A Markov chain model of day to day changes in the Canadian forest fire weather index

Abstract

A Markov chain is used to model day to day changes in the Fire Weather Index (FWI) component of the Canadian Forest Fire Weather Index System. The results of statistical analyses of 26 years (1963 through 1988) of fire weather data recorded at 15 fire weather stations located across the province of Ontario suggest that it is reasonable to partition the fire season into three subseasons and model day to day changes in the Fire Weather Index class within each subseason as a Markov chain of order 1.