Gravity Wave Variances Research Articles

Seasonal variations of gravity waves revealed in rawinsonde data at Pohang, Korea

Seasonal variations of gravity wave characteristics are investigated using rawinsonde data observed at Pohang observatory, Korea (36°2′N, 129°23′E) during the one-year period of 1998. Analysis is carried out for two atmospheric layers representing the troposphere (2–9 km) and stratosphere (17–30 km). There exist clear seasonal variations in amplitudes of temperature and wind perturbations and wave energy in the stratosphere, with their maxima in wintertime and minima in summertime. A strong correlation is found between the wave activity and the strength of the jet stream, but there is no clear correlation between the wave activity and the vertical gradient of static stability. The intrinsic frequency and vertical and horizontal wavelengths of gravity waves in the stratosphere are 2f–3f, where f is the Coriolis parameter, and 2–3 km and 300–500 km, respectively. The intrinsic phase velocity directs westward in January and northeastward in July. The vertical flux of the stratospheric zonal momentum is mostly negative except in summertime with a maximum magnitude in January. Topography seems to be a major source of stratospheric gravity waves in wintertime. Convection can be a source of gravity waves in summertime, but it is required to know convective sources at nearby stations, due to their intermittency and locations relative to floating balloons.

Inter-annual variation of gravity waves in the Arctic and Antarctic winter middle atmosphere

High latitude, middle atmospheric (28–80 km) gravity wave observations from the Microwave Limb sounder (MLS) on the Upper Atmosphere Research Satellite (UARS) are available during five Arctic and Antarctic winters between 1991 and 1997. Newer data from NOAA Advanced Microwave Sounding Unit (AMSU) instruments have been introduced recently, in order to extend the gravity wave measurements from 1998 to present. Analyses of MLS and AMSU data show substantial inter-annual variability in both magnitude and spatial patterns of gravity wave activity at both Arctic and Antarctic latitudes.

Gravity waves in the equatorial MLT region

Observations of MF radar winds made at Christmas Island ( 2 ∘ N , 157 ∘ W ) between 1990 and 1997 are used to produce daily values of gravity wave variances in the equatorial mesosphere and lower thermosphere (70–100 km). Gravity wave activity is modulated on number of time scales, ranging from diurnal through seasonal to intraseasonal and interannual. A semiannual variation is particularly evident. In this paper we examine the possibility of using the wave variances in the MLT to estimate the momentum flux spectrum near the source height in the troposphere. To test our ideas we focus on modeling observations in April 1993 and April 1994, months which had very different background winds and tides. A linear gravity wave model is used in combination with ray-tracing techniques to investigate the propagation of wave spectra through representative zonal wind and temperature fields, including the effects of tidal winds and temperatures. The magnitude and diurnal variation of observed wave variance are used to constrain the momentum fluxes as a function of height and time. A number of source distributions were tested using this technique. It was found that the observations impose rather strong constraints on the source-level spectrum. A spectrum that has a bias to toward eastward propagating waves gave the best correspondence with observations. Our results have significance for the design and constraint of gravity wave parameterization schemes used in climate models.

Interannual and seasonal variations of diurnal tide, gravity wave, ozone, and water vapor as observed by MLS during 1991–1994

The diurnal tide in the mesosphere and lower thermosphere (MLT) shows large seasonal and interannual variations. Despite recent modeling investigations, the underlying physical mechanisms for causing these variations remain unclear. This paper provides further observational constraints to tide-sensitive variables (H 2O, O 3, and gravity wave variances) used by the models, which are obtained simultaneously by upper atmosphere research satellite microwave limb sounder at altitudes below the MLT region. The strong quasi biannual oscillation and semiannual oscillation variations in these measurements reveal good correlations between the diurnal tide with other tide-sensitive variables, which should be taken into account for further modeling studies.

Mesoscale gravity wave variances from AMSU‐A radiances

A variance analysis technique is developed here to extract gravity wave (GW) induced temperature fluctuations from NOAA AMSU‐A (Advanced Microwave Sounding Unit‐A) radiance measurements. By carefully removing the instrument/measurement noise, the algorithm can produce reliable GW variances with the minimum detectable value as small as 0.1 K2. Preliminary analyses with AMSU‐A data show GW variance maps in the stratosphere have very similar distributions to those found with the UARS MLS (Upper Atmosphere Research Satellite Microwave Limb Sounder). However, the AMSU‐A offers better horizontal and temporal resolution for observing regional GW variability, such as activity over sub‐Antarctic islands.

Interannual variations of the mean wind and gravity wave variances in the middle atmosphere over Hawaii

Using simple numerical filters, wind variances with time scales of 0.1 –1 h and 1 –5 h are estimated from medium frequency radar wind observations at altitudes 70 –94 km over Hawaii (22°N, 160°W) during the years 1990–2000. The results show interannual variations of the mean wind and variances. The variances can be attributed to the presence of atmospheric gravity waves (GWs). At altitudes below 80 km a maximum of GW intensity was observed between the years 1992 and 1996. An average seasonal cycle in wind variances is calculated by compositing over the 11 years of data. The monthly residual values, obtained after subtraction of the average seasonal variations show quasi-biennial (∼27 month) and longer period oscillations. Maximum variances are observed in 1993 and 1998–1999 at altitudes above ∼84 km . A correlation is found between GW wind variances and the Southern Oscillation Index, which in turn reflects El Niño activity in the tropical Pacific. However, more analysis with several El Niño events is required to establish the correlation with greater confidence.

Geographical distribution and interseasonal variability of tropical deep convection: UARS MLS observations and analyses

Tropical deep convection and its dynamical effect on the tropopause and stratosphere are investigated using a suite of data from the Upper Atmospheric Research Satellite (UARS) Microwave Limb Sounder (MLS), including upper tropospheric humidity, cloud radiance, and gravity wave measurements. For this purpose, geographical distributions of temperature, water vapor, and cloudiness in the tropical tropopause layer (TTL) are compared with corresponding maps of gravity wave variance in the stratosphere. In addition, ECMWF global wind divergent and velocity potential fields as well as NOAA outgoing longwave radiation and CMAP rainfall data are analyzed to help pinpoint the locations of deep convection. We found that high‐altitude clouds near the bottom of TTL (∼147 hPa) are usually surrounded by high‐humidity air, and their spatial pattern and seasonal variability are closely associated with regions of vigorous summertime deep convection. Upward propagating gravity waves generated from these convection regions are shifted poleward by prevailing stratospheric winds. We estimate that tropical deep convection lifts ∼5% of the cloud tops to altitudes above 100 hPa and that most of the extreme deep convection events occur in the Western Pacific and Indian monsoon regions. Low‐temperature regions in the TTL are associated with, but often drift away from, the center of deep convection. Regions of water vapor maxima near the bottom of TTL are located directly above the deep convection centers, but this moisture behavior is somewhat reversed at the top of the TTL. The integrated picture derived from this study implies that convective scale motions could be important in affecting short‐term dehydration processes in the TTL. Our results also suggest that the spatial organization and temporal development of tropical convective systems will be better monitored with the follow‐on Earth Observing System (EOS) Aura satellite instruments and lead to improved understanding of the complex interaction of tropical convection with large‐scale dynamic and thermodynamic conditions.

Modulation of gravity waves by planetary waves (2 and 16 d): observations with the North American-Pacific MLT-MFR radar network

In an earlier study based upon medium frequency radar (MFR) data from Saskatoon (52°N) the variability of time-sequences of gravity wave (GW) variances was linked to tidal (12,24 h) and planetary wave (2 d) oscillations of the wind. Fifty days of data were chosen from each of winter, spring and autumn seasons for this most comprehensive assessment of wave interactions. While modulations of the GW variances were observed, the results indicated considerable intermittency in the strength and direction of waves from the GW sources. Here we extend the study to other sites in the MLT-MFR (mesosphere, lower-thermosphere) network, and focus upon the planetary waves (PWs) and their modulating influences upon the GW variances. The PW events include the 2 d waves as seen at Saskatoon, London, Hawaii and Christmas Island during the summer of 1994; and 16 d wave activity evidenced at Saskatoon and London throughout 1994, and also during 12 and 7 years of observations, respectively. The modulations of the sequences of GW variances (10– 100/150 min , 2– 6 h periods) are significant, and the phase-differences between the PW (2 d,16 d) oscillations in the time-sequences of the winds and of the GW variances allow the propagation directions for the GW fluxes to be inferred. These are eastward in summer months and westward in winter at MLT (60– 90 km) altitudes.

Modulation of upper mesospheric temperature inversions due to tidal-gravity Wave interactions

Two nights of coincident measurements with the University of Western Ontario's MF radar and Purple Crow Lidar have been extensively analyzed to illustrate the possible effects due to tidal-gravity wave interactions on upper mesospheric inversion layers. On both nights an inversion layer in the upper mesosphere disappears as the westward component of the tide increases. Inversions similar in height and magnitude to these two nights, as seen in the nightly mean temperature profile, occur in 11% of the measurements made by the Purple Crow Lidar since 1994. Of these inversions, 76% show similar behavior to the two nights considered here in detail. The statistical results suggest the upper mesospheric inversion occur in two types. The first type, illustrated by the individual nights considered in this study, are associated with increased gravity wave variance as the magnitude of the westward tide decreases. These inversions are persistence over many hours. The second type of inversion has a similar nightly mean structure, but is due to shorter (on the order of <2 h ) periods of heating with temperature changes much greater than that expected from tidal dissipation.

Mean winds, tides, and gravity waves during the westward phase of the mesopause semiannual oscillation (MSAO)

MF radar observations of mean winds, tides, and gravity waves (in the period range 30–120 min) over Tirunelveli (8.7°N, 77.8°E) during April 1993, when the mesopause semiannual oscillation (MSAO) was in the westward phase, are presented in this work. The focus herein is on the observed relationship between these dynamical parameters, which is explained in the context of the current understanding of the behavior of MSAO winds, tidal dissipation, and gravity wave‐tide interactions. The observed diurnal tide amplitude and phase variations with height indicate (1,1) wave damping and subsequent westward momentum deposition. The gravity wave variances are shown to be modulated by the diurnal tide. Though the analysis does not lead to an estimation of gravity wave contributions to the time‐mean zonal flow and its shears, the observations do suggest that the short‐term tidal variability contributes, either directly, or indirectly through the gravity wave‐tidal interactions, to the variation of the strength of the time‐mean westward flow associated with the MSAO.

Seasonal variation of gravity wave activity at 5–20 km observed with VHF radar at White Sands Missile Range, New Mexico

Wind observations at altitudes from 5 to 20 km obtained from the VHF wind profiler radar at the White Sands Missile Range (WS), New Mexico, are used to construct a climatology of gravity wave intensities over WS during the nearly 6‐year period from January 1991 through September 1996. The kinetic energy densities for both high‐frequency gravity waves (ground‐based period of 6 min to 2 hours) and low‐frequency gravity waves (period from 2 to 21 hours) are considered. These results are compared to earlier results from the middle and upper atmosphere (MU) radar in Japan. Comparisons of gravity wave variability at high and low frequency reveal both similarities and differences between the two sites. The primary differences occur at high frequency. The seasonal cycle in the high‐frequency gravity waves is less pronounced and less regular at WS compared to MU. The magnitude of the high‐frequency kinetic energy density is generally larger at WS compared to MU, and the relationship between the upper tropospheric jet stream winds and the high‐frequency gravity wave activity is not nearly so strong at WS as at MU. The primary high‐frequency gravity wave excitation mechanism at WS appears to change with season. There is evidence of a connection in winter and spring between the wind speeds in the middle troposphere and the high‐frequency gravity wave activity, suggesting a topographic source for these waves. Strong summertime high‐frequency variability at WS in the upper troposphere is most likely linked to waves launched by deep convection during the monsoon season in New Mexico. The results for low‐frequency gravity waves are more consistent between WS and MU. Similarities appear in the kinetic energy density, the nature of the seasonal cycle, and the correlation between the kinetic energy density and the mean winds.

A comparison of mean winds and gravity wave activity in the northern and southern polar MLT

Mean winds and waves observed in the mesosphere and lower thermosphere with MF radars located at Davis (69°S, 78°E) and Poker Flat (65°N, 147°W) are compared. Measurements covering the period from 1999 to mid 2000 show differences in the strength of the horizontal wind fields. In the southern hemisphere the zonal and meridional winds reach their maximum values near the summer solstice, but are delayed by 2–3 weeks in the northern hemisphere. Gravity wave variances also show significant differences, as do the strength of vertical velocities.

Seasonal variation of gravity waves with various temporal and horizontal scales in the MLT region observed with radar and airglow imaging

Abstract Seasonal variations of activity and horizontal propagation direction of gravity waves in the MLT region at mid-latitude over Shigaraki, Japan (35 N) are investigated for various time and spatial scales using the MU radar observations and recent airglow imaging observations. Semiannual variation with solsticial maxima and equinoctial minima are generally observed for short period (

Comparison of global distributions of zonal‐mean gravity wave variance inferred from different satellite instruments

Gravity wave temperature fluctuations acquired by the CRISTA instrument are compared to previous estimates of zonal‐mean gravity wave temperature variance inferred from the LIMS, MLS and GPS/MET satellite instruments during northern winter. Careful attention is paid to the range of vertical wavelengths resolved by each instrument. Good agreement between CRISTA data and previously published results from LIMS, MLS and GPS/MET are found. Key latitudinal features in these variances are consistent with previous findings from ground‐based measurements and some simple models. We conclude that all four satellite instruments provide reliable global data on zonal‐mean gravity wave temperature fluctuations throughout the middle atmosphere.

Seasonal variation of gravity waves observed with an OH CCD imager at Shigaraki (35°N,136°E), Japan

A wide field CCD imager (FOV: diagonal 128°, N-S 89°, E-W 88°) for OH airglow observation was installed at the MU radar site in Shigaraki (34.9°N, 136.1°E) in November 1996, and has been operated continuously since January 29, 1997. From the observational data in 161 nights from November 1996 to May 1998, gravity wave patterns shown in the OH images are analyzed. Typical horizontal wavelength, horizontal phase speed and observed period are about 5–50 km, 15–70m/s and 5–20 minutes, respectively. Seasonal variations of the parameters of these short period or small scale gravity waves are investigated. For the waves with horizontal wavelength larger than 17.5km, horizontal propagation direction shows clear seasonal change of eastward propagation in summer and westward in winter with a transition in the middle of March and the end of September. These results are consistent with that the horizontal propagation directions of these gravity waves are strongly affected by filtering effect due to the mean wind in the middle atmosphere. On the other hand, for the waves with horizontal wavelength smaller than 17.5km, seasonal variations of these parameters are not clear. Therefore, these waves are probably generated in situ in the mesosphere.

Interpretations of observed climatological patterns in stratospheric gravity wave variance

Observational analyses of gravity waves in the stratosphere have revealed various climatological patterns in gravity wave activity. Seasonal, geographical, and vertical variations have all been observed. In this work, a linear model of gravity wave propagation is applied to investigate the underlying causes of some of the observed patterns. A collection of monochromatic gravity waves that represent a broad spectrum of wavenumbers and frequencies is input at 6‐km altitude in the model. Propagation of the waves through realistic background atmospheric wind and stability fields is treated with linear ray theory and a simple saturation condition to limit amplitudes to stable values. The wave spectrum at the 6‐km source height is specified to be constant at all latitudes, longitudes, and times, so the variability that appears at higher altitudes is due entirely to background atmosphere variations. Before the model results are compared to the observations, the spectrum of waves is filtered in a way that mimics the limitations of each of the observation techniques. The filtering is described in terms of vertical wavelength and is referred to as the “observational filter.” In a vertically varying background wind, gravity waves are Doppler‐shifted in intrinsic frequency and refracted to different vertical wavelengths as they propagate vertically through the atmosphere. The observational filter and the wave refraction effects can thus couple in interesting ways that have not been explicitly considered in previous work. The model shows that this coupling can give rise to geographical, seasonal, and vertical variations in gravity wave observations without any variations in the spectrum or amplitude of gravity wave sources in the troposphere. Thus careful consideration of both the background wind profile and observational filter can greatly affect the interpretation of the observed climatological patterns in gravity wave activity.

Intraseasonal wind variability in the equatorial mesosphere and lower thermosphere: long-term observations from the central Pacific

Analysis of intraseasonal (10–100 days) oscillations in the equatorial mesosphere and lower thermosphere (MLT) is presented, based on over five years of velocity data acquired by a radar system at Christmas Island (2 °N, 157 °W), in the central Pacific. Strong peaks in the zonal winds are found at periods of ~60 days, ~35–40 days, and ~22–25 days. These peaks, as well as the mean annual variations of the activity within the various period ranges, are similar to 30–60 day and 20–25 day oscillations that occur in the equatorial troposphere. Weaker (but nonetheless clear) periodicities are also found in the meridional winds at ~60 days and ~35 days. A strong quasi-60-day variation is detected in gravity-wave variances, with much weaker signals at ~40 days and ~25 days. Strong variations in diurnal tidal amplitudes are observed with periods of ~60 days, ~40 days, and ~25 days.

A three‐year gravity wave climatology of the mesosphere and lower thermosphere over Kauai

A 3‐year study of gravity wave variability covering 1991, 1992, and 1993 was conducted using horizontal wind measurements collected by a partial reflection MF radar located on the island of Kauai, Hawaii (22 N, 160 W). High echo return rates between 70 and 98 km provided ample time resolution for the study of gravity wave variability in this height range. Horizontal wind measurements were band pass filtered for periods between 20 min and 6 hours, establishing a gravity wave band time series. Gravity wave variance and polarization calculations were carried out over various time frames and examined with regard to their response to season and background environment. Among the more interesting observed phenomena were the annual and semiannual oscillations in the gravity wave variances, a significant diurnal and semidiurnal variability in the variances, and an extended enhancement of these variances during the latter half of 1992 and the beginning of 1993.

Gravity Wave Variance in LIMS Temperatures. Part II: Comparison with the Zonal-Mean Momentum Balance

Abstract Zonal-mean gravity wave variance in the Limb Infrared Monitor of the Stratosphere (LIMS) temperature data is seen to correlate strongly with the residual term in the LIMS zonal-mean momentum budget throughout much of the observed mesosphere. This momentum residual is attributed to gravity wave momentum transport at scales that cannot be directly sampled by the LIMS instrument Correlation is highest in the vicinity of the fall and winter mesospheric jets, where both gravity wave variance and momentum residual reach their largest values. Correlation is also high in the Southern Hemisphere subtropical mesophere, where gravity wave variance and the momentum residual have broad temporal maxima during the easterly acceleration of the stratopause semi-annual oscillation (SAO). This subtropical correlation has important implications for the SAO eastward acceleration, which several studies suggest is forced by gravity wave momentum flux divergence. Correlation between gravity wave variance and inferred gr...

Gravity Wave Variability and Interaction with Lower-Frequency Motions in the Mesosphere and Lower Thermosphere over Hawaii

Radar observations of the mesosphere and lower thermosphere over Kauai, Hawaii, have provided measurements of the horizontal wind field since September 1990. Horizontal velocity variances in the gravity wave band of frequencies were computed for several averaging intervals over the entire dataset. The results, correlated with tidal and lower-frequency winds, show evidence of filtering of gravity waves by tidal and other lower-frequency motions. Spectra of gravity wave variances for two years of data display peaks at planetary wave as well as tidal periods, primarily near 12-, 24-, 48-h, and 16-day periods. The modulation of gravity wave variance during four distinct background regimes is examined: one in which the diurnal tide predominates; another in which the 2-day wave is large, a third in which the zonal mean wind and tidal amplitudes are large; and a fourth in which the zonal mean, tidal, and 2-day amplitudes are all small. Intervals in which the diurnal tide or the 2-day wave predominates show evidence of gravity wave modulation at semidiurnal, diurnal, and 2-day periods. Intervals in which the mean wind and lower-frequency amplitudes are small have much reduced gravity wave activity relative to intervals when they are large.