Fast estimation of the calorific values of forest fuels by near-infrared reflectance spectroscopy

Fuel calorific value is an important parameter in fire behaviour. The objective of this study is to evaluate the potential of using near-infrared reflectance spectroscopy (NIRS) for determining the calorific value of fuel. Fifty samples of various fuel components from the Mediterranean forest (17.1-24.6 MJ.kg-1) were used to establish a calibration equation between NIR spectra (1100-2500 nm) and calorific value. The coefficient of determination (r2) of the regression between predicted and measured values was 0.92 and the standard error of cross validation was 0.50 MJ.kg-1, i.e., twice the standard error of the measurement by the conventional method in a bomb calorimeter (0.23 MJ kg-1). Using this calibration equation, the calorific values of two sets of 15 samples collected weekly during the fire season from two shrub species, Arbutus unedo L. and Erica arborea L., were predicted. The r2 between NIRS-predicted and measured values were 0.79 and 0.76, respectively, and the standard errors of prediction were 0.36 and 0.21 MJ kg-1. These results show that NIRS is less accurate than the standard method but it can be used to quickly determine the calorific value of fuels when a large number of measurements are required.