The role of the fuel moisture content on the prediction of large wildfires using the Fire Weather Index system

Abstract

Fuel moisture content is one of the fundamental parameters in forest fire research and management given its implications for many aspects of fire danger systems. In Portugal, such as in many other countries, to classify the days with more favourable conditions for wildfires it is common to use the Canadian Forest Fire Weather Index System (CFFWIS) which is based on the estimation of moisture content of several fuel components. The mathematical structure of CFFWIS requires as input the daily meteorological parameters which are used to estimate the moisture content of the soil for different layers that are the primary outputs of the system – the fire moisture codes. The final output parameter of the system is the Fire Weather Index (FWI) which represents a measure of the fire danger due to meteorological conditions. The temporal scale of the study is from 2018 and 2021 and the study area is in Lousã, a central region of Portugal. In this work, in addition to the meteorological data, we will use as input the direct measurements of dead fuels in the CFFWIS and analyse its influence on the FWI. The fuel moisture content (mf) is determined through the sample collection of dead pine needles in Lousã. For this temporal scale, mf by sampling is significantly lower than the modelled mf using meteorological parameters. An advantage of using the mf measurements is to increase the range of FWI variation, giving a higher sensitivity to the index to more easily discriminate the days with high fire danger and large burned areas. Two methods are addressed: “FWI a” which represents the traditionally FWI determined only by meteorological parameters, and the “FWI b” which is determined with fuel moisture content measurements and with meteorological data for the days that we did not have measurements. The original FWI, based only meteorological parameters, is compared with the FWI determined using mf measurements. The different methods will be related with the number of fires and burned area to analyse their performance. The results show a good fit between mf and FWI for days with extreme weather conditions (mf<5%).