Gravity wave‐driven fluctuations in the O 2 atmospheric (0‐1) nightglow from an extended, dissipative emission region

The omission fluctuations of the O 2 atmospheric nightglow due to gravity waves, evanescent waves, and acoustic waves are calculated using the Eulerian model formalism of Schubert et al. (1991) from which the complex parameter <η> is derived: where I is the nightglow intensity, T I is intensity‐weighted temperature, the brackets denote an altitude integration over all emitting layers, an overbar denotes an unperturbed state and a prime denotes a perturbation about the unperturbed state. The state is assumed to form directly by the three‐body association reaction involving atomic oxygen and also via quenching of the intermediate state . Our calculated values of <η> are compared with those derived from the observations of Zhang (1991) and Zhang et al. (1992a). These observations, which were obtained during the AIDA (Arecibo Initiative in Dynamics of the Atmosphere) Act ′89 campaign at Arecibo (18°N) during the period covering April 5 to May 9, 1989, were made with the MORTI (mesopause oxygen rotational temperature imager) instrument and provide us with values of both <η> and the horizontal wavelengths of the gravity waves. For all of our results the undisturbed mesosphere is defined by the model output of Garcia and Solomon (1985) for a latitude of 18° and for the months of March and June, these being most relevant to the observations of Zhang. For the evanescent waves, there is essentially good agreement between the theory and observations, while for the internal waves the agreement is best for |<η>|. However, the comparatively larger errors in the observed values of |<η>| make it impossible to decide which chemical scheme used in the model provides the best results. The errors associated with the observed phases of <η> are apparently minimal, and cannot account for the differences between the observed and modeled phases of <η>. The sensitivity of our modeled values of <η> to certain rate constants and the quenching parameters is investigated and discussed. In particular, we demonstrate that our modeled values of <η> are sensitive to the quenching effects of atomic oxygen and hence also to the production mechanism of the .