Since March 2020, Master’s student Jingrui Mu has been analyzing COVID-19 data to gain insights into how the pandemic might spread. With guidance from her supervisor, Professor Mayer Alvo, Jingrui used both Bayesian spatial-temporal models as well as area-to-point Kriging methods to analyze the infection rate in Ontario. The Bayesian-temporal models take into account spatial and temporal effects on the spread of the virus in Ontario. They can be used to assess policy decisions where the objectives include limiting the number of people attending gatherings. The models can also be used to test for the significance of auxiliary variables.
Although originally used in geostatistics to predict the value in a geographic area for mining, soil, geology, and environmental science using a given set of measurements, Jingrui adapted the Kriging method to predict the spread of the SARS-CoV-2 virus. However, the basic Kriging method does not account for spatially varying population sizes. To address this limitation, Jingrui applied a variation of the Kriging method called the Poisson Kriging that incorporates the size and shape of administrative units, as well as the population density, to deduce infection rates or mortality rates and create isopleth risk maps. She applied the Poisson Kriging system to age-adjusted bi-weekly infection rates and mortality rates in different public health units. This novel approach showed how the virus spreads over time and helped estimate localized risk that can be used to detect zones of low and high risks.
As a next step, Jingrui intends to compare an ordinary kriged map to an Area-to-point Poisson kriged map, to determine which prediction method is more accurate in estimating local risk. Google published the COVID-19 Community Mobility report, which is a good dataset to explore the existence of a relationship between mobility at the province level and the evolution of the accumulated number of COVID-19 cases in Canada based on Bayesian spatial-temporal models. Jingrui has created a website dashboard to present these risk maps and hopes that these maps can help people understand how the coronavirus spreads over time. For her longer-term plans, Jingrui was recently accepted to a doctoral program at McGill University, where she will start in the fall of 2022.
Jingrui is thankful for Prof. Alvo’s guidance and believes that his mentorship made all the difference in her success. She advises prospective graduate students to never give up on any research ideas and to always discuss them with a supervisor.
