Comparing Air Pollution Modelling and Monitoring for Bushfire Smoke Health Impact Assessment in Australia

摘要

Fine particulate matter (PM2.5) is causally associated with cardiovascular and respiratory morbidity and mortality. Over recent decades the Sydney Greater Metropolitan Region population of five million have been repeatedly exposed to extreme levels of PM2.5 in smoke from bushfires, planned and unplanned. To date, estimates of historical exposure and related health effects have been based only on sparse monitors. The aim of this study is to compare existing estimates of exposure and health impact with those based on Chemical Transport Modelling (CTM) and blended surfaces; and to determine whether any differences are important for understanding the public health impact of bushfire smoke. Three methods of estimating PM exposure attributable to bushfires in Sydney will be compared. First, exposure based on only PM monitor interpolation will be compared to exposure based only on CTM. The former of these assessments is the subject of a paper under embargo until mid-April 2018. Comparisons of the magnitude and distribution of exposure and health effects will be presented. Developing results from a third 'blended' exposure surface will also be presented. This third exposure estimate is based on Bayesian blending of the monitor and CTM surfaces. A main aim of this work is to determine whether CTM predictions of future exposure can be bias-corrected based on the previous relationship between monitors and CTM, i.e. can predictions of future exposure be improved by taking into account recent model performance? Health impact will be quantified in terms of attributable premature mortalities and hospitalisations for cardiovascular and respiratory disease. The distribution and magnitude of these health effect estimates from the three exposure models will be compared to determine if there is a meaningful difference for understanding the public health impact of bushfire smoke.