Multi-scale CyberGIS Analytics for Detecting Spatiotemporal Patterns of COVID-19

Abstract

The ongoing COVID-19 pandemic has put the world into an unprecedented situation causing massive impacts on people’s daily life and businesses. While there exist numerous geospatial data portals and resources for fighting against the COVID-19 crisis, a major challenge is that few of these are based on rigorous geospatial analysis and modeling as many simply report COVID-19 cases and summary statistics. To tackle this challenge, it requires detailed data on populations, health care services, and high-risk settings, along with decision support tools for geospatial analytics and place-based intervention. Dynamic and interactive mapping informed by rigorous geospatial analysis and modeling is urgently needed for understanding how COVID-19 spreads across a number of spatial and temporal scales in various population contexts and for supporting decision making to mitigate the spread and minimize negative impacts. However, such geospatial analysis and modeling are often computation- and data-intensive, and thus require integration with and enablement by cyberGIS—geospatial information science and systems (GIS) based on advanced cyberinfrastructure. Therefore, this study aims to develop a cutting-edge cyberGIS approach to answer the following two research questions: (1) where are spatiotemporal clusters of COVID-19 death cases in the United States? and (2) what is the correlation between COVID-19 death case clusters and related socioeconomic factors? To answer these questions, we conducted a space-time kernel density estimation (STKDE) to explore spatiotemporal intensities of COVID-19 deaths in the United States. The results from STKDE provide a multi-scale understanding of the COVID-19 severity by identifying a number of spatiotemporal clusters centered around urban areas. Furthermore, the additional spatial correlation analysis based on STKDE results shows that the clustering regions with higher death rates are related to certain population characteristics (e.g., racial and socioeconomic status), highlighting the impacts of health disparities.