Extracting Spatio-Temporal Patterns of Taxi Trips During COVID-19: a Case Study of Chicago
In 2020, the whole world experienced unprecedented challenges caused by COVID-19, and the United States is one of the most impacted areas. With the high contagiousness of COVID-19, human mobility became the catalyst of the pandemic. The rapid growth of big geo-data also provides an opportunity and a challenge to explore human mobility during a pandemic. This research extracts spatio-temporal urban dynamics from floating car data (FCD) in Chicago during and before COVID-19. The taxi trip records are aggregated by the hour at the community area level, so there is a taxi trip time series in each community area. We applied a time series decomposition method, Seasonal-Trend decomposition using LOESS (STL), to analyze taxi trips’ spatiotemporal patterns. STL can divide the original time series into different components, including trend, seasonality, and residuals. We also clustered the trend and seasonal effects of time series in different community areas of Chicago. The results show that time series decomposition is useful for extracting short-term and long-term patterns in urban dynamics. The comparison of the spatio-temporal patterns of taxi trips before and during COVID-19 indicates a sharp decrease and less diversity of human mobility during a pandemic.
Human mobility, Floating car data, Time series decomposition, COVID-19
Wang, B. (2021). <i>Extracting spatio-temporal patterns of taxi trips during COVID-19: A case study of Chicago</i> (Unpublished thesis). Texas State University, San Marcos, Texas.