Temperature Profile Inversion From CO_2 Spectral Intensities Through Levenberg-Marquardt Optimization and Tikhonov Regularization

摘要

In this study, we present an inverse calculation model based on the Levenberg-Marquardt optimization method with Tikhonov Regularization to reconstruct CO_2 temperature profile and average concentration from measured line-of-sight spectral intensity data. The measured spectra were synthesized through calculations from HITEMP2010, the high-temperature spectral database, for the CO_2 4.3 μm band. Although it has been shown Tikhonov regularization is suitable for the ill-posed inverse problem, it is difficult to select an appropriate regularization parameter, especially for the nonlinear problems. A new regularization selection method based on the combination of the L-curve criterion and the discrepancy principle is proposed and shows good generality for different temperature profiles inversion.