Mobility data to aid assessment of human responses to extreme environmental conditions

Benjamin J Hatchett,Tarik Benmarhnia,Kristen Guirguis,Kristin Vandermolen,Alexander Gershunov,Heather Kerwin,Andrey Khlystov,Kathryn M Lambrecht,Vera Samburova

doi:10.1016/s2542-5196(21)00261-8

Abstract

Mobility data to aid assessment of human responses to extreme environmental conditions

Highlights

Mobile phone-derived human mobility data are publicly available in support of tracking the effect of interventions to control community spread of SARS-CoV-2.1–3 Previous work leveraged mobility data that used to be proprietary to examine responses to extreme events including wildfires and hurricanes.[4]
We applied county-level mobility data in the USA to reveal the power of this application during various extreme environmental conditions including hurricanes, wildfire smoke, and winter weather
We de-seasonalised daily, county-level, stay-at-home metrics provided by Facebook’s Data for Good Movement Range Maps[6] to remove the weekly cycle[7] and present data as anomalies by subtracting the long-term median for the period June 21, 2020, to July 18, 2021. This period follows the cessation of stayat-home orders due to COVID-19 in the country

Summary

Introduction

Mobile phone-derived human mobility data are publicly available in support of tracking the effect of interventions to control community spread of SARS-CoV-2.1–3 Previous work leveraged mobility data that used to be proprietary to examine responses to extreme events including wildfires and hurricanes.[4]. We applied county-level mobility data in the USA to reveal the power of this application during various extreme environmental conditions including hurricanes, wildfire smoke, and winter weather (figure). Concerns mounted regarding the risks posed by hurricane landfalls to communities struggling to contain COVID-19.8 Using mobility data from Puerto Rico and Louisiana, we found exposure to a hurricane corresponds with decreased mobility as people sheltered in place (figure A).

Results

Conclusion