Evaluating soil-driven decay of preservative-treated wood with short-wave infrared hyperspectral imaging

This study evaluated the use of a short-wave infrared (SWIR) hyperspectral camera to monitor time-dependent fungal decay in preservative-treated and untreated Scots pine sapwood over a year-long terrestrial microcosm (TMC) experiment, to assess its potential for the inspection of wooden utility poles. Previous hyperspectral studies of wood decay focused on basidiomycete monocultures in sterile laboratory conditions. This research extended that work by exposing Scots pine sapwood to soil, without treatment and with one of four preservative treatments (creosote, two oil-based and one water-based), with retentions adjusted to target a 20% mass loss, not commercial levels. Spectral changes from decay were analysed using difference spectra, principal component analysis, and partial least squares regression (PLSR). Treated and untreated wood exhibited largely similar spectral characteristics up to ∼20% mass loss, with an increasing lignin-related contribution at higher mass loss. A PLSR model with three latent variables predicted mass loss with a root mean squared error of 2.1 percentage points in replicate cross-validation. However, one oil-based preservative leached, making the decay-related mass loss of those specimens difficult to predict. Under lab conditions, SWIR hyperspectral imaging detected decay-related spectral changes, which indicates that it could eventually support inspection of wooden utility poles.