This work is a part of a larger study to retrospectively identify the cause of extreme particulate matter (PM) air pollution events and to assess the health effects of landscape fire smoke in 6 Australian cities occupying 3 distinct bioregions. Here, we present preliminary results for the effects of landscape fire smoke on daily all-cause, cardiovascular and respiratory mortality in Sydney, located in a subtropical bioregion, and surrounded by fire prone eucalypt forest and heath lands.
Pollution events for PM10 and PM2.5 were defined as any day from January 1997 through December 2006 where the 24 hour city-wide average concentration was above 99th percentile. These dates were cross-checked against newspaper archives, government agency records, and remote sensing data to evaluate whether they were related to landscape fire smoke, dust, or other sources. A case-crossover design was used to estimate the effects of PM10 on daily mortality for landscape fire smoke event days and nonevent days. Effects were adjusted for temperature, dew point temperature, and flu epidemics. Analyses were stratified by age and sex.
Of the 56 days above the PM10 99th percentile, causes for extreme PM concentrations were identified for 86% of the event days and, of these, 98% were attributed to landscape fire smoke. Landscape fire smoke event days (lag 1 day) were associated with a 4.7% (95% confidence interval: 0.1%–9.5%) increase the odds of all-cause mortality. The magnitude of the effect of PM10 on mortality was larger on landscape fire smoke event days compared to nonevent days, but neither was statistically significant.
Landscape fire smoke is the primary source of extreme PM pollution in Sydney. Preliminary results suggest that the magnitude of the association between PM10 and mortality may be different on landscape fire smoke event days compared to nonevent days.