Many gases prove to have characteristic absorption properties which give the possibility of gas detection with spectrometer methods in the infrared region. The method utilizes the temperature difference between the air and the natural sources in the background e.g ground, vegetation and sky. This report contains statistics concerning the temperature difference between air and different types of background. The temperature difference is calculated from radiance data from FOI´s database. The results shows that with 50 % probability the temperature difference is 1-2 °C or larger from measurements with a background of grass, meadow or forest. By comparing the statistics of the temperature difference and performance parameters of existing spectrometer systems for gas detection the adequacy of these systems can be evaluated. Furthermore, an empirical and a semi-empirical model for prediction of temperature difference are described. Predictions with the semi-empirical model generally show good correspondence between calculated and measured temperature difference for the considered cases. The best result is achieved for a flat, homogeneous background e.g asphalt. Predictions with the empirical texture model are less successful in producing correspondence with the data. Deviations between the models and measured data are quantified and compared.