Impact of the 2019/2020 Australian Megafires on Air Quality and Health

Abstract

The Australian 2019/2020 bushfires were unprecedented in their extent and intensity, causing a catastrophic loss of habitat, human and animal life across eastern-Australia. We use a regional air quality model to assess the impact of the bushfires on particulate matter with a diameter less than 2.5 μm (PM_2.5) concentrations and the associated health impact from short-term population exposure to bushfire PM_2.5. The mean population Air Quality Index (AQI) exposure between September and February in the fires and no fires simulations indicates an additional ∼437,000 people were exposed to "Poor" or worse AQI levels due to the fires. The AQ impact was concentrated in the cities of Sydney, Newcastle-Maitland, Canberra-Queanbeyan and Melbourne. Between October and February 171 (95% CI: 66-291) deaths were brought forward due to short-term exposure to bushfire PM_2.5. The health burden was largest in New South Wales (NSW) (109 (95% CI: 41-176) deaths brought forward), Queensland (15 (95% CI: 5-24)), and Victoria (35 (95% CI: 13-56)). This represents 38%, 13% and 30% of the total deaths brought forward by short-term exposure to all PM_2.5. At a city-level 65 (95% CI: 24-105), 23 (95% CI: 9-38) and 9 (95% CI: 4-14) deaths were brought forward from short-term exposure to bushfire PM_2.5, accounting for 36%, 20%, and 64% of the total deaths brought forward from all PM_2.5. Thus, the bushfires caused substantial AQ and health impacts across eastern-Australia. Climate change is projected to increase bushfire risk, therefore future fire management policies should consider this.

Keywords: Australia; PM2.5; air quality; bushfire; health impact assessment; wildfire.

Figure 1

Particulate matter with a diameter less than 2.5 μm (PM_2.5) fire emissions (Tonnes day⁻¹) across Australia between March 2019 and March 2020 from the FINN near‐real time fire emission data set. The timeseries shows the 2010–2018 daily mean PM_2.5 emissions (gray) and the 2019–2020 daily mean PM_2.5 emissions (maroon). Inset map: Total PM_2.5 fire emissions (Tonnes km⁻²) across eastern Australia between March 2019 and March 2020.

Figure 2

(a) Observed (black) and simulated (dotted magenta and dashed cyan) daily mean particulate matter with a diameter less than 2.5 μm (PM_2.5) concentrations. Simulations shown are no fires (dashed cyan) and fires (dotted magenta). The mean PM_2.5 concentration from all 80 observational sites across eastern‐Australia is shown for the model and observations. (b) The same as above but for individual cities (Sydney, Newcastle, Canberra, and Melbourne). The observed (black) and simulated (dotted magenta and dashed cyan) daily mean PM_2.5 concentrations are shown for each city. The total number of observational sites in each city is also shown on the left of each panel.

Figure 3

Monthly mean percentage of particulate matter with a diameter less than 2.5 μm (PM_2.5) attributable to fires, calculated as ${\text{PM}}_{2.5\text{fires}}-{\text{PM}}_{2.5\text{no}\text{fires}}/{\text{PM}}_{2.5\text{fires}}$ using the fires and no fires simulations. Monthly mean PM_2.5 concentrations from the fires simulation are also over plotted in contours for reference.

Figure 4

(a) Daily population exposure (in millions and %) to New South Wales Air Quality Index (AQI) values across eastern‐Australia (fires simulation) between September 1 and January 31. AQI values correspond to particulate matter with a diameter less than 2.5 μm (PM_2.5) concentrations of 0–8.5 (V. Good), >8.5–16.75 (Good), >16.75–25 (Fair), >25–37.5 (Poor), >37.5–50 (V. Poor), >50 (Hazardous), all in μg m⁻³. More information on how the AQI is calculated is available in Table S7 in Supporting Information S1. (b) Daily population‐weighted bushfire PM_2.5 exposure (in μg m⁻³) across all states in the model domain (red) and regionally for Victoria (green), Australian Capital Territory blue (yellow) and Queensland (purple) (fires‐no fires simulation) between September 1 and January 31.

Figure 5

(a) Daily population exposure (in millions and % of total population) to New South Wales Air Quality Index (AQI) values in individual cities (Brisbane (Queensland), Sydney (NSW), Newcastle‐Maitland (NSW), Canberra‐Queanbeyan (ACT), and Melbourne (Victoria)) between September 1st and January 31. AQI values correspond to PM_2.5 concentrations of 0–8.5 (V. Good), >8.5–16.75 (Good), >16.75–25 (Fair), >25–37.5 (Poor), >37.5–50 (V. Poor), >50 (Hazardous), all in μg m⁻³.More information on how the AQI is calculated is available in Table S7 in Supporting Information S1. (b) Daily population‐weighted bushfire PM_2.5 concentration (in μg m⁻³) in the cities of Brisbane (blue), Newcastle‐Maitland (purple), Sydney (green), Canberra‐Queanbeyan (yellow), Melbourne (gray), and Adelaide (orange) (fires‐no fires simulation) between September 1 and January 31.

Figure 6

(a) Estimated increase in the number of deaths brought forward across model domain (red) and the regions of Victoria (green), Australia Capital Territory (ACT) (blue), New South Wales (NSW) (yellow), and Queensland (purple) due to particulate matter with a diameter less than 2.5 μm (PM_2.5) from bushfires (fires only) between October 1 and January 31. Shading indicates the 95% confidence intervals of the estimate. The number of deaths brought forward due to bushfire PM_2.5 (fires only) (red) between October 1 and January 31 is also broken down by region (b) and city (c) and the total number of deaths is shown above the bars. The estimated number of deaths brought forward in each region (b) due to bushfire PM_2.5 (fires only) (red) in this study are compared to the Borchers Arriagada et al. (2020) (indigo) and Ryan et al. (2021) (gold) estimates for the same period.