10:45 AM - 12:15 PM
[HTT16-P03] Spatio-temporal analysis of urban population during the COVID-19 pandemic using 1 km-gridded daytime population data
Keywords:COVID-19, Urban uses, Dynamic Time Warping
Since the worldwide spread of COVID-19 in late 2019, Japan has implemented various measures, including voluntary curfews, to restrict people's movements and behaviors. Although the increase in the number of vaccinated individuals appears to have lessened the impact of COVID-19 on urban activities, it is likely that local governments will continue to suggest avoiding population concentration by stimulating social distancing, such as through telecommuting and online shopping. Therefore, to understand how the population distribution and movement patterns have been affected by the pandemic, this study aims to analyze the spatio-temporal patterns of the daytime population in Tokyo, Kanagawa, Saitama, and Chiba prefectures from 2019 to 2021, the year of the COVID-19 pandemic.
The 1 km-gridded monthly daytime population data, published by the Ministry of Land, Infrastructure, Transport, and Tourism, were applied to Dynamic Time Warping (DTW), a technique for estimating the similarity of time-series patterns. By clustering the results of DTW, this study compared time-series patterns to identify similarities, rather than simply delineating temporal changes. The resulting clustering map shows that prior to the COVID-19 pandemic in 2019, people tended to be concentrated in the city center. However, during the first emergency declaration period (March-May 2020), the population was restricted to residential areas and their vicinity, and a similar trend persisted from May onward. Understanding the spatio-temporal pattern of the daytime population in the post-COVID-19 era is meaningful in considering how cities should be utilized in a society that coexists with infectious diseases.
The 1 km-gridded monthly daytime population data, published by the Ministry of Land, Infrastructure, Transport, and Tourism, were applied to Dynamic Time Warping (DTW), a technique for estimating the similarity of time-series patterns. By clustering the results of DTW, this study compared time-series patterns to identify similarities, rather than simply delineating temporal changes. The resulting clustering map shows that prior to the COVID-19 pandemic in 2019, people tended to be concentrated in the city center. However, during the first emergency declaration period (March-May 2020), the population was restricted to residential areas and their vicinity, and a similar trend persisted from May onward. Understanding the spatio-temporal pattern of the daytime population in the post-COVID-19 era is meaningful in considering how cities should be utilized in a society that coexists with infectious diseases.