No abstract

In the nighttime equatorial F region, zonal neutral air winds and gravity waves propagating across the magnetic field tend to drive the ionization as if there were no field. When the phase trace speed of a gravity wave equals the drift speed of the ionization, strong ionization perturbations may occur (Whitehead's spatial resonance effect). The resonance effect is described by a parametric differential equation. Without diffusion, it yields perturbations in the form of periodic δ functions. Therefore a coefficient of turbulent diffusion is introduced to limit the perturbation amplitude. Then the equation can be solved in terms of partial Fourier sums. Numerical calculations with reasonable F region parameters indicate that under spatial resonance conditions, gravity waves cause considerable secondary flows of ionization as a result of nonlinear effects. The contours of constant ionization density show steep or even breaking wave fronts similar to those observed by HF radar and radio propagation experiments in the large‐scale structure of the equatorial spread F. The spatial resonance mechanism produces field‐aligned tubes of decreased ionization density which may be the primary cause of ‘plasma bubbles’ detected by satellite and rocket probes.

We present evidence that shows some aspects of the global atmospheric dynamic responses to the eruption of Mount St. Helens on May 18, 1980. Although events such as volcanic eruptions may excite a number of acoustic‐gravity wave modes in the atmosphere, the observed surface pressure perturbations and distant ionospheric perturbations can be explained only in terms of propagation of Lamb modes with a horizontal propagation velocity slightly above 300 m/s. Results from model computations show good agreements with the observational data. Ground level pressure perturbations created by this event are only slightly smaller than those created by the historical Great Siberian Meteor.

A timescale analysis indicates that at noctilucent cloud (NLC) heights near 83 km, the growth and decay of ice particles and their transport by tide-and gravity-wave-associated vertical velocity components are the dominant processes that determine the particle size distribution. Other processes including sedimentation are of minor importance, so that the formation of NLCs should, in general, not depend on atmospheric conditions at the mesopause, in agreement with recent findings from simultaneous lidar and rocket experiments. Then a simple NLC model can be constructed consisting essentially of a partial differential equation which describes the temporal behavior of the particle size distribution in particle parcels moving with the wave-associated air velocity. Contrary to previous ones, the present model yields a steady state oscillation within an integration time of 1 day because the optically active ice particles have lifetimes of a few hours only. NLC simulations yield particle concentrations, mean radii. and scatter ratios which are in good agreement with observational results. The model also predicts an occasional coexistence of small populations of optically active particles and large populations of microscopic particles, which may be of basic importance for the interpretation of recent observations indicating a more or less tight coupling between NLCs and simultaneously occurring polar mesosphere summer echoes.[1980], pure ice crystals are assumed to result from ionrecombination nucleation, and mixed ice-dust particles Copyright 1998 by the American Geophysical Union. Paper number 98JD02070. 0148-0227 / 98 / 98J D-02070 $ 09.00 form via water vapor deposition on dust cores and coagulation of pure ice crystals with dust. Further models have been described by Jensen and Thomas [1988, 1994] and Jensen et al. [1989]. They are based essentially on the model of Turco et al. but contain considerable simplifying assumptions. A common feature of all these models is that for different reasons the simulated NLCs depend strongly on the mesopause temperature minimum. In the model of Turco et al. [1982] this is due to the fact that the computed NLC height almost coincides with the temperature minimum, which is assumed to be about 5 km below the actual temperature minimum measured by L•'bken and yon Zahn [1991] in the summer months at a latitude of 69øN. As at consequence, the model predicts an enhancement of the ion nucleation rate with decreasing mesopause temperature, which in turn yields larger particle concentrations but smaller particle sizes because of the limited wetter vapor supply. In the models of Jensen et al. [1989], on the other hand, the mesopause is assumed at heights above 87 km and the strong dependence of simulated NLCs on the temperature minimum results from the assumption that the nucleation of NLC particles occurs mainly, but not exclusively, where the air is coldest. Recent studies of Hansen and yon Zahn [1994] and La'bken et al. [1996] on NLCs and the thermal structure in the mesopause region base...