A multi-objective time series analysis of community mobility reduction comparing first and second COVID-19 waves

TL;DR

This study compares community mobility reductions during the first and second COVID-19 waves across various localities in America and Europe using Google mobility data, employing multi-objective visualization to analyze lockdown effects.

Contribution

It introduces a multi-objective comparison method for mobility reduction analysis across different localities and COVID-19 waves using Pareto-compliant visualization techniques.

Findings

Mobility reduction varies significantly between localities.

Temporal evolution of mobility reduction differs across regions.

Lockdown effects are context-dependent and vary over time.

Abstract

With the logistic challenges faced by most countries for the production, distribution, and application of vaccines for the novel coronavirus disease~(COVID-19), social distancing~(SD) remains the most tangible approach to mitigate the spread of the virus. To assist SD monitoring, several tech companies have made publicly available anonymized mobility data. In this work, we conduct a multi-objective mobility reduction rate comparison between the first and second COVID-19 waves in several localities from America and Europe using Google community mobility reports~(CMR) data. Through multi-dimensional visualization, we are able to compare in a Pareto-compliant way the reduction in mobility from the different lockdown periods for each locality selected, simultaneously considering all place categories provided in CMR. In addition, our analysis comprises a 56-day lockdown period for each locality and COVID-19 wave, which we analyze both as 56-day periods and as 14-day consecutive windows. Results vary considerably as a function of the locality considered, particularly when the temporal evolution of the mobility reduction is considered. We thus discuss each locality individually, relating social distancing measures and the reduction observed.