This paper presents a new algorithm for the Temperature-Emissivity Separation (TES) task in LWIR hyperspectral imagery. The joint retrieval of both physical quantities from the measured radiance represents an ill-posed problem. Moreover, the atmosphere modifies the self-emitted radiance from the object, thus making the estimation more complicated. The proposed technique solves such indeterminateness exploiting an optimization procedure, by estimating the best temperature that minimizes the atmospheric-residual features inside the emissivity spectral shape. In order to perform the atmospheric correction of the at-sensor radiance, a Look-Up-Table (LUT) of atmospheres is provided as input of the algorithm. The latter allows both to solve the TES problem, and to estimate the water vapor content of the monitored scenario. After the theoretical overview of the algorithm, evidence of the goodness of the proposed technique is provided. Results obtained for different scenarios, both in summer as well as in wintry atmosphere, suggest the efficiency of the proposed solution.