Verification and Assessment of Automated NFDRS Forecasts Based on Gridded Weather Forecasts and Satellite Derived Fuel Parameters

During the summer of 1999, a prototype Automated Next-Day Fire Weather Forecast Process was field-tested. Forecast fields were high resolution grids (1km) of temperature, relative humidity, wind speed and cloud cover for 21Z (next day). Other forecast fields were the 24 hour (21Z to 21Z) maximum and minimum temperature and relative humidity, hours of precipitation, and accumulated precipitation. The forecast weather stream was blended with the NFDRS Fuel Model Map, the weekly Relative Greenness map (both at 1-km resolution), and 30-arc-second terrain data to produce 5 NFDRS forecast maps (Energy Release, Burning Index, Spread Component, Ignition Component, and Keetch Byram Drought Index) at 1-km resolution.

This paper reports the NFDRS forecast index verification statistics based on afternoon fire weather (NFDRS) observations reported daily to the Weather Information Management System. The paper also reflects on the operational requirements of grid-based versus point-based forecast processes and discusses the implications to both forecasters and end users of the information