Assessing the health burden of bushfire smoke in Australia

Background Wildfires are increasing in frequency and severity in many regions of the world. Wildfires, like other biomass combustion processes, produce large amounts of smoke which degrades air quality and causes many adverse population health impacts. Current practices in Australia and elsewhere, to reduce the risk of future wildfires, include deliberate burning of the landscape (known as prescribed burning or hazard reduction burning), also potentially producing large amounts of smoke, and affecting human health. Yet the burden of smoke – from wildfires and prescribed burns - are seldom considered when assessing the trade-offs between these two types of landscape burns. There are important knowledge gaps that make assessment of these trade-offs difficult: (1) there is not a clear understanding of the relative health risks from landscape fire smoke compared with urban pollution largely derived from industry and vehicle emissions; (2) there has not been an adequate assessment of the health burden and health costs attributable to this source of air pollution or how this compares to other types of biomass pollution (e.g. wood heater emissions); and (3) there is a lack of understanding of the relative health burden and health costs from prescribed burning smoke and wildfire smoke taking into account the differences in frequency of fires, the amount of smoke produced and number of people exposed. In this thesis I aimed to provide new information using health economic analyses, to help fill these crucial gaps in existing evidence to support fire management practices that can achieve the greatest public health protection while minimising adverse public health impacts. Specific aims of this thesis 1. To synthesise existing evidence regarding climate change and heat-related extreme weather events - wildfires and heatwaves - in Australia (Chapter 2) 2. To systematically synthesise and meta-analyse existing evidence regarding the association between landscape fire smoke and asthma-related outcomes (Chapter 3) 3. To estimate the health burden and costs of biomass smoke in Tasmania, Australia, comparing the contribution of the two main emission sources, wood heaters and landscape fires (Chapter 4) 4. To estimate the health burden attributable to smoke during the 2019-20 Black Summer bushfires in southeast Australia (Chapter 5) 5. To compare the relative health impacts and costs of exceedances of air quality standards for particulate matter attributable to wildfires and prescribed burns in Western Australia (Chapter 6) 6. To evaluate the historical smoke-related health impacts and costs of wildfires and prescribed burns for a twenty year period in the state of New South Wales, Australia (Chapter 7) Methods Aim One (Chapter 2) was addressed through a review of the literature exploring the impacts of climate change on wildfires and heatwaves, and the impact of these on human health in Australia. Aim Two (Chapter 3) was addressed with a systematic review and meta-analysis of the literature exploring the asthma-related outcomes associated with short-term exposure to fine particulate matter (PM2.5) from landscape fire smoke (LFS). Aims Three to Six (Chapters 4-7) were addressed through the application of a health impact assessment framework to assess the health and economic impacts of bushfire smoke PM2.5 in Australia. Publicly available air quality and health data were used to model exposure and estimate the number of cases for each outcome. Australian-based values were used to monetise the health impacts. Whenever required, machine learning methods were used to impute missing data. Results In the literature review presented in Chapter 2, I found that a large proportion of natural hazard related health impacts (60% of deaths) were from wildfires or heatwaves, even when deaths and morbidity due to bushfire smoke were not accounted for. In the systematic review and meta-analysis presented in Chapter 3, I demonstrated that the LFS-related PM2.5 was positively associated with asthma hospitalisations and emergency department visits, with females being more susceptible than males for ED visits, and that there was an increasing association by age groups for hospital admissions and ED visits. In this xxiv Assessing the Health Burden of Bushfire Smoke in Australia study I also found that health risk coefficients for LFS PM2.5 and asthma-related outcomes might be higher than those more usually associated with typical urban PM2.5 mixtures. In modelling studies presented in Chapters 4-7, I substantially advanced understanding of the health burden and health costs of bushfire smoke in Australia. In Chapter 4 I demonstrated that in Tasmania, most of the burden attributable to biomass smoke is due to wood heater smoke (74% of impacts), but that in extreme bushfire years such as 2016 and 2019, impacts from landscape fire smoke are quite substantial and comparable to those of wood heater smoke. In Chapter 5 I estimated that during the Black Summer bushfires there were 417 premature deaths, 3151 cardiorespiratory hospitalisations and 1305 asthma ED attendances attributable to bushfire smoke exposure in New South Wales, Queensland, the Australian Capital Territory and Victoria. These results were of significant policy impact and used to inform several state and national inquiries. In Chapter 6 I found that in Western Australia, total health costs from exposure on days with PM exceedances (271 days between 2002 and 2017) were mainly dominated by prescribed burns (51%) followed by wildfires (41%). Finally, in Chapter 7 I found that total historical smoke-related health costs in New South Wales, Australia, were higher for wildfires (82.1%) compared to prescribed burns (17.9%). Nevertheless, the per hectare health costs were higher for prescribed burns ($AUD 477 per hectare) compared to wildfires ($AUD 104 per hectare). Conclusions Overall, I found that the magnitude of the health burden of bushfire smoke in Australia is substantial, comparable to other important pollution sources, and that prescribed burns could produce a similar level of health burden to that from wildfires. The findings in this thesis underscore the importance of the public health impacts from exposure to landscape fire smoke, and the need to incorporate smoke impacts and management into prescribed burn and other fuel management practices and wildfire responses. More generally these studies illuminate the complexity of understanding and measuring the trade-offs between wildfires and fuel management programs design to control them