Abstract-Lee DeCola -University of Maryland, Baltimore County
Title: Space-time analysis of the US COVID epidemic
Title: Space-time analysis of the US COVID epidemic
Abstract:
Abstract:
The non-profit usafacts.org posts daily COVID-19 counts for US counties, which as of 2020 July 28 totaled 4,259,226 cases and 146,858 deaths since 2020 January 23. I used cyclical decomposition to produce filtered US, state, and county trends that show how the epidemic began with few cases and deaths, followed by a rapid rise to fluctuating behavior with a weekly cycle. Temporal trend estimates were kriged into an animated predicted surface that visualizes incidence originating in several centers and then spreading across the contiguous US county network. Spatial dependence models can be used to forecast the spread of the epidemic, and they provide uncertainty measures that must accompany narratives based on often unreliable epidemiological reports. The open-source R and QGIS environments were used.
The non-profit usafacts.org posts daily COVID-19 counts for US counties, which as of 2020 July 28 totaled 4,259,226 cases and 146,858 deaths since 2020 January 23. I used cyclical decomposition to produce filtered US, state, and county trends that show how the epidemic began with few cases and deaths, followed by a rapid rise to fluctuating behavior with a weekly cycle. Temporal trend estimates were kriged into an animated predicted surface that visualizes incidence originating in several centers and then spreading across the contiguous US county network. Spatial dependence models can be used to forecast the spread of the epidemic, and they provide uncertainty measures that must accompany narratives based on often unreliable epidemiological reports. The open-source R and QGIS environments were used.