Using charcoal, ATR FTIR and chemometrics to model the intensity of pyrolysis: Exploratory steps towards characterising fire events

This study describes a multivariate statistical model (derived using partial least squares regression, PLS-R) that derives charring intensity (reaction temperature and duration) from the attenuated total reflectance (ATR) Fourier Transform Infrared (FTIR) spectra of charcoal. Data for the model was obtained from a library of charcoal samples produced under laboratory conditions at charring intensities (CI) relevant to wildfires and a series of feedstocks representing common tree species collected from Australia. The PLS-R model developed reveals the potential of FTIR to determine the charring intensity of charcoal. Though limited by the differences between laboratory-produced charcoal and the more heterogeneous and less-structured charcoal produced in a wildfire, the method was tested against fossil charcoal from a well-dated sediment core collected from Thirlmere Lakes National Park, Australia and showed a distinct change in CI that can be related to other climatic and environmental proxies. We suggest that the method has the potential to offer insights into the conditions under which natural charcoal is formed including the modelling of charring intensities of fossil charcoal samples isolated from sediments, archaeological applications or characterisation of contemporary fire events from charcoal in soils. [Display omitted] •Statistical model linking charring intensity (CI) to FTIR spectra of charcoal•Model derives CI irrespective of species used (species independent).•Fast and inexpensive method for deriving CI from charcoal formed in wildfires•Allows characterisation of past or contemporary fire severity from charcoal