Daily PM<sub>2.5</sub> and Seasonal-Trend Decomposition to Identify Extreme Air Pollution Events from 2001 to 2020 for Continental Australia Using a Random Forest Model
Nicolas Borchers-Arriagada,
Geoffrey G. Morgan,
Joseph Van Buskirk,
Karthik Gopi,
Cassandra Yuen,
Fay H. Johnston,
Yuming Guo,
Martin Cope,
Ivan C. Hanigan
Affiliations
Nicolas Borchers-Arriagada
Menzies Institute for Medical Research, University of Tasmania, Hobart, TAS 7000, Australia
Geoffrey G. Morgan
Centre for Safe Air, NHMRC Centre for Research Excellence, 17 Liverpool Street, Hobart, TAS 7000, Australia
Joseph Van Buskirk
Sydney School of Public Health, Faculty of Medicine and Health, University of Sydney, Sydney, NSW 2006, Australia
Karthik Gopi
Centre for Safe Air, NHMRC Centre for Research Excellence, 17 Liverpool Street, Hobart, TAS 7000, Australia
Cassandra Yuen
Centre for Safe Air, NHMRC Centre for Research Excellence, 17 Liverpool Street, Hobart, TAS 7000, Australia
Fay H. Johnston
Menzies Institute for Medical Research, University of Tasmania, Hobart, TAS 7000, Australia
Yuming Guo
Centre for Safe Air, NHMRC Centre for Research Excellence, 17 Liverpool Street, Hobart, TAS 7000, Australia
Martin Cope
Centre for Safe Air, NHMRC Centre for Research Excellence, 17 Liverpool Street, Hobart, TAS 7000, Australia
Ivan C. Hanigan
Centre for Safe Air, NHMRC Centre for Research Excellence, 17 Liverpool Street, Hobart, TAS 7000, Australia
Robust high spatiotemporal resolution daily PM2.5 exposure estimates are limited in Australia. Estimates of daily PM2.5 and the PM2.5 component from extreme pollution events (e.g., bushfires and dust storms) are needed for epidemiological studies and health burden assessments attributable to these events. We sought to: (1) estimate daily PM2.5 at a 5 km × 5 km spatial resolution across the Australian continent between 1 January 2001 and 30 June 2020 using a Random Forest (RF) algorithm, and (2) implement a seasonal-trend decomposition using loess (STL) methodology combined with selected statistical flags to identify extreme events and estimate the extreme pollution PM2.5 component. We developed an RF model that achieved an out-of-bag R-squared of 71.5% and a root-mean-square error (RMSE) of 4.5 µg/m3. We predicted daily PM2.5 across Australia, adequately capturing spatial and temporal variations. We showed how the STL method in combination with statistical flags can identify and quantify PM2.5 attributable to extreme pollution events in different locations across the country.
