“It’s fire season in the forests and wildlands of America.” So began an article by Barry Cipra (Fighting Fire with Data, SIAM News, July 2004). I recalled this article after hearing about the tragic events in the forest fires in Arizona earlier this week, and by the news of the fires near Colorado Springs that threatened the homes of a number of my former colleagues.
Cipra’s article reported on efforts by mathematicians and computational scientists working with NCAR (the National Center for Atmospheric Research) to develop real-time data-driven simulations of wildfires for use by fire fighters in the field. The simulations employ a coupled weather-wildfire model. The article was based in part on a talk at the 2003 SIAM Conference on Computational Science and Engineering by Janice Cohen, “Computational Science and Engineering Aspects of Wildland Fire Modeling.” I wondered what had become of that work and whether those tools had found their way into the hands of fire fighters.
Then today a received a copy of press release from the University of Arizona that shows that indeed research has continued and that mathematical/computational tools like those described a decade ago are indeed being used, and are playing a role in analyzing what went wrong in the tragedy at the Yarnell Hill Fire.
As the press release points out, the problem with forecasts based on weather models is that the predictions are expressed in terms of percentages, and decisions must be made in the light of these uncertainties.
It all indicates the further need for research to reduce the uncertainties and to better understand the risks.
James M. Crowley
Executive Director, SIAM