On Krassovsky's ratio for ter-diurnal hydroxyl oscillations in the winter polar mesopause

Past observations of ter-diurnal oscillations in hydroxyl brightness and rotational temperature of the winter polar mesopause have revealed two apparently conflicting facets. The 8 h wave is a commonly observed feature, possibly the most recurring tide of that region. On the other hand, temperature changes were observed to lead changes in brightness, implying a negative phase of Krassovsky's ratio. (Krassovsky's ratio is the ratio of normalized airglow brightness variation to normalized temperature variation.) This qualitatively supports a gravity wave explanation according to predictions of current dynamical-chemical models of fluctuations in airglow emissions. But the ter-diurnal oscillations are not likely due to gravity waves because of the persistence of the oscillations. Here a statistically-significant sample (23 individual days) of clear ter-diurnal waves of the winter polar mesopause over Eureka, NWT (80°N), is considered in order to establish the range of observed amplitude and phase of Krassovsky's ratio and compare with the predicted values. The observed amplitude of Krassovsky's ratio was 2.6–8.1 with the majority of cases (74%) at 4±1. The phase of Krassovsky's ratio was negative for all 23 cases, i.e. changes in temperature led changes in brightness. In the majority of cases the negative phase was significantly different from zero. Theoretical calculations for two different model atmospheres (both isothermal and non-isothermal) were performed using a current model of tidally driven hydroxyl airglow fluctuations. The predicted phase of Krassovsky's ratio was slightly positive (≤ +30°) for an 8 h gravitational zonally symmetric tide of meridional index 2–5 and a migrating tide of meridional index 3–7. It is concluded that the theoretical predictions of Krassovsky's ratio of oscillations in hydroxyl emission induced by zonally symmetric tides and migrating tides as applied to the two model atmospheres do not agree with the above observations of the winter polar mesopause. Neither do the theoretical predictions of Krassovsky's ratio for gravity wave-driven fluctuations in OH nightglow from an extended, dissipative emission region match the observed value for the vast majority (87%) of days with clear ter-diurnal oscillations. The consistently negative phases over Eureka probably reflect unique conditions in the very high latitude winter mesopause region characterized by strong lapse rates in the emission layer or background minor constituent con