Fire Risk Modeler       


 Jianbin Tang photo

Fire Risk Modeler - overview

Fire Risk Modeler
IBM Research – Australia


Bushfires in Australia are frequently occurring events during the summer season mainly due to Australia’s hot and dry climate. Record high temperatures, strong winds, low relative humidity, and high drought conditions usually precede the most devastating fires. These parameters form the core of a weather forecast.

In 2009, a standardized Fire Danger Rating (FDR) was adopted by all Australian states. FDRs are a feature of weather forecasts that help the community to prepare and/or take actions to prevent fire related disasters.

What is a Fire Risk Modeler?

Fire Risk Modeler (FRM) is a set of software tools that provide functional capabilities to simulate, identify, and rate geographical regions that could be potential sources of fire. It does so by integrating various sources of data, including real-time weather forecasts, and establishing a realistic relationship between those data sources in the form of a risk model. FRM provides output that could be (a) visualized on a geographic map in real-time (e.g., KML visualizer), (b) fed into a different module for further analysis (e.g., a fire-front modeler), or (c) simply stored as historical scenarios and used in future simulations.

How does FRM work?

FRM computes the McArthur Forest Fire Danger Index (FFDI) for each geographic region of interest. In Australia the McArthur Forest Fire Danger Index (FFDI) (McArthur 1967) is widely used to forecast the influence of weather on fire behavior, and the Australian Bureau of Meteorology routinely issues forecasts of Grassland and Forest Fire Danger Index (GFDI and FFDI) for use by fire authorities.

The computation of the FFDI is based on fine-grained data obtained from:

1)Real-time and simulated weather forecast (weather simulation is carried out locally at our lab using IBM Deep Thunder at fine-grained resolution).

2)�� Real and simulated topology, vegetation, infrastructure data (e.g., external sources, such as BOM,

3)�� Real-time and simulated events and incident data (e.g., planned burns data from Department of Environment and Primary Industries (DEPI), Country Fire Authority (CFA), Emergency Services Telecommunications Authority (ESTA)).

After computing the FFDI, the visualization of the fire danger rating is based on the following values and color code:

Fire Danger Rating

FFDI range







Very High




Code Red




The capabilities of the Fire Risk Modeler software tools could be used for the following benefits:

1)Real-time simulation of the relationship between various data: fine-grained localized weather condition, soil moisture content and drought factor, vegetation, topology, local events, and so forth. This leads to the forecast of Fire Danger Ratings and identification of potential ignition points and/or areas.

2)�� Real-time simulation of parameters (e.g., wind speed, area, terrain) at varying degree of resolution. E.g., gridded regions could have varying areas.

3)�� Integration of planned and unplanned events (real and simulated), such as planned burns in the area, to forecast regions of potential danger.

4)�� Easy integration of FRM results with other components via web-services.

Related Efforts

Fire Services Commissioner Victoria and IBM


1)The Information Interoperability Blueprint

2)The Victorian Network for Emergencies (VINE) Reference Architecture

3)Visualization of Planned Burns, a demonstrator.