Geographical and Temporal Weighted Regression: Examining Spatial Variations of COVID-19 Mortality Pattern Using Mobility and Multi-Source Data

Abstract

The COVID-19 pandemic has caused devastating impacts on public health and society worldwide. The transmission of the virus is facilitated by increased mobility, and specific populations with pre-existing health conditions are at a higher risk of COVID19 mortality. However, the spatial and temporal effects of health conditions and mobility on COVID-19 mortality are not fully understood. To address this gap, this project used the Geographical and Temporal Weighted Regression (GTWR) model to examine the influence of mobility and health-related factors on COVID-19 mortality in the United States. This analysis included significant demographic and health-related variables and compared the GTWR model's performance to the Multi-scale Geographically Weighted Regression (MGWR) model.

This study found that human mobility and specific health conditions have a significant spatial impact on COVID-19 mortality. The GTWR model allowed us to consider the temporal and spatial variability of COVID-19 mortality and identify different patterns in the association between the explanatory variables and COVID-19 mortality. The GTWR model's performance was superior to that of the MGWR model, indicating that the GTWR model is more effective at analyzing the spatial and temporal patterns of COVID-19 mortality. This model can provide a more comprehensive analysis of the impact of mobility and health-related factors on COVID-19 mortality, which can assist in identifying high-risk areas and developing effective policies to mitigate the spread of the virus.

This study underscores the importance of considering the spatial and temporal variability of COVID-19 mortality in analyzing the factors that influence its mortality. The results have important implications for public health interventions that target vulnerable populations and areas with high COVID-19 mortality rate. In conclusion, the GTWR model is a valuable tool for analyzing COVID-19 data and identifying high-risk areas. Considering the spatial and temporal variability of COVID-19 mortality, the results provide valuable insights to policymakers, which can inform effective decision-making regarding the pandemic.