Broadband chemical species tomography: Measurement theory and a proof-of-concept emission detection experiment

This work introduces broadband-absorption based chemical species tomography (CST) as a novel approach to reconstruct hydrocarbon concentrations from open-path attenuation measurements. In contrast to monochromatic CST, which usually involves solving a mathematically ill-posed linear problem, the measurement equations in broadband CST are nonlinear due to the integration of the radiative transfer equation over the detection spectrum. We present a transfer function that relates broadband transmittances to a path-integrated concentration, suitable for tomographic reconstruction, and use a Bayesian reconstruction technique that combines the measurement data with a priori assumptions about the spatial distribution of the target species. The technique is demonstrated by reconstructing a propane plume, and validating the results by point concentration measurements made with a flame ionization detector.