The sensitivity of fire activity to inter annual climate variability in Victoria, Australia

Sarah Harris, Neville Nicholls, Nigel Tapper, Graham Mills

Research output: Contribution to journalArticleResearchpeer-review


Climate change is expected to have an impact on fire activity in many regions around the globe.The extent of this can only be determined by first establishing the relationship between climate and fire activity. This study relates observed changes in fire activity in Victoria to observed changes in antecedent and concurrent climate parameters – maximum temperature, rainfall and vapour pressure, using data for 1972–2014. A first-difference approach was adopted to estimate the amount by which the observed changes in the climate parameters would have altered the fire activity in the absence of other confounding effects. This study provides a method for examining the sensitivity of fire activity to changes in climate parameters without the need to consider the complex response of fuel dynamics to future climates and changes in fire regime or fire management. We used stepwise multiple-regression to determine the months whose climate parameters explained much of the variance in the total number of fires (TNF) and area burned in a fire season. The best performing fire–climate models explained almost two-thirds of the variation in year-to-year variability of fire activity. The significant explanatory ability of the fire–climate models established in this study reveals the combination of climate parameters that closely relates to the observed year-to-year changes in fire activity, and this may provide an additional valuable resource for fire management planning. Further, we explored the role changes in climate have had on the trend in fire activity. Natural logarithm of area burned and mean fire size have not significantly increased over the study period, but the TNF has significantly increased. We find that the observed increase in maximum temperatures and decrease in rainfall account for 26% of the observed increase in TNF for the 1972–2014 period. Therefore, most of the upward trend found in fire numbers must be due to factors other than climate (i.e. changes in fire occurrence, reporting/recording, land and fire-management changes). Additionally, this study concludes that total area burned should have also increased significantly due to the observed changes in climate and that improved fire-management practices may be offsetting this expected increase in the area burned. Finally, using the relationship established in this study between fire numbers and climate parameters, we estimate that a 2°C increase in mean monthly maximum temperatures could be expected to lead to a 38% increase in fire numbers.
Original languageEnglish
Pages (from-to)146-160
Number of pages15
JournalJournal of Southern Hemisphere Earth Systems Science
Issue number1
Publication statusPublished - 11 Jun 2020

Cite this