Estimating error in wind speed measurements for experimental fires

Abstract

Most experimental fires, by nature, are small scale ([Formula: see text]100 m), and rate of spread measurements are taken over periods of several minutes. The aim of empirical fire modellers is to ascribe a single measure of rate of forward spread over a period to a single scalar measure of wind. The actual wind affecting the fire is unmeasurable; its value must be estimated from remote anemometry. Observation and consideration of the spatial and temporal statistics of the wind has allowed confidence limits to be placed upon the accuracy with which the measured wind reflects the wind acting on the fire front. Experimental data to verify these estimates was gathered during Project Vesta, a study into high-intensity fires in dry eucalypt forests. An equation that quantifies the accuracy of the estimate of wind affecting the fire front is given. The accuracy increases with time scale, size of the fire front, and density of anemometry. When applied to a measured wind speed taken some distance from the fire, it gives a useful estimate of the likely variation of the corresponding wind at the fire front.