EO/IR/FLIR sensor performance models currently employ a thermal difference metric(s) to predict target detection, recognition, and identification ranges in conjunction with minimum resolvable temperature difference (MRT) curves. In this paper, we present a target, atmosphere, background, and sensor-specific (TABSS) thermal difference metric, minimizing shortcomings and deficiencies of other thermal difference metrics currently used in thermal imaging system performance models. This metric is parametrically compared with other (Delta) (Tau) metrics. We also investigate target, background, and scene pixel variances behavior as the scene maps to a fewer number of pixels, which reveals potential applications in clutter metrics as well as detection, recognition, and identification range predictions. Finally, we survey current status of sensor performance models to seek an application of the TABSS (Delta) (Tau) metrics. We find that this metric will enhance the current thermal imaging system performance models to accurately predict detection, recognition, and identification ranges not only when the thermal difference is large, but especially when the thermal difference is small.