Full utilization of hyper-spectral imagery in the thermal infrared region requires a high-quality calibration, and the calibration quality requirement increases with the power and sophistication of the retrieval algorithms to be employed. Here we examine calibration issues associated with an airborne imaging Michelson Fourier Transform Infrared (FTIR)Spectrometer operating in the long-wave infrared (LWIR) region. In addition to the fundamental challenge of extracting a weak signal of interest from a complex background, problems which arise in such an instrument include pointing jitter, detector non-linearity, sampling position errors and etalon effects within the focal plane array. In each case, a frequent, high-quality calibration can ameliorate these problems. We discuss several hyper-spectral data analysis techniques, and how our calibration strategy, incorporating both ground and on-board calibration systems, improves the sensitivity of the retrievals.