| description abstract | Eastward propagating disturbances over the equator are diagnosed in two independent Nimbus-7 LIMS (Limb Infrared Monitor of the Stratosphere) data sets. They are evident consistently at several pressure levels throughout the stratosphere and account for much of the temperature variance in the tropics. The disturbances, which can be seen in wavenumbers 1?3, are in phase and symmetric about the equator, latitudinally evanescent, and have short-moderate vertical phase structure, 10?40 km, which progresses downward. Wavenumber 1 has spectral components which propagate eastward at periods of 6.7?8.6 days (54?69 m s?1) and 3.5?4.0 days (115?135 m s?1). Wavenumber 2 exhibits eastward propagating variance at periods of 6.0?7.5 days (31?39 m s?1) and 3.8?4.3 days (55?62 m s?1). The faster waves appear principally in the upper stratosphere. These features are in reasonable agreement with the structure and dispersion characteristics of simple, quasi-separable Kelvin modes. With the exception of the slower wavenumber 1 feature, reported earlier by Hirota, these components are newly documented for the middle and upper stratosphere. Interpretations of wave structure in terms of refractive properties of the basic flow are supported by the zonal-mean winds for the period. Power structures exhibit several maxima and minima in height, with phase variations across the maxima slower than across the minima. This behavior, supported by the longer vertical wavelengths, suggests that some reflection may be occurring. A rapid phase variation is evident in both wavenumbers 1 and 2 near the stratopause, overlying a region of magnified amplitude. The latitudinal structure at this level, can be seen to contract as well. Such behavior is suggestive of disturbance focusing, due to Doppler shifting to small intrinsic frequencies, and attendant wave absorption. Coincident with this region of enhanced power and steep phase tilt, is a layer of sharp westerly shear, which, as reported by Leovy and others, descends over a period of weeks. The concurrent observation of the two phenomena supports earlier suggestions that Kelvin waves are instrumental in the westerly acceleration of the semiannual oscillation. | |
