Oral Presentation AMOS Annual Meeting and International Conference on Tropical Meteorology and Oceanography

A New National Fire Danger Rating System for Australia (#226)

Alex Holmes 1 , John Runcie 1 , Stuart Matthews 1 , Jennifer Hollis 2 , Belinda Kenny 1 , Saskia Grootemaat 1 , Paul Fox-Hughes 3 , Samuel Sauvage 3
  1. NSW Rural Fire Service, Sydney Olympic Park, NSW, Australia
  2. NSW Rural Fire Service, Queanbeyan, NSW, Australia
  3. Bureau of Meteorology, Hobart, Tasmania, Australia

Currently, in Australia, fire danger ratings are based on the empirically derived McArthur FFDI developed in the 1960s. This index accounts for neither the variation in vegetation or climates of Australia nor advancements in remote sensing and fire modelling. In July 2014, state fire services and the Commonwealth government agreed to develop a new National Fire Danger Rating System (NFDRS) based on current science and builds upon decades of research into fire behaviour.

The research prototype takes a modular approach calculating fire danger based on fire behaviour metrics for eight major fuel types for which suitable fire behaviour models were available: forest, grass, savanna, spinifex, mallee-heath, heath, pine, and buttongrass. These major fuel types were sub-divided on the basis of vegetation structure (e.g. wet and dry forest) and classified into several hundred individual fuel types, to differentiate variation in fuel attributes. Weather forecasts were used to calculate rate of spread, fireline intensity, flame height and spotting on a 1.5 km grid at hourly intervals for each 24-hour period. Similar to the FFDI, index values were then categorised into six risk levels unique to each fuel type. These were based on: fire behaviour, burn implication, fire suppression and containability and consequences.

A national trial with participants from each state ran between October 2017 - March 2018. Ratings were displayed in static ratings tables and maps and on an interactive website that included live incident feeds. Participants collected information on the behaviour of, and response to, live incidents then assigned each a rating category based on the detailed fire danger rating tables. Historically significant incidents and the Bureau of Meteorology’s weather reanalysis dataset were also used to characterise and evaluate the system retrospectively. The performance of the system was evaluated based on its ability to predict these observed ratings for actual fires.