The Canadian wildfire forecasting model can be downloaded at http://www.firegrowthmodel.ca/prometheus/software_e.php for those curious. I've always found these predictions lacking due to limitations in weather forecasting.
Cool, thanks for sharing! Yeah, there are also issues reconciling the different scales of weather data (hundreds of meters to kilometers resolution), fuel data (~30 m) , and fire dynamics (sub-centimeter). It's still pretty amazing what they've been able to do with what they have, and weather forecasting has gotten so much better in the last decade! It's a pretty exciting time for the field.
She has another article about the basics of modeling: