Some Statistical Problems Inspired by Air Pollution

Across the mid-to-late 20th century, epidemiologists developed an understanding of the impacts of ambient air pollution on human health, inspired both by observed conditions due to chronic exposure, and by extreme climate events which provided natural (extreme) experiments. However, as an ecological problem, it proved difficult to validate acute exposure impacts - there simply wasn't enough data! In this talk I will discuss a series of connected problems in applied statistics that I have explored, worked on, or been connected to across the last decade, all stemming from a common question: given broad population-level measurements of air pollution and climate, and aggregate population-level health counts, what can we estimate about the acute impacts of air pollution exposure on human health? Many of these problems become time series problems, so there is a distinct flavour of those methods in this talk.

About Wesley Burr

Wesley Burr is an Associate Professor and Department Chair in the Department of Mathematics & Statistics at Trent University, Peterborough, Ontario. He completed his PhD in statistics under David J. Thomson at Queen's University in 2012, and after several years working as a Postdoctoral Research Fellow in Health Canada's Population Studies Division in Ottawa, joined Trent University in 2016, becoming Chair in 2022. Wesley has been a member of the Statistical Society of Canada for almost 15 years, and has recently ended six years serving on the executive of the Statistics Education Section. He remains Chair of the Statistics Education Awards committee. He is also on the executive of The International Environmetrics Society (TIES), and serves as Associate Director (Small Institutions) for the Canadian Statistical Sciences Institute. His research interests include time series analysis and spectrum estimation, environmental epidemiology, forensic statistics, statistics education, and any interesting problems in the analysis of scientific data.