The ability to computationally predict wildfire spread would enable game-changing advances in fire prevention and optimized firefighting approaches. However, predicting wildfire spread is difficult because the large range of relevant length scales (from mm to km) is computationally intractable for a single simulation, and because there are many unknowns in how the landscape and weather quantitatively impact the combustion processes. These challenges are exacerbated by the complex nature of wildfires, which make controlled experimental datasets for model validation difficult. This thesis develops and applies diagnostic techniques, based on dual frequency comb spectroscopy (DCS), for non-intrusive quantification of wildfire combustion to create controlled datasets for developing and validating wildfire models.

DCS is an emerging broadband, high-resolution method of absorption spectroscopy that can simultaneously probe thousands of absorption features from many molecules, enabling accurate measurements of multiple species in transient high-temperature environments. We implement a near-infrared dual-comb spectrometer to measure water-vapor emissions and gas temperatures for the pyrolysis and flaming combustion of douglas fir. The data is combined with mass-loss history and surface temperature measurements to quantify the pyrolysis and combustion across a range of moisture contents. While these experiments demonstrate the promise of DCS in wildfire experiments, the number of measured species is limited by the weaker near-infrared absorption features accessible by the DCS. New mid-infrared DCS technology enables measurements in an optimal wavelength region for multi-species detection, but complicates the analysis due to the overlapping nature of absorption fingerprints in this region.

To address these challenges, two absorption analysis techniques are introduced. A new absorption spectroscopy technique eliminates the need to account for the laser intensity in absorption spectroscopy by converting the measured transmission spectrum to a modified form of the time-domain molecular free induction decay using cepstral analysis. A second technique allows researchers to scale reference absorption cross-sections to arbitrarily higher pressures, removing the need for reference cross-sections that match the pressure of the experiment.

We apply the two analysis techniques to mid-infrared DCS of pyrolyzing wildfire-relevant fuels to quantify in situ concentrations of eleven molecules and gas temperature, and investigate the thermal decomposition of different species of wood.