Daily-activity-dependency of effective reproduction number in COVID-19 pandemic: direct modelling from GPS data

TL;DR

This study models COVID-19 infection risk in Tokyo using GPS data, revealing how different activities and locations influence the reproduction number, and emphasizing targeted restrictions for effective control.

Contribution

It introduces a direct modeling approach linking GPS-derived population density to infection risk, highlighting activity-specific and location-dependent effects on COVID-19 spread.

Findings

Stay-out activities have over 28 times higher infection risk than other activities.

Location-specific contributions significantly affect the effective reproduction number.

Restricting high-risk locations or activities can sufficiently suppress COVID-19 transmission.

Abstract

During the COVID-19 pandemic, governments faced difficulties in implementing mobility restriction measures, as no clear quantitative relationship between human mobility and infection spread in large cities is known. We developed a model that enables quantitative estimations of the infection risk for individual places and activities by using smartphone GPS data for the Tokyo metropolitan area. The effective reproduction number is directly calculated from the number of infectious social contacts defined by the square of the population density at each location. The difference in the infection rate of daily activities is considered, where the `stay-out' activity, staying at someplace neither home nor workplace, is more than 28 times larger than other activities. Also, the contribution to the infection strongly depends on location. We imply that the effective reproduction number is sufficiently suppressed if the highest-risk locations or activities are restricted. We also discuss the effects of the Delta variant and vaccination.