize human activity in green spaces in the Greater Toronto Area, Canada. The goals of our study were to describe i) a methodological example of how anonymized GPS data could be used for human-nature research and ii) associations between park features and human activity. We describe some of the challenges and solutions with using this activity index, especially in the context of green spaces and biodiversity monitoring. We found the activity index was strongly correlated with visitation records (i.e., park reservations) and that these data are useful to identify high or low-usage areas within green spaces. Parks with a more extensive trail network typically experienced higher visitation rates and a substantial proportion of activity remained on trails. We identified certain land covers that were more frequently associated with human presence, such as rock formations, and find a relationship between human activity and tree composition. Our study demonstrates that anonymized GPS data from smart devices are a powerful tool for spatially quantifying human activity in green spaces. These could help to minimize trade-offs in the management of green spaces for human use and biological conservation will continue to be a significant challenge over the coming decades because of accelerating urbanization coupled with population growth. Importantly, we include a series of recommendations when using activity indexes for managing green spaces that can assist with biomonitoring and supporting sustainable human use.

In urban areas, green spaces represent important places for recreation, preservation of biodiversity, and delivery of ecosystem services, such as managing stormwater and reducing extreme heat. How people use green spaces and their impact on urban biodiversity is not well understood, particularly because it is difficult to monitor human activity. We used anonymized GPS data from smart devices to quantify green space use in Southern Ontario. We found a strong correlation between our estimates of mobile device activity and green space visitation rates determined from reservation data. We also found that users often spent most of their time on trails and that there were correlations between human activity and tree composition. We provide one of the first analyses exploring how people use urban green spaces using GPS data and the potential link to urban biodiversity.

Funding: This research was funded by a Post-Doctoral Fellowship awarded to AF by the Center for Urban Environments and School of Cities at the University of Toronto, Canada. GX was funded by an Ontario Graduate Scholarship, a Center for Environmental Research in the Anthropocene Graduate Fellowship, and NSERC CREATE funding (# 401276521) awarded to JSM. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Introduction

Cities are rapidly expanding, creating new challenges for managing green spaces. More than half of the global population currently lives in cities and that number is projected to increase to almost 90% by the end of the century [1,2]. As cities increase in size and area, green spaces including remnant natural areas, protected reserves, and urban parks, face growing stressors from human activity. Direct human use of green spaces can negatively impact urban wildlife including trampling, litter, the introduction of non-native species, and pollution [3–6]. However, both managed and unmanaged green spaces are important for city residents as a place for exercise, recreation, socialization, and supporting mental well-being [7–10]. This need has been amplified during the COVID-19 pandemic as people have increasingly sought out green spaces when indoor areas were closed or with increased risk of infection [11–13]. Thus, managing green spaces is a delicate balancing act between utility for people and the conservation of biodiversity.

One of the main limitations in effectively managing green spaces is the uncertainty around how and when people use these areas. Some managed parks use a reservation-based system with controlled points of entry, whereas other green spaces have multiple unrestricted access points. Trails are created to facilitate human movement and reduce disturbance to biodiversity, but visitors will still venture off-trail or erode new paths of easily navigable terrain [14]. Determining areas of high disturbance (i.e., high traffic), potential off-trail use, and overlap with sensitive species, can all be achieved through understanding human activity in green spaces. However, capturing human activity at a resolution fine enough for management, such as less than 100 x 100 m, is challenging. Typical methods for quantifying human activity include record-keeping visitors at entrance points, video monitoring, or post-hoc assessment of visitor impacts, such as campsite use [15–17]. Unfortunately, these data often neglect any spatial component of what visitors do outside of control points. Visually tracking visitors as they move within green spaces can be both cost-prohibitive and potentially intrusive. Using social media can be effective to track actions and activity from geotags of publicly shared images, posts, or tweets [18,19], but these data can be biased towards individual behaviours and be biased towards intentional points of interest [20]. With the widespread adoption of mobile smartphones and other smart devices, using anonymized GPS data can be an effective tool in determining patterns in the use of green spaces.

Connecting biodiversity observations to smart device activity data can pose a unique set of challenges beyond validating human activity patterns. Any correlation between species and human activity could occur because of multiple pathways including i) the species is relatively resilient to human disturbance and thus persists when others cannot, ii) the species or species’ habitat is attractive to visitors, iii) the property coincidentally is dominated by this species and is very accessible (e.g., walking distance to residential areas), or iv) any combination of these three factors. Teasing apart which of these pathways is relevant can be challenging because correlations between human activity and biodiversity may be because of aesthetic appeal or accessibility. While GPS data from smart devices have broad spatial and temporal coverage across a region, biodiversity data is often restricted to long-term monitoring plots that are static in location or multiple experimental sites that are short-term [21]. Biodiversity data are also rarely collected daily or cover a broad spatial area, presenting a challenge when trying to connect these two disparate types of data (e.g., evaluating the effects of human activity on biodiversity). Additionally, biodiversity surveys are often conducted away from areas with high human activity (e.g., trails, playgrounds, picnic areas) in more naturalized areas, reducing the chance that any overlapping human activity would be recorded. Using community science (e.g., iNaturalist, e-bird, Bumble Bee Watch) can be an effective tool for obtaining surveys with broad spatial and temporal coverage of green spaces [22,23], but these types of data are inherently correlated with smartphone use because of the mobile applications they require. A preliminary exploration of biodiversity and GPS data from smart devices would include examining the relative use of land cover types in green spaces to determine if certain areas, particularly where there is sensitive habitat, receive disproportionate levels of human activity.

Management of urban green spaces can be complex trying to balance different property types, land covers, and public uses. We partnered with Conservation Halton (www.conservationhalton.ca), a local conservation authority within the Conservation Ontario (www.conservationontario.ca) network responsible for natural areas, protected reserves, and urban parks in the regional municipality of Halton. Conservation Halton is responsible for managing a watershed that spans 1000 km2 of land containing different ecosystems, including forests (106 hectares; 7.5% coverage), riparian vegetation (7 km in total length; 14.5% coverage), and grasslands (130 hectares; 10.9% coverage) [24]. To improve ecosystem service delivery, Conservation Halton is working to increase natural land coverage above a particular threshold but must also balance naturalization with the provisioning of recreational opportunities [24]. Maintaining this balance is challenging since there are over 1.2 million visitors annually to the Conservation Areas due to their proximity to large urban centers (e.g., Hamilton, Milton, Mississauga, and Oakville) [25] with many of these municipalities being the fastest-growing cities in Canada [26]. Common social features of this landscape are recreational activities such as hiking, dog walking, cross-country skiing, and picnicking. Management of the 53 diverse properties within Conservation Halton’s jurisdiction represents some of the common challenges associated with land managers responsible for urban green spaces.

GPS data from smart devices can be a powerful tool in managing green spaces, but methods are needed that can properly quantify patterns of human activity. The purpose of our study was to describe how an activity index of anonymized GPS data from smart devices could be used for human-nature research, particularly looking at the associations between park features and human activity. GPS data from smartphones has been used previously to estimate trail use, green space access, and outdoor recreation patterns [13,27,28], but many studies rely on volunteer participants representing a fraction of green space users. Using Mapbox Movement (www.mapbox.com/movement-data) we obtained an anonymized activity index representing human density aggregated to 100 x 100 m grid cells and two-hour windows. In the following study, we developed methods for the synthesis, management, and analysis of anonymized GPS data from smart devices in Conservation Halton green spaces by answering the following three questions:

[END]
---
[1] Url: https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1010725

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/