Dynamic Population Estimation Using Anonymized Mobility Data

TL;DR

This paper introduces a Bayesian model that leverages anonymized mobility data and static census data to estimate dynamic populations consistently across various spatial and temporal resolutions, addressing limitations of existing power law models.

Contribution

The paper presents a novel Bayesian approach for population estimation that remains accurate across different resolutions, unlike traditional power law models.

Findings

Model provides consistent estimations across resolutions

Outperforms power law models in flexibility

Enables real-time population monitoring

Abstract

Fine population distribution both in space and in time is crucial for epidemic management, disaster prevention,urban planning and more. Human mobility data have a great potential for mapping population distribution at a high level of spatiotemporal resolution. Power law models are the most popular ones for mapping mobility data to population. However,they fail to provide consistent estimations under different spatial and temporal resolutions, i.e. they have to be recalibrated whenever the spatial or temporal partitioning scheme changes. We propose a Bayesian model for dynamic population estimation using static census data and anonymized mobility data. Our model gives consistent population estimations under different spatial and temporal resolutions.