A comparison is undertaken of theories for the gravity wave induced fluctuations in the intensity of airglow emissions and the associated temperature of the source region. The comparison is made in terms of Krassovsky's ratio &eegr;E for a vertically extended emission region (&eegr;E is the ratio of the vertically integrated normalized intensity perturbation to the vertically integrated normalized intensity-weighted temperature perturbation). It is shown that the formulas for &eegr;E in the works by Tarasick and Hines (1990) and Schubert et al. (1991) are in agreement for the case of an inviscid atmosphere. The calculation of &eegr;E using the theory of Tarasick and Hines (1990) requires determination of their function &khgr;; we show that &khgr; is simply related to the ''single-level'' Krassovsky's ratio &eegr; of Schubert et al. (1991). The general relationship between &khgr; and &eegr; is applied to a simple chemical-dynamical model of the O2 atmospheric airglow and the altitude dependence of these quantities is evaluated for nonsteady state chemistry. Though the Tarasick and Hines (1990) formula for &eegr;E does not explicitly depend on the scale heights of the minor constituents involved in airglow chemistry, &eegr;E implicitly depends upon these scale heights through its dependencies on chemical production and loss contained in &khgr;. We demonstrate this dependence of &eegr;E for the OH nightglow on atomic oxygen scale height by direct numerical evaluation of &eegr;E; in this case the dependence originates in the chemical production of perturbed ozone.