Predicting the Burn : How Mathematics Can Help Us Fight Wildfires

By Pabodha Wanniarachchi

Recently, I had the opportunity to attend the Undergraduate Research Symposium and Innovation Exhibition (URSIE 2025) held at the Faculty of Science, University of Kelaniya. Walking through the exhibition area, I was surrounded by innovative ideas, but one poster in particular caught my eye because of its real world application to environmental disasters.

Here is a review of the research project that explored how we can use mathematical modeling to predict how wildfires spread.

Research Details

• Research Title : Modelling Wildfire Boundary Movements Using Level Set Method
• Paper ID : 03
• Authors : P.W.G.D.T. Wijeratna & S.H.D.S. De Silva
• Department : Department of Mathematics, University of Kelaniya

What is the Problem?

We all know that wildfires are uncontrollable and pose massive risks to ecosystems, infrastructure, and human lives. A major challenge for firefighters and disaster management teams is predicting exactly where the fire will go next.

The researchers pointed out that most existing models for predicting fire spread are too simple. They often focus on just a few factors, like wind speed or the slope of the land, which makes them inaccurate when dealing with the complex reality of a raging forest fire.

The Solution : The Level Set Method

The core of this research involves a mathematical concept called the Level Set Method.

To put it simply, instead of trying to track individual particles of fire, this method looks at the fire as a changing geometric shape (a boundary) that evolves over time. The researchers used MATLAB to solve complex equations that simulate this boundary movement.

What makes this specific model impressive is that it doesn’t just look at one thing. It combines several crucial factors to create a more realistic prediction:

• Wind Speed & Direction: How the air pushes the flames.
• Terrain Slope: Fire travels differently uphill versus downhill.
• Fuel Type & Moisture Content: How dry or dense the vegetation is.
• Bayesian Correction: This is a statistical method that updates the predictions as new data comes in, making the model “smarter” over time.

Why Does This Matter?

The results displayed on the poster showed 2D contours of the fire front evolving. The researchers found that their model yielded more accurate and realistic wildfire spread predictions compared to existing models.

For a student looking at applied sciences, this is a great example of how abstract mathematics (like differential equations) can be translated into tools that save lives. By improving prediction accuracy, this model can help fire management teams make better decisions on how to allocate resources and where to evacuate people.

Conclusion

Paper ID 03 was a highlight of the exhibition for me. It successfully demonstrated that with the right mathematical approach, we can better understand and mitigate the chaos of natural disasters.