SA54A-01 INVITED
Gravity Wave Breaking and Turbulence in the MLT: Implications for Transport, Diffusion, and the Turbulent Prandtl Number
Gravity waves (GWs) account for the majority of turbulence and transport in the MLT due to their amplitude
growth with altitude and instabilities that occur for all wave amplitudes and intrinsic frequencies. Linear theory
provides a useful guide to initial instability structures and growth rates, and also to likely consequences of
turbulent mixing. Neither linear theory nor observations can readily address all of the contributions to mixing
and transport in an evolving 3D motion field, however, and we must seek guidance from numerical studies
that define these flows with high resolution and precision. Given evidence for large-amplitude GWs in the
MLT, we have performed direct numerical simulations (DNS) that resolve the full GW and turbulence
spectrum for GW parameters and Reynolds numbers relevant to the MLT. These simulations exhibit rapid
instability development, a competition between 2D and 3D instability modes, large GW amplitude reductions,
and spatially-localized turbulence. Turbulence is strongly correlated with the GW phase and highly variable in
intensity. Both 2D GWs and 3D turbulence contribute to fluxes of heat and momentum, but the contributions
are often of opposite sign. Turbulent mixing is strong, but impacts the velocity and thermal fields to different
degrees. The turbulent Prandtl number, a key quantity in GCM descriptions of unresolved small-scale
dynamics, is assessed for breaking GWs having subcritical and supercritical amplitudes and is found to be Pr
~ 3 – 8 during active breaking, confirming expectations of earlier linear theory and consistent with values
required by large-scale models.
http://www.cora.nwra.com/dave/GWBmovies.html
SA54A-02 INVITED
Estimation of the vertical diffusion coefficient in MLT region from the migrating diurnal tides observed by TIMED
We use TIMED global temperature and wind data from 2002-2007 to investigate the damping of the diurnal tide. Horizontal winds are measured by TIDI and are calculated from SABER temperature measurements. The SABER derivation of tidal wind depends on the damping and, therefore, the discrepancy between TIDI and SABER determinations of the wind tides can be used to determine the damping. For example, the amplitude of diurnal tide of the meridional wind exhibits very obvious asymmetries between the northern and southern hemispheres. The estimated vertical diffusion coefficient is that which gives the best agreement between the SABER and TIDI diurnal tide winds. The results show that there can be very large diffusion coefficients in very narrow vertical regions. The peak of the diffusion coefficient is at higher altitude in summer than in winter in both hemispheres. The derived diffusion coefficients exhibit obvious asymmetries between the northern and the southern hemispheres. In both hemispheres, the diffusion coefficients are larger at 40 than at 20 degrees latitude.
SA54A-03
Eddy turbulence and the double mesopause
We study heating and cooling in the upper mesosphere due to dissipation of gravity waves as described by eddy turbulence, eddy conductivity, and chemical heating and cooling. We estimate the role of these processes in seasonal variations of the temperature in the upper mesosphere and in the production of the double mesopause. Also, we model the impact of these processes on atomic oxygen density in conjunction with published experimental data. An additional temperature maximum in the upper mesosphere (the 'double mesopause') can be produced due to heating by eddy turbulence. According to our model, there is a close connection between the temperature peak in the mesosphere and the O density distribution. A rocket experiment confirms this connection. Applying our model to the case, a breaking gravity wave observed during the rocket campaign gives excellent agreement with both the temperature and oxygen profiles observed. The splitting of the O density profiles can occur in fall and spring seasons when the double mesopause is observed.
SA54A-04
Dissipation Mechanisms for Ducted Gravity Waves in the Mesosphere and Lower Thermosphere
Gravity waves of short period (4-12 minute) and small scale (15-40 km horizontal wavelength) are frequently observed in mesopause airglow imaging experiments. The propagation of small-scale gravity waves, however, is strongly affected by local background structure and dynamics associated with large-scale waves and tides, and existing background conditions [Fritts et al., JASTP, 68, 247, 2006]. Vertical variations of temperature or wind can also produce ducts, in which trapped waves are frequently observed to propagate [e.g., Isler et al., JGR, 102(D22), 26301, 1997; Walterscheid et al., JASTP, 61, 461, 1999]; spatially- periodic structure of large-scale waves and tides may lead to formation of alternating layers of evanescence and ducted propagation [e.g., Snively et al., JGR, 112, A03304, 2007; Fritts and Janches, JGR, 113, D05112, 2008]. At higher altitudes, and for larger wave magnitudes, interactions between the small and large scales become increasingly nonlinear. Small-scale perturbations to large-scale wave fields may contribute to the formation of dynamic or convective instabilities [e.g., Fritts and Alexander, RG, 41(1), 1003, 2003, and references cited therein]; strong wind flows may also provide critical layers to some portion of the wave spectrum that would otherwise be propagating or ducted. Ducted or evanescent waves can produce strong airglow signatures, due to long vertical wavelengths and strong induced vertical fluid perturbations [Hines and Tarasick, GRL, 21(24), 2729, 1994], and form a significant fraction of observed waves in airglow data. Although commonly assumed to be horizontally- propagating, waves that appear locally trapped may still be subject to vertical tunneling [e.g., Sutherland and Yewchuk, JFM, 511, 125, 2004, and references cited therein]. Stable ducted waves may therefore tunnel into adjacent ducts, or into regions that may include critical layers, or alternatively facilitate instability or breaking, leading to turbulence and mixing. These processes may contribute to slow dissipation of the stored wave energy. Using numerical and analytical models, we investigate case studies of ducted wave dissipation under MLT conditions.
SA54A-05 INVITED
In-Situ Turbulence Measurements in the High-Latitude MLT-Region
Since the beginning of the nineties our research group has launched a total of 40 sounding rockets carrying ionization gauges for the high resolution measurement of neutral density fluctuations. These fluctuations are a suitable tracer for turbulent velocity fluctuations. Since such measurements are made at very high spatial resolution (< 1 m) they can be used to derive the spectral content of the turbulence field from which in turn the turbulent energy dissipation rate can be reliably derived. After a short introductuction to the instrumental and analysis technique we will present mean results of the seasonal and geographical variation of turbulence at high northern latitudes, insights into the mechanism of wave mean flow interaction, and discuss selected instability events. We will further show that modern analysis techniques using wavelet transforms enable us to characterize the actual vertical extent of turbulence layers which can be extremely narrow (order of 100 m). From our measurements we can also tentatively characterize the statistics of turbulence in the MLT indicating that turbulence is highly intermittent. Comparing these results to model estimates from a high resolution GCM, we suggest that this intermittency partly reflects the gravity wave dynamics of the MLT. Future questions to be addressed are among others the relation of turbulent energy dissipation to turbulent mixing and the horizontal structure of turbulent cells.
SA54A-06
Estimation of eddy momentum and thermal diffusivities from lidar measured gravity wave momentum and heat fluxes
Momentum and heat fluxes derived from high-resolution lidar measurements of wind and temperature in the mesopause region at Starfire Optical Range (35N, 106W) were used to derive eddy momentum and thermal diffusivities based on linear parameterization theory. Both seasonal and altitude variation of these diffusion coefficients were derived. It is found that both coefficients vary with season with minimum in equinoxes and maximum in solstices. They also vary with altitude, with kinetic diffusion increasing and thermal diffusion decreasing from 85 to 100 km.
SA54A-07
Enhanced eddy diffusion and large eddies above the nominal turbopause height
Eddy diffusivity measurements from sounding rocket chemical releases have been very limited but have shown evidence for enhanced diffusion above the nominal turbopause height in both cases in which detailed analyses have been carried out. Specifically, enhanced diffusion rates, larger than the molecular diffusion rate, were found up to altitudes above 110 km. We present a summary of the earlier results and argue that the large eddies that are frequently observed with resonance lidars in the altitude range above 100 km may contribute significantly to the turbulent diffusion and transport in that altitude range. Specifically, the overturning features have characteristic time scales of several hours and vertical scales greater than 2-3 km and are qualitatively similar to the so-called "large eddies" found in the planetary boundary layer (PBL). The latter are known to contribute significantly to the eddy diffusive transport in the lower atmosphere. We compare the overturning feature characteristics in the mesosphere and lower thermosphere (MLT) region with those in the PBL and discuss the implications for eddy diffusive transport in the upper atmosphere.
SA54A-08 INVITED
Impact of Eddy Diffusivity on Seasonal Variations of the Thermosphere
Thermospheric neutral density and composition exhibit a strong seasonal variation, with maxima near the equinoxes, a primary minimum during northern hemisphere summer, and a secondary minimum during southern hemisphere summer. This pattern of variation is described by thermospheric empirical models. However, the mechanisms are not well understood. The annual insolation variation due to the Sun-Earth distance can cause an annual variation; large-scale inter-hemispheric circulation can cause a global semiannual variation; and geomagnetic activity can also have a small contribution to the semiannual amplitude. However, simulations by the NCAR Thermosphere-Ionosphere-Electrodynamics General Circulation Model (TIE-GCM) indicate that these seasonal effects do not fully account for the observed annual/semiannual amplitude, primarily due to the lack of a minimum during northern hemisphere summer. A candidate for causing this variation is change in composition, driven by eddy mixing in the mesopause region. Other observations and model studies suggest that eddy diffusion in the mesopause region has a strong seasonal variation, with eddy diffusion larger during solstices than equinoxes, and stronger turbulence in summer than in winter. A seasonal variation of eddy diffusion compatible with this description is obtained. Simulations show that when this function is imposed at the lower boundary of the TIE-GCM, neutral density variation consistent with satellite drag data, and O/N2 consistent with measurements by TIMED/GUVI, are obtained. These model-data comparisons and analyses indicate that turbulent mixing in the mesopause region may contribute to seasonal variation in the thermosphere, particularly the asymmetry between solstices that cannot be explained by other mechanisms.
SA54A-09
Is OH in the Upper Mesosphere a Sensitive Indicator for Eddy Diffusion? New Insights From SHIMMER Data
The Spatial Heterodyne Imager for Mesospheric Radicals (SHIMMER) on STPSat-1 allows the observation of the diurnal variation of mesospheric hydroxyl (OH) with high vertical resolution. In our initial study, we compared the OH diurnal cycle measured in the summer of 2007 around 55 degrees north with a one dimensional photochemical model calculation without taking into account vertical transport. We found excellent agreement at 74km, but significant differences at higher altitudes. Sensitivity studies we have done suggest that inclusion of vertical transport, as given by an eddy diffusion (Kzz) coefficient, will change the shape of the calculated OH diurnal cycle. Here we explore this further using both model and data and address the question as to whether OH could be used as a diagnostic for Kzz. Calculations from the WACCM (Whole Atmosphere Community Climate Model) suggest significant latitudinal variation in Kzz. This suggests that SHIMMER data at tropical latitudes might provide an important constraint. In addition, the model data OH discrepancy is reflected in a similar (but opposite) discrepancy with TIMED/SABER ozone data. Given the well known anticorrelation between odd oxygen and odd hydrogen, this is an important validation of our conclusions derived from SHIMMER. Ultimately, we seek to provide a self-consistent model formulation which is consistent with SHIMMER OH, SABER O3 and WACCM gravity wave specifications.
SA54A-10
Model Case Study of Convectively Generated Waves in the Maritime Continent and Northern Australia
The Maritime Continent and northern Australia is a complex region incorporating a myriad small and large islands, narrow peninsulas, inner seas, and surrounding oceanic and continental areas . As a consequence of its location and unique geography the Maritime Continent-north Australia is extremely convectively active, and has been referred to as the boiler-box of the atmosphere. Deep convection generates a broad spectrum of waves that can propagate to high altitudes and attain wave breaking amplitudes. The enhancement of eddy diffusivity in the summer is thought to be linked to increased convective activity. We present a case study of convectively-generated waves over the Maritime Continent as simulated by a regional model. The characteristics of the waves and sources on time scales ranging from hourly to the diurnal and on spatial scales ranging from ~10-500 km are presented. The implications of this case study for our understanding of eddy diffusivity on shorter time and spatial scales is discussed.