Identifying Human Mobility Patterns Changes During Public Health Emergencies

Objectives This study aims to evaluate the changes in human mobility during a public health emergency.

Methods Employing detailed spatiotemporal big data on human mobility from China and the USA, we constructed three mobility indices: Mobility change index, distance change index, and volatility change index. We investigated the mobility characteristics within China's Guangdong-Hong Kong-Macau Greater Bay Area (GBA) and the United States' San Francisco Bay Area (SBA) during the epidemic. The singular value decomposition (SVD) algorithm was applied to identified underlying structures and patterns of mobility in these regions.

Results The results show that: (1) The GBA outperformed the SBA in terms of human mobility control with a greater decline in movement volumes, smoother volatility in daily movement and shorter average travel distances during the pandemic. (2) Human mobility patterns in the GBA were influenced by both the Chinese Spring Festival holiday and public health policies, from which the daily travel patterns, returning home travel characteristics and returning-to-work characteristics were identified. Human mobility patterns in the SBA show strong regularity (including daily travel characteristics, weekday travel characteristics and weekend travel characteristics), and public health policies do not have profound impacts.

Conclusions We quantify the changes in human mobility patterns under different epidemic control measures in two Bay Areas of China and the United States, which is essential to assess and identify intervention effectiveness. It also provides important evidences and references for various infectious disease control in the future.