Joe McNorton, ECMWF Climate Scientist: So historically for fire forecasting we use what's called the fire weather index and this is a simple 8physics-based model where we use four weather 9variables temperature, wind, 10precipitation and humidity to forecast the chance that if a fire does occur how intense it will be. So, what we know from that is that it doesn't 11account for a lot of things. It doesn't account for fuel, it doesn't account for ignition sources and things like that. So what we try to do here is we try to incorporate more data into a machine learning framework.
CNN Narrator: For example, in the recent LA wildfires, unusual wet weather conditions 12leading up to the fires caused 13ample 14vegetationgrowth that was then made 15flammable by exceptionally dry weather in winter.
¡°So, you see?¡±
This new model would be able to take those factors into account and find the specific areas most at risk.
Francesca Di Giuseppe, ECMWF Principal Scientist: This new method, the 16probability of fire, having the memory of the fuel 17abundance in their 18formulation, allowed to really identify those regions that could be much more affected compared to simpler methods that only consider weather. And this is why our prediction in this case was much more 19precise and 20pinpointed the exact location when very close to Los Angeles where fire really occurred.
CNN Narrator: The danger wildfires pose is growing. Just last year, wildfires forced the 21displacement of 800,000 people, the highest numbers since records began back in 2008. Researchers hope the model can be used by fire officials to identify areas at the most risk of fires and prevent them before they start, saving lives and homes in the process.
