Abstract
Decision models involving prescribed fire (the use of fire as a silvicultural tool) were used to analyze the utility of four types of weather information as forecasts: climatology, current weather observations (used as a persistence forecast), special fire-danger forecasts issued by the National Weather Service, and a combination of the current weather observations and the fire-danger forecast.
Data from three seasons of observations (collected at weather stations between 6 500 and 8 000 ft in elevation in New Mexico and Arizona) were used in decision models to indicate that decisions based upon the combination of current weather observations and fire-danger forecasts have the greatest utility.
Forecasts for complex events (i.e., those involving temperature, humidity, and wind speed) showed lower utility gains than simple prescriptions (i.e., those based on temperature alone).
