How LiDAR is changing fire predictions

For National Science Week, the Country Fire Service (CFS) is shining a light, quite literally, on the technology helping firefighters better understand bushfire risk. Firefighting might seem all about hoses, trucks, and protective gear, but at its core, it is driven by science. Physics, chemistry, and technology all play a part in predicting how a fire will behave and how to keep people safe. For National Science Week, the CFS is showcasing one of the newest tools helping firefighters do exactly that.

LiDAR, which stands for Light Detection and Ranging, is a remote sensing method that uses safe, pulsed lasers to measure distances. When a laser pulse hits an object, whether it is a leaf, tree branch, or rooftop, it bounces back to a sensor. The system records how long the light takes to return and repeats the process hundreds of thousands of times per second. The result is a detailed three-dimensional map of the environment.

Project Manager AFDRS, Simeon Telfer, explains that in the CFS context, the technology is focused on vegetation and its relationship to buildings. ‘We can accurately, down to millimetre precision, measure where the houses are, where the trees are, and turn that into a fuel model. The more accurately we know the fuel, the better we can predict fires,’ he said.

The Emesent LiDAR scanner.

Mapping fuel and predicting fire

In bushfire science, three main factors shape how a fire behaves: terrain, weather, and fuel. While terrain and weather have relatively good measurements and models, fuel, which is the vegetation that sustains a fire, is much more complex to measure. Fuel is not just about the amount of vegetation, but also its structure, such as how shrubs connect to tree canopies, how dense ground litter is, and how close it sits to assets like homes or community buildings.

Traditionally, this has been done through visual assessments or manual fuel sampling, where trained observers estimate the type and quantity of fuel in an area. Simeon said both methods have limitations. ‘Visual measurements are not very precise, and they are not always accurate. They are subject to observer bias. The way you assess a site might be influence by your recent experience. The alternative is manual sampling, but this can take several hours per site, which isn’t practical in most circumstances. The LiDAR does not have an opinion; it just models what is there much more quickly than manual fuel sampling.’

An example of the image generated by the LiDAR scan.

By scanning an area with LiDAR, CFS can create a “digital twin,” which is a snapshot of the environment in three dimensions that captures height, structure, and detail. This digital twin can be re-analysed as needed without returning to the field. If a new method for measuring fuel is developed, the same data set can be used to apply it, saving time and resources.

On foot, on road, or from the air

The LiDAR unit used by CFS is called an Emesent Hovermap, which offers flexibility in how data is collected. It can be worn on a backpack and carried through bushland to record detailed ground-level information, mounted on a vehicle to scan from the roadside, or attached to a drone to capture the tops of trees and other large structures. Each method has its strengths. Ground-based scanning picks up low vegetation in greater detail, while aerial scanning covers large areas quickly.

Project Manager AFDRS, Simeon Telfer, wearing the Hovermap LiDAR scanner at a prescribed burn in Lincoln National Park.

Processing is fast, and in under an hour, you can have a detailed 3D model. ‘A walk might take 10 to 30 minutes, depending on the area,’ Simeon said. ‘Back in the office, it takes about five minutes to plug it in, and another 10 to 30 minutes to process the data, depending on the size of the scan. Then you have a detailed map of the fuel and vegetation structure.’

From school evacuation plans to prescribed burns

Over the last six months, the CFS has already applied LiDAR in a variety of real-world situations. In several South Australian schools, scans have been used to map the distance from classrooms to surrounding vegetation, helping school leaders develop bushfire action plans. ‘In some cases, the data has shown buildings would be relatively safe during a bushfire. In others, it has shown they might need to close the school on days of certain Fire Danger Ratings,’ Simeon said.

LiDAR is also proving valuable for prescribed burn planning. By accurately mapping fuel before a planned burn, fire managers can select the right weather conditions to ensure the burn achieves its goal of reducing fuel loads to lower future fire risk.

Scientific monitoring equipment on a prescribed burn on Kangaroo Island.

The role of science in firefighting

For Simeon, National Science Week is an opportunity to remind people that firefighting is built on science. ‘Fire is basically a physical process. It falls into physics and chemistry,’ he said. ‘LiDAR is a new technology that is really going to help in fire science and our understanding of how fire affects the environment, people, and houses.’

By combining LiDAR with other fire modelling tools and even artificial intelligence (AI), the accuracy of fire behaviour predictions can improve significantly. In the future, machine learning could instantly classify vegetation and structures from 3D scans, giving firefighters better information faster.

Simulation of fire spreading towards a house based on a 3D LiDAR Scan.

The CFS is proud to highlight how science and technology are shaping the future of bushfire management. National Science Week celebrates the role of science in everyday life, and for the CFS, science is not confined to laboratories. It is part of the decisions made on the fireground, from predicting fire behaviour to planning prescribed burns, and it helps keep communities safer.