Integrating global land cover and human mobility data to understand human-nature interactions

Preprint soon on ArXiv - link will be posted here as soon as the preprint is available.

Interactive model results

https://annecathrinel.github.io/nature-mobility/

LPS Abstract

One of the grand challenges of sustainability science is understanding the principal trade-offs between human well-being and the natural environment. Such trade-offs are dependent on how well-being benefits are derived from spending time in nature and how such use of nature may in turn threaten biodiversity. This knowledge is crucial for human activities to be sustained while keeping the planet in a habitable state for our species. One poorly understood aspect of socio-ecological systems is cultural ecosystem services (CES). CES are co-produced by people undertaking activities in nature and are generally defined as intangible benefits people obtain from nature exposure. One of these non-material benefits that can be obtained from human-nature interactions is human well-being. However, we still have a poor understanding of how well-being benefits emerge from human-nature interactions because we lack detailed observations of the cumulative exposure to nature for a large number of individuals. Smartphone users collect a wide range of information, such as geolocation, on their devices which they can choose to make available to apps they use. Associating such data with nature exposure and environmental features can contribute to understanding socio-ecological systems. We have developed a unique dataset of human movement across various land cover types, which offers a detailed view of global human mobility with 25 billion data points across five million individuals living in 245 countries. This smartphone dataset contains anonymized GPS location data sampled hourly for approximately five million individuals collected from 2017 to 2019 by a global smartphone and electronics company and users’ self-reported age, gender, and country of residence. Thus, providing the sociodemographic context for understanding human-nature interactions. We qualified the habitat exposure of individuals using the ESA WorldCover data (at 10-m resolution) to determine the ecotype of individuals. We find that individuals could be classified in persistent ecotype groups defined by the frequency distribution of habitat use. This work provides the foundation for developing spatio-temporal CES exploitation decision models to identify habitats and their socio-ecological contexts most likely to yield well-being benefits to users, without endangering biodiversity. This research highlights how remote sensing products can be combined with human mobility data to increase our understanding of socio-ecological systems and their dynamics for improved decision-making.