Internal temperature inversion in semi-transparent materials based on spectral radiative transfer modeling

Non-contact temperature measurement in semi-transparent materials presents significant challenges due to optical inhomogeneities and depth-dependent thermal gradients. To address these limitations, this study proposes a spectral inversion model, which is based on the scattering matrix and the fluctuation-dissipation theorem. The proposed model achieves high-precision reconstruction of internal temperature distribution in multilayer materials by integrating the simulated annealing algorithm with the local ensemble averaging methodology. The method demonstrates robust performance under multiplicative uniform noise levels ranging from 0.5% to 5%, with maximum relative errors remaining below 0.67% at 5% noise. To assess its applicability, numerical simulations were performed on five-layer thin-film structures with distinct optical constants, successfully reconstructing temperature distributions. The proposed framework offers a non-contact method for accurate thermal profiling in complex media where conventional infrared thermometry is inadequate.