Regression analysis of wildfire suppression

Abstract

One of the most important aims of forest fire research is how to better controlforest fires. One way to attain this aim is to develop better decision tools forestimating how much water is required for extinguishing a fire with a specific setof conditions. There has in the past generally been little research conducted withrespect to water requirements when suppressing a wildfire. A number ofexperiments were conducted in the early ‘70s where a spray rig with a nozzlearrangement was used in order to apply water uniformly across a fuel bed. Twostudies regarding the required duration of water application related to fires inbuildings were conducted in the past. Both studies were based upon statisticaldata from fire brigades; the duration of water application was determined as afunction of the fire area. This paper encompasses an analysis of more than64,000 wildfires occurring in Sweden between 1996 and 2009. The valuescalculated from the obtained formulations with regression analysis were found tobe in good agreement in the following cases: The amount of water as a functionof the extinguishing time, the extinguishing time as a function of the fire area;and the fire area as a function of the flame height. It was observed that thederived equations from the analysis in the above cases could provide an accuratecomputation. These relationships could be of considerable use when developingdecision tools for wildfire suppression, optimizing the use of resources duringthe suppression activities.