Search

The effects of Doppler shifting caused by the mean wind field on the separability of gravity wave spectra are analyzed using the gravity wave models developed recently by Gardner et al. (1993). If the intrinsic joint vertical wave number (m) and temporal frequency (ω) spectrum is separable, Doppler effects associated with even small mean winds will destroy separability of the observed joint (m, ω0) spectrum, particularly at high vertical wave numbers. The effect can have a substantial impact on the measured m-spectrum especially when the observations are not sensitive to the lowest or highest observed frequencies. Data averaging or low-pass temporal filtering can result in substantial reduction of the spectrum magnitude at high m values while data prewhitening or high-pass temporal filtering can result in significant distortion of the spectrum at low m values. These results have important implications for observational tests of the various gravity wave spectral models.

Gravity wave models for the horizontal wave number spectra of atmospheric velocity and density fluctuations are derived by assuming that both saturated and unsaturated waves obey the polarization and dispersion relations and that the joint vertical wave number (m) and temporal frequency (ω) spectrum is separable. In the wave number region h* = ƒm*/N ≤ h ≤ m*, where h = (k2 + l2)1/2, k is the zonal wave number, l is the meridional wave number, ƒ is the inertial frequency, N is the buoyancy frequency, and m* is vertical wave number of the largest scale saturated wave, the model predicts that the two-dimensional isotropic spectrum is proportional to h−(p+1) and the one-dimensional spectra are proportional to k−p and l−p when the temporal spectrum is proportional to ω−p. In the saturation region h, k, l, m > m*, the spectra are proportional to h−(q+1), k−q and l−q when the saturated m spectrum is proportional to m−q. The joint (k, l, m) and (k, l, ω) model spectra are not separable. The models are consistent with existing observations of horizontal wave number spectra in the lower stratosphere and upper mesosphere.

The UIUC CEDAR Rayleigh/Na lidar was operated for approximately 160 h on 30 nights in January, March, and April 1989 at the Arecibo Observatory (18°N, 67°W) as part of the AIDA Act '89 Campaign. During this period 38 quasi-monochromatic gravity waves were observed in the stratopause region (25-55 km) and 62 waves were observed in the mesopause region (80–105 km). The event rates in both regions are approximately half those observed at the midlatitude site of Urbana. The characteristics of the waves in both regions are similar. Measured vertical wavelengths range from 1.1 to 17 km, vertical phase velocities from −6 to −270 cm s−1, observed periods from 5 min to 65 h, and amplitudes (relative atmospheric density variations) 0.4% to 17%. The wave amplitudes in the stratopause region are on average half the values for waves in the mesopause region with similar periods and vertical wavelengths. Vertical wavenumber spectra of density perturbations in both regions exhibit power-law dependencies with slo...

The airborne Na lidar and Christmas Island MF radar were used to measure the spectra of gravity wave density and wind perturbations during ALOHA-90 on the 25 March mission. Wave activity was especially strong near Christmas Island with measured wind variances between 1,000 and 1,400 (m/s){sup 2}. The vertical and meridional wave number spectra of horizontal wind perturbations exhibited power law shapes with slopes of {minus}3.17 and {minus}2.22, respectively. The temporal frequency spectrum appeared to be influenced by strong Doppler effects and had a shallow slope near {minus}1. The magnitudes of the spectra were 4.9 {times} 10{sup 5} (m/s){sup 2}/(cyc/m) at the vertical wave number 2{pi}/(4 km), 8.4 {times} 10{sup 6} (m/s){sup 2}/(cyc/m) at the meridional wave number 2{pi}/(200 km) and 7 {times} 10{sup 5} (m/s){sup 2}/(cyc/m) at the temporal frequency 2{pi}/(1 h).

We present vertical and horizontal wave number spectra of density perturbations in the upper stratosphere (25–40 km) and the upper mesosphere (∼80–105 km) inferred from ∼45 hours of airborne Na/Rayleigh lidar observations in the vicinity of Hawaii. The mean density variance is (0.92%)2 for the upper stratosphere and (6.1%)2 and for the upper mesosphere. The mean vertical shear variance and Richardson's number for the upper mesosphere are (6.1%/km)2 and 0.69, respectively. The observed vertical wave number m and horizontal wave number k spectra exhibit power law dependencies. The mean slopes of the vertical wave number spectra are −2.5 for the upper stratosphere and −3.1 for the upper mesosphere. The mean slope of the horizontal wave number spectra is −1.9 for the upper mesosphere. The mean characteristic vertical and horizontal wavelengths inferred from the spectra are 13.4 and 5810 km, respectively. In all cases the magnitudes of the upper stratosphere m spectra are more than an order of magnitude smaller than those of the upper mesosphere at all observable scales. In the stratosphere the m spectra exhibit significant energy at low wave numbers that are less than the values expected for the characteristic wave numbers. This suggests the presence of gravity wave sources capable of generating significant energy at long vertical scales. The source is believed to be geostrophic adjustment of the jet stream. A large-amplitude, highly coherent, quasi-monochromatic, short horizontal scale (λh ∼ 16 km) oscillation in the Na layer was observed on the March 22 flight. The oscillation resembled short horizontal scale structures often observed in OH airglow images and was found to be consistent with a gravity wave interpretation.