SMALL-SCALE FLOW, TURBULANCE AND MIXING (IAPSO,

IAVCEI, IAMAS)

Location: Watson Building G23 LTA

Location of Poster: Bridge Poynting/Watson

Friday 30 July AM

Presiding Chair: Peter G. Baines (CSIRO Atmospheric Research, Aspendale, Australia)

JSP49/E/02-B5 Invited 0830

TIDAL INTERNAL WAVES GENERATION OVER SUBMARINE RIDGES, ENERGY TRANSFER TO HIGH FREQUENCY INTERNAL WAVES AND TURBULENCE

Eugene Morozov (Shirshov Institute of Oceanology, 36, Nakhimovsky st., 117851, Moscow, Russia, e-mail: internal@redline.ru)

We analyze tidal internal waves generated near the Emperor Ridge in the North Pacific. Internal waves are studied on the basis of data of Russian and US moorings set from both sides of the ridge. A great amount of CTD hydrographic measurements are also available in the region. This allows us to analyze the oceanic turbulence and fine structure which occurs in the ocean due to breaking internal waves. The stations near the ridge show well expressed fine structure with homogeneous steps of tens of meters, the distant stations are more smooth and demonstrate no finestructure. Calculations of spectra from these vertical profiles give different levels of spectral densities. Those near the ridge are significantly above the GM background dropped background spectra. We interpret this as breaking of intensive tidal internal waves, generation of high frequency internal waves, and formation of fine structure. The results of the observations are confirmed by a non-linear numerical model of the generation of tidal internal waves over the ridge. The model describes their propagation and dissipation with generation of short period internal wave trains and turbulence. It is shown that internal tides propagating from submarine ridges strain the vertical structure of hydrographic fields, which can be registered by usual CTD-profiling. Analysis of CTD profiles over great squares in the oceanic basins with account for these properties can indicate areas of intermittent turbulence in the ocean which is caused by breaking internal waves. This analysis of standard oceanographic data is less expensive than special measurements which include deployment of moored buoys in the ocean.

JSP49/E/03-B5 0850

REGULAR BURSTS IN THE SEDIMENT-STRATIFIED, OSCILLATING BOTTOM BOUNDARY LAYER: MODEL RESULTS

Boris A. KAGAN and Konstantin B. Utkin (both at Shirshov Institute of Oceanology, Russian Academy of Sciences, St.Petersburg Branch, 30 Pervaya Liniya, 199053 St.Petersburg, Russia, email: kgn@GK3103.spb.edu)

A model based on the 2.5-level turbulence closure scheme and on a parameterization of the processes of sedimentation and entrainment at the top of the bed-load layer is employed to simulate the vertical structure of the sediment-stratified, oscillating bottom boundary layer (BBL). The model predicts regular bursts arising during an oscillation cycle every time when, under certain conditions, entrainment of suspended particles from the bed-load layer is terminated. The regular bursts differ from the random bursts associated with the intermittency of small-scale turbulence in the viscous and buffer sublayers in many respects. In particular, the former are due to the interaction between hydrodynamic and sedimentological processes; more specifically, to the release of the turbulent kinetic energy expended for maintaining particles in suspension. Also, unlike the random bursts, the regular ones do not produce significant contributions to the Reynolds stress, and their timescale is a function of the suspension Froude number, the Reynolds number, the nondimensional critical friction velocity at which sediment particles begin to go into suspension, the relative settling velocity of suspended particles and the Strouhal number. For fixed values of the first four above-mentioned parameters, the regular burst timescale is the smaller the greater is the Strouhal number or, all other factors being the same, the greater is the oscillation frequency. That is apparently why no bursts have been observed in the BBL beneath wind waves, whereas in the tidally induced BBL these have become common.

JSP49/W/04-B5 0910

SEAMOUNT TURBULENCE: GENERATION AND DECAY

Iossif Lozovatsky (P.P. Shirshov Institute of Oceanology, Russian Academy of Sciences Moscow, 117851, Russia; also at Department of Mechanical and Aerospace Engineering, Arizona State University, Tempe, AZ, 85287-9809, USA, i.lozovatsky@asu.edu

Turbulent measurements above summits and at the rims of seamounts Erving and Ampere in Eastern Atlantic revealed an enormous enhancement of turbulent mixing with the mean kinetic energy dissipation rate of = 6x10-8 W/kg. The mean eddy and scalar diffusivities are in the range of (3.8-5.1)x10-4 m2/s with 90% confidence limits of (3.1-6.1)x10-4 m2/s. This implies that the averaged mixing rate in a 60-mile diameter and 150 m high region surrounding seamount Erving is 30-60 times than that of the far field value. At the place of generation, the diffusivities, averaged over the stratified water column were as high as (1-2)x10-3 m2/s; a similar result was reported for Fieberling seamount (e.g., Kunze and Tool, 1997). Our observations showed that the turbulent "columns" are possibly detached from the bottom. Their origin appears to be related to the internal-wave shear, rather than to the bottom-roughness induced turbulence. More regular, frictional boundary layers occupy about 30 % of the water column above the summit of the Ampere seamount. The dissipation rate within these stratified, non-fully mixed layers was high, (1-9)x10-7 W/kg, and corresponding scalar diffusivities ranged (5-70)x10-4 m2/s. Turbulence upstream of the seamount is almost unaffected by the topography, showing a low background level of 6x10-10 W/kg. Downstream of the summit, decreases from 6x10-8 to about (1-2)x10-9 W/kg, over a distance of 36 miles. This decay can be reasonably approximated by an inverse function of the horizontal distance normalized by the buoyancy mixing scale LB at the site of turbulence generation. The data taken from other regions, e.g. turbulent wakes downstream of the Ampere seamount and small Howland Island, also follow the power low trend (x/LB)-1, showing the versatility of the results obtained.

JSP49/W/22-B5 0930

ABYSSAL TURBULENCE

ROBERT B. SCOTT: Mathematics, UCL, Gower Street, London, WC1E 6BT, UK

Deep ocean turbulent mixing is believed to be important in controlling the large scale meridional overturning rate. Munk suggested in 1966 that deep ocean mixing may be much stronger near boundaries than in the interior, and this has been vindicated by direct measurements of microscale turbulence. While direct measurements reflect local mixing rates, one must use budget studies to indirectly infer the time average volume integrated mixing rates. Wunsch has suggested using the source of mechanical energy as a means of inferring the abyssal mixing rates. Recently it has been demonstrated that the flux of mechanical energy from the winds to the surface geostrophic flow F(Kg) can be well quantified using TOPEX/Poseidon (TP) data and recent gravity models. Turbulence theory suggests that the geostrophic energy must be dissipated by interactions with the bottom. Thus F(Kg), along with tidal forcing, are important in that they provide a power source for driving small scale, abyssal turbulence and mixing. In this work, the time mean field of F(Kg), and the factors affecting the reliability of this estimate, are reviewed for the World Ocean. The global mean is comparable to the mean rate of sun and moon tidal forcing. The new results are the presentation of the time variability of F(Kg) integrated over the World Ocean and various subregions. The time varying component in some cases exceeds the time mean and can result in short periods of atmospheric braking of oceanic surface flow. Factors controlling the variability in F(Kg) are identified for the TP time period. We also speculate on the longer time scale variability of power sources driving abyssal turbulence and the effects on global circulation and northern climate.

JSP49/W/10-B5 0950

DOUBLE DIFFUSIVE MIXING PROCESSES IN THE OYASHIO FRONTAL REGION

Jiro Yoshida, Hideki Nagashima and Moh Nagasaka(all at Department of Ocean Sciences, Tokyo University Fisheries, 4-5-7 Konan, Minato, Tokyo Japan 108-8477, email:jiroy@tokyo-u-fish.ac.jp;nagasima@bimelan.tokyo-u-fish.ac.jp)

Along the north eastern coast of Japan, the Kuroshio(warm and salty water) and the Oyashio (relatively cold and fresh water) directly contact and forms the so-called perturbed area. Warm core rings are sometime detached from the Kuroshio in this area and flows northward. Between this perturbed area and the Oyashio, clear frontal region is formed and is sometimes called as Oyashio Front. At this frontal region, vertical profiles of temperature and salinity often show complex structures such as temperature inversion layers, step structures and so on. These structures sometimes traced over several kilometers in horizontal direction, and were considered to be interleaving processes driven by double diffusive convection. In the present study, focusing upon these interleaving processes, the time history of layer formation is analyzed by intensive XBT and CTD surveys. The characteristic features of the layer are discussed statistically, and the effective horizontal and vertical diffusivities are estimated by simple advection-diffusion model. These results are compared with the previous ones obtained by Meddy studies, C-SALT , and NATRE, and the role of double diffusive mixing in this area is also significant and might affect the decay processes of Warm core rings

JSP49/W/08-B5 1010

INTERNAL WAVES, TURBULENCE AND INTRUSION OF SHELF WATER INTO THE KUROSHIO NEAR THE SHELF BREAK IN THE EAST CHINA SEA

Takeshi MATSUNO, Shotaro Ohsaki and In-Seong Han(Faculty of Fisheries, Nagasaki University, Nagasaki, 852-8521 Japan, email:matsuno@net.nagasaki-u.ac.jp)

Measurements of short time scale fluctuations with moored current meters were carried out near the shelf break in the East China Sea. Short time scale internal waves with periods of 10 to 30 minutes, which is slightly larger than the Brunt-Vaisala period, were frequently detected. Some of internal waves have borelike shape in time series of temperature and current velocity. Propagation speed of the borelike internal waves was 0.5 or 0.6 m/s which was estimated from two moorings with distance of 200 m. Shorter time scale fluctuations often follow the internal waves. This suggests that breaking of internal waves would generate turbulent fluctuations.

These turbulent fluctuations could make vertical mixing, however, we did not find distinct occurrence of vertical mixing just after the significant turbulent fluctuations. However, we sometimes found evidences of offshore intrusion of shelf-bottom high turbid water into the Kuroshio. The intrusion tends to be developed when predominant pycnocline is formed just above the shelf break. It is suggested that enhancement of the intrusion should depend on intensification of internal waves associated with turbulent fluctuations, though we did not find a direct evidence yet. Such cross-shelf fluxes related to internal waves and turbulent fluctuations might play an important role on material and energy transports between the shelf water and the Kuroshio.

JSP49/E/07-B5 Invited 1050

STRATIFIED FLOW OVER TOPOGRAPHY: THE ROLE OF SMALL SCALE ENTRAINMENT AND MIXING IN FLOW ESTABLISHMENT

David FARMER (Institute of Ocean Sciences, Sidney, B.C., V814B2 Canada, email: farmerd@dfo-mpo.gc.ca) and Laurence Armi (Scripps Institution of Oceanography, La Jolla, California 92093-0230 USA, email:larmi@ucsd.edu)

Stratified flow over topography is examined in the context of its establishment from rest. A key element of numerical and steady state analytical solutions for large amplitude topographic flow is the splitting of streamlines which then enclose a trapped wedge of mixed fluid above the rapidly moving deeper layer. Measurements have been acquired which illustrate the development of this wedge and the role played by small scale instabilities and mixing formed initially by the acceleration of subcritical stratified flow over the obstacle crest. The volume of trapped fluid progressively increases with time, permitting the primary flow to descend beneath it over the lee face of the obstacle. Throughout the evolution of this flow, small scale instability and consequent entrainment would seem to be a prime candidate for producung the weakly stratified wedge, thus allowing establishment of the downslope flow to take place. Velocity structure of instabilities within the entrainment zone is observed and the associated entrainment rate determined. The entrainment is sufficient to produce a slow downstream motion within the upper layer and a density step between the layers that decreases with downstream distance. The resulting internal hydraulic response is explained in terms of a theory which accommodates the spatially variable density difference across the sheared interface. The measurements described here were acquired in a coastal inlet subject to gradually changing tidal currents. It is proposed that the observed mechanism for flow establishment also has application to atmospheric flow over mountains.

JSP49/L/01-B5 Invited 1110

STRATIFIED FLOW OVER TOPOGRAPHY: HYDRAULIC FRONTS

Laurence ARMI (Scripps Institution of Oceanography, La Jolla, California 92093-0230 USA, email: larmi@ucsd.edu) and David Farmer (Institute of Ocean Sciences, Sidney, B.C., V814B2 Canada, email: farmerd@dfo-mpo.gc.ca)

When stratified fluid is strongly forced over topography, the resulting flow includes a splitting streamline which then encloses a trapped wedge of mixed fluid above the rapidly moving deeper layer. In response to changes in the forcing, the location of this bifurcation or hydraulic front was observered to move either upstream or downstream. The movement of the hydraulic front is in response to either an increase or decrease of the tidal flow speed or as a result of changes in the density of fluid in the wedge itself. The position of the hydraulic front is explained in terms of a quasisteady theory proposed earlier for two-layer flows and the turbulent structures on the hydraulic front are compared with those observed in recent laboratory experiments on an accelerating baroclinic shear layer. The measurements described here were acquired in a coastal inlet subject to gradually changing tidal currents. It is proposed that the observations also have application to atmospheric flows over mountains.

JSP49/W/12-B5 1130

SPATIAL EVOLUTION OF HYDRAULICALLY CONTROLLED FLOWS TO DOWNSLOPE CURRENTS

GENO PAWLAK and Larry Armi (School of Oceanography, University of Washington USA)

The developing region of a steady downslope current in transition from a hydraulically controlled arrested wedge flow is examined through a set of laboratory experiments. The mixing and entrainment mechanisms at the shear interface are explored with the use of imaging techniques. The initial, unstable accelerating region, characterized by rapid development, low shear layer Richardson numbers and high entrainment rates, is quickly limited by the effects of stratification which suppress the vortex pairing mechanism. However, the continuing source of energy from the accelerating flow destabilizes the mixing layer to a shredding type of vortex interaction. The shredding mechanism results in a weakly entraining, well mixed shear layer which evolves towards a steady downslope current for a constant bottom slope. Rates of entrainment of fluid into the turbulent mixing region for each regime are calculated for the various slopes considered.

JSP49/P01-B5 Poster 1210-01

ENERGY TRANSFER AND MIXING NEAR A DENSITY INTERFACE

A. A. BIDOKHTI and M.Saghafi (both at institute of Geophysics, Tehran University, P.O. Box 14155-6466: Tehran, lran. email: bidokhti@chamran.tu.ac.ir)

The problem of mixing and energy transfer across a density interface some distance from the source of turbulent energy, has important applications in geophysical flows. Quite a few experiments have been done, in order to understand arid quantify the processes involved (e.g Turner 1976, Linden 1981 and McGrath et.al. 1997).

Results are presented from experiments in which, mechanical grid turbulence is introduced in a homogeneous layer in contact with a stratified layer. Flow visualisation and direct measurements of salinity field (using fast response salinity probes) near the interface, are presented. With a step profile in density, we obtain an entrainment velocity proportional to Ri-n (where Ri is the overall Richardson number and n ≈ 1). Engulfment, eddy impengment are important when the interface is close to the turbulence source. As the distance increases ( for Ri>30), internal waves breaking near the interface seems to occur more often (e.g. McGrath et. al. 1997). However the case with stratified lower layer (with a linear profile), shows that interfacial internal waves are more pronounced than before. Also some of the energy is transferred to stratified layer and appear as internal waves which seem to have interactions. Internal waves exist in the entire depth of stratified layer. Salinity signals show that the amplitude of the waves decay with depth exponentially. Modulations of the signals indicate the presence of different modes, but the dominant frequency is close to N, the buoyancy frequency.

JSP49/W/05-B5 Poster 1210-02

GENERATION AND REFLECTION OF PERIODIC INTERNAL WAVES: EXACT SOLUTIONS AND LABORATORY EXPERIMENTS

Yuli D. Chashechkin, Yuri S. Ilyinykh and Yuri V. Kistovich (all at the Laboratory of Fluid Mechanics of the Institute for Problems in Mechanics of the RAS, Moscow, prospect Vernadskogo, 101-1, 117526, Russia, E-mail: chakin@ipmnet.ru)

We investigate analytically and experimentally 2D and 3D periodic internal waves generation problem in an exponentially stratified viscous fluid. The wave source is plane strip or vertical cylinder tube periodically oscillating along its surface. The linearized governing equations are solved by integral transform method. The exact boundary conditions on the surface of a body as well as the governing equations are satisfied if, in addition to propagating internal wave beams, internal boundary layers on the emitting surfaces are taking into account. On the basis of these two eigen-forms of fluid motion we calculate wave amplitudes and evolution of the wave shape along the beam. The number of maxima in the wave amplitude cross-section depends on the ratio of source width by intrinsic viscous wave scale. The distance from the source where the bi-modal beam is transformed into the uni-modal one is defined. The spatial decay of the wave is different due to the geometry of the problem. Schlieren visualization and probe measurements of a periodic wave pattern confirm the theoretical results for the far wave field. The absolute values of calculated and measured wave amplitudes differ by a factor less then 1.3. Indirect evidences of the internal boundary layers are presented.

JSP49/W/13-B5 Poster 1210-03

STOKES FLOW INDUCED BY PERIODIC INTERNAL WAVES

Yuli D.Chashechkin and Yuri V. Kistovich (both at the Laboratory of Fluid Mechanics of the Institute for Problems in Mechanics of the RAS, Moscow, prospect Vernadskogo, 101-1, 117526, Russia,

E-mail: chakin@ipmnet.ru )

We study problem of propagation non-linear 2D periodic internal waves in exponentially stratified viscous fluid. The solution of complete set of governing equations is expanded in series on harmonics of the main frequency. The average flow is described by the term with zero frequency. As a seed disturbance the infinitesimal solution of linearized equations is taken. In the second order approximation the explicit expression for stream function of the main Stokes flow is received as quadratic differential form of the stream function of seed infinitesimal waves. As an example we calculate Stokes flow induced by a beam of periodic internal waves generated by a point dipole source. Explicit formula are received in the intrinsic frame where one of the axes is directed along a beam and the other is directed transversely. Total discharge in a beam is absent, that is fluid is moving in one direction in the upper part of a beam and in the opposite direction in the lower part. Patterns of stream lines are calculated for different models of mass source (dipole, quadruple). Stokes flow induced by a single beam tends to zero when kinematic viscosity goes to zero. In a domain of intersection of wave beams from several sources Stokes flow exists even in an ideal fluid. Calculated pattern of stream function describing reflection of an periodic wave beam off a sloping solid plane includes both reflected internal waves and a boundary current on the reflected surface

JSP49/W/06-B5 Poster 1210-04

ESTIMATES OF DEEP-OCEAN TIDAL DISSIPATION FROM THE TOPEX / POSEIDON GLOBAL TIDE MODELS

JOHN A CHURCH1,2 Ole B. Anderson3, Richard Coleman1,2,4 and Neil J. White1,2 1Antarctic CRC, Hobart, Tasmania, Australia 2CSIRO Marine Research, Hobart, Tasmania, Australia 3National Survey, Geodetic Division, Copenhagen, Denmark 4University of Tasmania, Hobart, Tasmania, Australia

The TOPEX/POSEIDON models of global sea-surface tidal elevation are used to estimate tidal currents, assuming linear bottom friction. For the M2 tide, the derived currents satisfy the continuity equation to a rms residual of about 0.2 mm s-1. However, there is a clear correlation between these residuals and the underlying topographic features. From the surface height fields, the derived current fields and the residuals from the continuity equation, the tidal energy budget can be computed. The energy input by the astronomical potential almost exactly balances the divergence associated with the tidal currents. Residuals from the energy equation, ie the deep water dissipation, are positive almost everywhere and are maximum over major topographic features, such as the mid-ocean ridge in the South Atlantic, around New Zealand and near a number of regions where internal tide generation is known to be strong. The bottom dissipation estimates are combined with hydrographic data to estimate vertical mixing rates in the deep ocean.

JSP49/E/11-B5 Poster 1210-05

TIDAL ENERGY AVAILABLE FOR DEEP OCEAN MIXING: BOUNDS FROM ALTIMETRY DATA

GARY D. EGBERT (College of Oceanic and Atmospheric Sciences, Oregon State University, 104 Ocean. Admin. Bldg, Corvallis, OR 97331, email: egbert@oce.orst.edu) Richard D. Ray (Goddard Space Flight Center, NASA, Greenbelt, MD, email: rray@geodesy2.gsfc.nasa.gov)

Maintenance of the large-scale thermohaline circulation has long presented a problem to oceaonographers. Observed mixing rates in the pelagic ocean are an order of magnitude too small to balance the rate at which dense bottom water is created at high latitudes. Recent observational and theoretical work suggests that much of this mixing may occur in hot spots near areas of rough topography (e.g., mid-ocean ridges and island arcs). Barotropic tidal currents provide a very plausible source of energy to maintain these mixing processes. Topex/Poseidon satellite altimetry data have made precise mapping of open ocean tidal elevations possible for the first time. We can thus obtain empirical, spatially localized, estimates of barotropic tidal dissipation. These provide an upper bound on the amount of tidal energy that is dissipated in the deep ocean, and hence is available for deep mixing. We will present and compare maps of open ocean tidal energy flux divergence, and estimates of tidal energy flux into shallow seas, derived from T/P altimetry data using both formal data assimilation methods and empirical approaches. With the data assimilation methods we can place formal error bars on the fluxes. Our results show that 20-25% of tidal energy dissipation occurs outside of the shallow seas, the traditional sink for tidal energy. This suggests that up to 1 TW of energy may be available from the tides (lunar and solar) for mixing the deep ocean. The dissipation indeed appears to be concentrated over areas of rough topography.

JSP49/W/16-B5 Poster 1210-06

ENERGY TRANSFER IN THE SMALL-SCALE OCEANIC INTERNAL WAVE SPECTRUM

Ryo FURUE (Center for Climate System Research, University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8904, Japan, email: furufuru@ccsr.u-tokyo.ac.jp)

We examine energy transfers within a high-wavenumber portion of the oceanic internal wave spectrum by direct numerical simulations. The domain is a rectangular box of 100 m x 100 m x 128 m and the flow field is initialized with the Garrett-Munk (GM) spectrum. The GM flow field of larger horizontal scales is represented as a horizontal shear flow of an infinite horizontal scale. We compare cases with and without this shear flow to assess effects of the large scale GM flow field on the energy transfer. We find that wavenumber-local interactions dominate and transfer energy mainly to high horizontal wavenumber regions and that the modeled rate of energy transfer (1-4 x 10^-10 W/kg) is comparable to observed values in the main thermocline. The contribution of the large-scale flow is found to be not very large although the presence of the large-scale flow does enhance vertical transfer by Doppler spreading mechanism.

From these results, we propose the following scenario of the energy transfer. There are two paths of energy flux in the spectrum. In one, energy is transferred to vertically small scales at horizontally medium scales, and ultimately to dissipation scales, by scale-separated interactions such as described by the eikonal approximation. In the other, energy is transferred to horizontally small scales, where strong wavenumber-local interactions increasingly dominate and transfer energy to dissipation scales.

JSP49/W/17-B5 Poster 1210-07

THE IMPACT OF NON-LINEAR WAVES ON THE DISSIPATION OF INTERNAL TIDAL ENERGY AT A SHELF BREAK

Mark E. Inall (University of the Highlands and Islands project, Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, P.O. Box 3, Oban, PA34 4AD, UK, Email:mei@dml.ac.uk), Tom P. Rippeth (School of Ocean Sciences, University of Wales Bangor, Anglesey LL59 5EY, UK), Toby J. Sherwin (UCES, University of Wales Bangor, Anglesey LL59 5EY, UK)

The vertical and temporal structure of the dissipation of turbulent kinetic energy within the internal tide at a location 5 km shoreward of the shelf break on the Malin Shelf has been determined using a combination of the free-fall FLY profiler and Acoustic Doppler Current Profilers (ADCPs). Two distinct internal wave regimes were encountered: Period I in which large amplitude high frequency non-linear internal waves (NIWs) occurred (around neap tides), and Period II in which the internal wave spectral continuum was not dominated by any particular frequency band (around spring tides). Empirical Orthogonal Function (EOF) analysis shows that for the low frequency waves 76% of the variance was described by mode 1, rising to 95% for the high frequency waves. During period I, the dissipation and vertical mixing were dominated by the NIWs, with 70% of the dissipation occurring in the bottom boundary layer. During period II, the depth integrated dissipation was more evenly distributed throughout the tidal cycle, whereas vertical mixing was concentrated in one hour long episode of elevated thermocline dissipation coincident with weakened stratification. During both I and II approximately 30% of the total measured dissipation occurred within the thermocline when averaged over 12.4 hours, the remainder occurred within the BBL. Tidal average values for depth integrated dissipation and vertical eddy diffusivity for regime I (II) were 1.1x10-2 Wm-3 (4.0x10-2 Wm-3) and 5 cm2s-1 (12 cm2s-1) respectively.

JSP49/W/19-B5 Poster 1210-08

OBSERVATIONS AND K-EPS MODEL SIMULATIONS OF TURBULENT MIXING IN STRATIFIED TIDAL FLOW

Jossy P. JACOB and Hartmut Peters (both at Rosenstiel School of Marine and Atmospheric Science, University of Miami, USA. hpeters@rsmas.miami.edu); Helmut Baumert (Hydromod Scientific Consultants, Wedel, Germany. baumert@hydromod.de)

This paper addresses the question how well second-moment turbulence closure models can simulate tidally driven mixing in fast flowing shallow estuaries. We employ a 1-D numerical k-eps model (Burchard and Baumert, 1995, JGR Oceans, 100, 8523), nudging the model velocity and stratification towards the observations. This is necessary because the estuarine momentum and mass balances are not one-dimensional, advective terms being unknown or poorly known. As a result of the nudging, simulated and real turbulence evolve in very similar shear and stratification. The modeled turbulent kinetic energy and dissipation rate are recast into a turbulent length scale and dissipation, both to be compared quantitatively with observations. Extensive observations of the turbulent dissipation rate and of turbulent overturning scales were made in the Hudson River in 1994/95 in conditions of shallow water (15 m depth typical), energetic tides (velocities of 1-2 m/s), mostly strong stratification (2-18 psu top-to-bottom salinity difference) and strong shear. Shear, stratification and turbulent mixing showed pronounced tidal and fortnightly variations.

The k-eps model is able to simulate turbulent dissipation rates and length scales fairly well. Comparing both parameters to their observed counterparts is crucial to determining adjustable model constants.

JSP49/W/14-B5 Poster 1210-09

TURBULENCE PARAMETERIZATION FOR PBL DISPERSION MODELS IN ALL STABILITY CONDITIONS

Gervásio Annes Degrazia (Universidade Federal de Santa Maria -UFSM, Santa Maria-RS-Brazil, CEP 97.119-900, Email: degrazia@ccne.ufsm.br) CLÁUDIA JACONDINO DE CAMPOS (Universidade Federal de Pelotas-UFPel, Pelotas-RS-Brazil, CEP 96.010-900, Email: cjcampos@ufpel.tche.br)

Accounting for the current knowledge of the Planetary Boundary Layer (PBL) structure and characteristics, a new set of turbulence parameterizations to be used in atmospheric dispersion models has been derived. The expressions for the vertical profiles of the velocity standard deviation, Lagrangian and Eulerian length scale and time scale and diffusion coefficient are proposed. The classical statistical diffusion theory, the observed spectral properties and measured characteristics of energy containing eddies are used to estimate these parameters. The results of this new method are shown to agree with previously determined parameterizations. In addition, these parameterizations give continuous values for the PBL at all elevations and all stability conditions from unstable, neutral to stable. It is the aim of this work to present the general derivations of these expressions and to show how they compare to previous results. Finally a validation of the present parameterization, obtained by inserting it in a Lagrangian particle model, will be shown. The Copenhagen data set is simulated. This data set is particularly suited for this validation, since most of the Copenhagen tracer experiments were performed in stability conditions that are the result of the relative combination of wind shear and buoyancy forces. As a consequence, a parameterization scheme, able to deal contemporary with neutral and slightly convective condition, is preferred.

JSP49/E/06-B5 Poster 1210-10

COEFFICIENT OF TURBULENT DIFFUSION AT A TROPOSPHERIC ALTITUDES

George JANDIERI (department of Physics, Georgian Technical University, 77 Kostava str., Tbilisi 380075, Georgia, email: jandieri@access.sanet.ge) Anzor Gvelesiani (Institute of Geophysics of the Georgian Academy of Sciences, 1 Rukhadze str., Tbilisi 380093, Georgia) Zhuzhuna Diasamidze, Vladimir Glonti and Mzia Diasamidze (all three at Department of Physics, Batumi State University, 35 Ninoshvili str., Batumi, Georgia)

New statistical model of the distribution of pollutant mean concentration is presented at lowest atmospheric turbulent layers. Statistical description of turbulent diffusion is payed attention by means of stochastic diffusion equation. Source of pollutant is deterministic function of coordinates and time. Velocity field describes homogeneous and stationary process. General expression of passive impurity concentration distribution is obtained. Coefficient of turbulent diffusion is expressed in terms of coefficient of molecular diffusion and arbitrary correlation tensor of the vector field of velocity fluctuation. Knowledge of the components of the turbulent diffusion coefficient makes it possible to restore the values of spatial-temporal parameters typical of a stochastically inhomogeneous medium. In particular, distribution of pollutant at anomalous meteorological conditions such as calm is of interest. Numerical simulations of normalized concentration distribution against distance are carried out at stationary and nonstationary cases. Globul formation mechanism is observed in nonstationary case for different values of the turbulent flow velocity. They have different linear dimentions and existence time.

JSP49/E/10-B5 Poster 1210-11

THE EFFECT OF DIFFUSION INSTABILITY IN A LOWER TURBULENT ATMOSPHERIC LAYERS

George JANDIERI (department of Physics, Georgian Technical University, 77 Kostava str., Tbilisi 380075, Georgia, email: jandieri@access.sanet.ge) Anzor Gvelesiani (Institute of Geophysics of the Georgian Academy of Sciences, 1 Rukhadze str., Tbilisi 380093, Georgia) Vladimir Glonti, Guram Varshalomidze and Mzia Diasamidze (all three at Department of Physics, Batumi State University, 35 Ninoshvili str., Batumi, Georgia)

Turbulence favours the mixing of different impurities in random media. Interaction of impurities leads to temporal evolution of medium inhomogeneous structure and it passes to the metastable state. Diffusion instability is investigated by means of solution of stochastic integro-differential equation for the mean concentration of impurities in a randomly inhomogeneous media. Potential energy (PE) of interaction between impurities plays significant role. If PE is positive we can introduce effective diffusion coefficient while it is not justified at any chemical reactions. If PE is negative, critical concentration is exist from which diffusion instability is beginning. Model of PE is supposed using similarity method. Function characterizing PE has potential well at small distances and has wavy character at great distances. New approximation of spatial-temporal distribution of impurity concentration is proposed by solving stochastic diffusion equation using Picard’s method. Concentration fluctuations are caused by velocity pulsations. Incompressibility and isotropy of turbulent medium are not use. Knowledge of statistical parameters of fluctuating media and distribution of the mean concentration of impurities allow us to restore characteristic linear dimensions of diffusible particles.

JSP49W/07-B5 Poster 1210-12

DIFFUSING OF A PASSIVE TRACERS IN RANDOM FLOWS. SOME RESULTS OF NUMERICAL SIMULATION.

KONSTANTIN KOSHEL and Olga Alexandrova (both at Pacific Oceanological Institute, FEB RAS, 43,Bultiyskay str., Vladivostok, 690068, Russia, Email: wave@online.vladivostok.ru)

The diffusion of a passive tracers in a two-dimensional random velocity field is considered. On the basis of diffusion equations solution as a functional integral the method of numerical simulation of a problem is offered. By numerical simulation dynamics of formation of cluster structures is studied. The role of molecular diffusion in such situation and degree of effects connected with potential and solenoidal components of velocity fields is estimated. The behavior of a tracers as in diffusion approximation, and for infinite temporary radius of a correlation is investigated.

All calculations are done for one realization of a velocity field. For other realizations the distribution of an impurity essentially varies, however common behaviour difined by parameters of a problem is saved. The results of simulation will well be agreeed with analytical results obtained by V.I. Klyatskin and A.I. Saichev (JETP. 1997. V. 111. N 4. P. 1297. In Russian). The molecular diffusion renders noticeable influence to distribution of partiñles of an impurity, thereby increasing common square of area occupied by impurity, however influence of this effect to compression of partiñles along trajectorny is unsignificant. On the other hand an incompressible component in a velocity field poorly influences effect of grouping of partiñles in clusters, but essentially decelerates compression of partiñles along trajectorny. Results for case of a velocity field with an infinite radius of correlation on time allow to assume, that the qualitative behaviour of an impurity in this case is saved, however numerical and it is possible also functional performances of process vary. ‰

JSP49/W/11-B5 Poster 1210-13

DECOMPOSITION OF WIND VELOCITY FLUCTUATIONS INTO WAVE AND TURBULENT SIGNALS USING WAVELET TRANSFORM

Nisia Krusche (Dept. Geociencias, Fundação Universidade De Rio Grande, C.P. 474, 96201-900 Rio Grande Rs, Brazil) And Osvaldo L. L. Moraes (Dept. Física, Universidade Federal De Santa Maria, 97119-900 Santa Maria, Rs, Brazil)

In the stably stratified atmospheric boundary layer, spectra might be affected by several kinds of Waves. Since the problem of turbulent diffusion is closely related to the spectra of wind velocity Fluctuations, there is a need to separate the contributions of those waves from that of Turbulence. Discrete wavelet transform was applied to temperature and vertical velocity Fluctuations measured in candiota, rs (31o28´s, 53o 40´w), sampled with a frequency of 10 Hz. In this data set, waves have been detected through cospectral techniques. As a result, a Wave was identified in both series, which was 90o out of phase, as expected. Therefore, this Method may be used to decompose the series in two signals, one containing the wave and Other corresponding to random turbulence of the flow.

JSP49/W/23-B5 Poster 1210-14

OVERTURNING LENGTH SCALE AS AN INDICATOR OF THE WASTEWATER NEAR FIELD

Vlado MALACIC (Marine Biological Station, National Institute of Biology, Piran, 6330, Slovenia, email: malacic@nib.si)

During a period of calm weather and stratified water column of a shallow (depth = 21 m) semi-enclosed sea, six surveys of the wastewater near-field of two adjacent submarine diffuser outfalls, with a low sewage ouflow rate (usually less than 220 l/s), were conducted with a fine-scale CTD probe. The survey area of 780m_740m was monitored with the free-falling probe at 30-31 stations. Vertical velocity of the probe was nearly constant, 1 m/s, and the data was retrieved at a frequency of 50 Hz, providing a vertical resolution of 2.5 cm.

When using a fine-scale CTD it has been observed that from the vertical distribution of 'virtual' displacements (Thorpe displacements) of particles with a sampled density, by which each particle would gain a statically stable position, the length scale of overturning events could be estimated, through an appropriate averaging process.

Within the turbulent wastewater field the fluctuations of temperature were of the order of 0.1 oC, and of salinity less than 0.05 psu, generating overturning. The vertical extent of the overturning events could well be determined using the maximum vertical displacement, or the Thorpe length scale. Both scales also indicate the thickness of the neutrally buoyant subsurface wastewater field. Distribution of these length-scales over the diffussors show a clear picture of the wastewater field(s) of a thickness smaller than 1m in windless, stratified conditions, which is otherwise hard to obtain without costly experiments (dye tracers).

JSP49/E/01-B5 Poster 1210-15

PROPAGATION AND DISSIPATION OF TIDAL INTERNAL WAVES IN THE NORTH PACIFIC ON THE BASIS OF THE MEGAPOLYGON DATA, AN EXPERIMENT WITH 170 MOORED BUOYS

Eugene MOROZOV (Shirshov Institute of Oceanology, 36, Nakhimovsky st., 117851, Moscow, Russia, e-mail: internal@redline.ru)

Tidal semidiurnal internal waves generated over the Emperor Ridge in the North Pacific are analyzed on the basis of a vast cluster of moorings in the area between 38 - 42 N and 152 - 158 E. The cluster of 170 moored buoys was set for two months in this region with a distance of 24 miles between moorings. The current and temperature meters were set at 120 m and 1200 m levels. This cluster of moorings with current meters forms an antenna to study eddies and tidal internal waves and their variability. The maximum distance between the utmost western and eastern moorings exceeded 450 km. The wave lengths of tidal internal waves vary between 130 and 150 km, while the change of the direction of their propagation is insignificant. Westerly waves are dominating indicating that the Emperor Ridge is the source area. The cluster of buoys was divided into two parts to obtain independent data sets for evaluating wave parameters. The wave lengths and direction of propagation measured at both parts of the study area gave approximately the same results. The results did not change in the analysis of the first and the second months of the time series. The maximum amplitudes of the waves in the eastern part of the study area exceed 40 m, whereas in the western part they were about 15-20 m or less. Decay of the waves is estimated as 5% of the energy over one wave length. Similar results were obtained from the data of US moorings in this area and also from the moorings east of the Emperor Ridge.

JSP49/W/24-B5 Poster 1210-16

A NUMERICAL EXPERIMENT ON THERMOBARIC CONVECTION

Miyako Naya and Hideki Nagashima (Department of Ocean Sciences, Tokyo University of Fisheries, Konan,Minato, Tokyo Japan 108-8477. E-mail:nagasima@bimelan.tokyo-u-fish.ac.jp)

Predictions of vertical convection and deep water formation in the polar region need to include the nonlinear equation of state. In the present study, a numerical experiment on thermobaric convection is carried out based on the non-linear equation of state in which the thermal expansion coefficient is a function of pressure_@together with temperature and salinity. A two-dimensional non-hydrostatic model is used with the calculating domain of 8km in horizontal and 4km in vertical directions. The onset of convection is well explained by a linear stability analysis of Rayleigh-Benard problem and the convection develops and reveals thermobaric enhanced turbulence. Next, the double diffusive effect is taken into consideration, because the typical vertical profile in the polar sea shows high activity of double diffusion. In the numerical model, the vertical diffusivities of heat and salt are parameterized by using empirical flux law (Kelly:1990). The result shows that the double diffusive effect reduces the onset of thermobaric convection.

JSP49/W/20-B5 Poster 1210-17

STIMULATION OF LARGE-SCALE ANTICYCLONIC CIRCULATION IN THE OCEAN VIA SALT FINGERS CONVECTION

Anatoly Pereskokov (Russian Research Institute of Hydrometeorological Information - World Data Centre, 6 Korolyov St , Obninsk 249020, Russia, email: peres@meteo.ru)

A further progress in understanding of the nature of the ocean large-scale circulation relates first of all to the estimate of the processes effectiveness generating the mixing in the ocean picnocline. Diapicnic transfer due to salt fingers convection appears to be the significant climate-forming factor (Pereskokov, Fedorov, 1985). Evidence of this is the displacement of regions of the increased temperature from equatorial latitudes to the subtropics with the increase in depth. It is shown that in the core of waters with thermohaline conditions favourable for the development of salt fingers convection which is found in all oceans in the layers between horizons 300 and 600 m, their volumetric shares in the North Pacific, South Pacific, and the Indian ocean are close to 70% and increases to 80% in the South Pacific and almost to 93% - in the North Atlantic. In such a case, an overwhelming majority of the values of density ratio lie within of 1.7-5.0 for the entire World Ocean and within of 1.7-2.5 - for the Atlantic. These limits for the density ratio are typical for conditions in which the effect of salt fingers is most pronounced. Via this vertical heat transfer from upper layers to deep waters salt fingers convection stimulates large-scale anticyclonic circulation contributing to the development of cores of warm (with less density) waters in subtropics. Probably it is not easy to find the numerical support of this hypothesis, but the fact itself that it is already on the level of 200 m that the wind energy can account for only 30% of the observed values of the current speed, is indirect evidence in favour of the assumption made.

JSP49/W/15-B5 Poster 1210-18

RICHARDSON NUMBER AS AN INDICATOR OF TURBULENT STRATIFICATION AND LARGE EDDY DIFFUSION

Prof .KPR Vittal Murty.(DCM,INPE,Av.dos Astronautos1758,caixa postal 515,Sao jose dos campos-sp-BRAZIL, cep-12201-010,email murty@met.inpe.br)

Richardson number(Ri),a non-dimentional number which is a ratio of thermal and kinetic forces is an indicator of surface layer turbulence.It is closely associated with turbulence production terms in the equation of turbulent energy balance.In this article the turbulent state of the surface layer is studied as a function of Ri based on Pantanal gradient observation data. Pantanal is located in the westren part of central Braziland is well noted for its flora and fauna. An International Pantanal Experiment(IPE_1) was conducted in May 1998,which is a part of broad experimental program to study the weather and climate of this region.The gradient observations of wind and temperature obtained in IPE_1were used in the present study.The three different forms of Ri namely bulk, gradient and flux form are estimated and compared. The various functions of Ri as envisaged by Pristley, Pasquill, Gurevitch and Riderwere estimated and discussed. The variation of horizontal velocity variance with Ri was studied for a clear and cloudy day.This shows an Increase of variance with decrease of Ri and the slope which also varies indicate the large eddy diffusion in the area

Friday 30 July PM

Presiding Chair: Peter G. Baines (CSIRO Atmospheric Research, Aspendale, Australia)

JSP49/E/09-B5 Invited 1400

TURBULENT CONVECTION AND MIXING IN A LABORATORY MODEL OF THE CONVECTIVELY-DRIVEN FLOW IN PARTIALLY-ENCLOSED SEAS.

T. MAXWORTHY and Th. Grimm, Department of Aerospace and Mechanical Engineering, University of Southern California. Los Angeles, CA 90089-1191, USA.

Turbulent convection and mixing play a major role in a variety of natural hydrodynamical systems. A special class of such systems are those in which convection drives exchange flows in a channel that is closed at one end and has a lateral contraction and/or a sill at the other, open, end. Typical examples of such "partially-enclosed seas" are the Red and Mediterranean Seas, the Persian Gulf and the fjords that indent many coastlines. The present work focuses on the spatial distribution and scaling of the density difference (g') between the inflowing and outflowing fluid layers at the exit and the characteristics of the turbulent mixing that takes place at the closed end of the sea.

Using a long, water-filled channel, fitted with buoyancy sources at its upper surface, experiments were conducted to investigate the influence of the geometry of the strait and the channel as well as the magnitude of the buoyancy flux. Firstly, we have found that the scaling law:

g' = const. Bo^2/3 x/h^4/3

best describes the distribution of the observed density difference along the channel. Where Bo is the buoyancy flux, x the distance from the closed end of the channel and h its depth at the open end (sill). The numerical value of the constant depends on the hydraulic conditions at the exit and these, in turn, depend on the details of the channel geometry. This scaling law holds for the experimental results and appears to be valid for a number of natural systems as well. Secondly, we discuss, in detail, the flow conditions that are responsible for the turbulence and mixing, at the closed end of the channel, that ultimately account for the observed density distribution.

JSP49/E/08-B5 1420

PARTICLE DISPERSION AND MIXING INDUCED BY BREAKING INTERNAL GRAVITY WAVES

P. BOURUET-AUBERTOT, C. Koudella & C. Staquet DAMTP, U. of Cambridge, Silver Street, Cambridge, CB3 9EW, UK Lab. de Physique, E.N.S. de Lyon, 46 allee d'Italie, F69364 Lyon Cedex 07 L.E.G.I., BP 53, F38041 Grenoble Cedex 9

Our study focuses upon diapycnal mixing induced by a breaking gravity wave. The primary wave we consider is of small amplitude (statically stable), a case for which the breaking process mostly involves two-dimensional instabilities. The dynamics of the waves have been previously analyzed by means of two-dimensional direct numerical simulations and three-dimensional calculations will also be reported. Diapycnal mixing is inferred both from particle dispersion and from potential energy budgets, and the two methods are found to lead to the same prediction for the diapycnal diffusivity. This is of particular interest regarding the interpretation of in situ measurements. The diapycnal diffusivity is found to depend on the turbulent Froude number squared, a dependency which we shall relate to the instability process of the primary wave.

JSP49/E/05-B5 1440

MIXING IN A STABLY STRATIFIED SHEAR LAYER

Chantal STAQUET (LEGI, BP 53, 38041 Grenoble cdx 9, France, Chantal.Staquet@hmg.inpg.fr)

We have investigated the mixing properties of a stably stratified shear layer in two and three dimensions, using a new analysis of mixing proposed by Winters et al. (1995) and Winters and d'Asaro (1996). This analysis provides an exact expression for the diapycnal flux of density; it relies on a stable density profile obtained by sorting adiabatically the fluid parcels so that the heaviest has the lowest altitude. The interest in the new analysis is that mixing rates, such as the mixing efficiency or the diapycnal diffusivity, can be computed instantaneously whatever the flow regime, even if unsteady and strongly nonlinear. We found that the maximum value, as a function of time, of the flux Richardson number depends upon the vertical scale of the instability triggering turbulence, being larger for a larger vertical scale. When the vertical scales of the flow collapse and a weakly nonlinear regime is eventually reached, the predictions of the analysis coincide with those by the Osborn-Cox model. Interestingly, the mixing efficiency relaxes toward an asymptotic value close to 0.25, when the flow is three-dimensional or two-dimensional with strong stratification. In the latter case, the normalized turbulent diffusivity is found to depend on the inverse of the Richardson number, a parameterization that is identical to that often used in large scale circulation models. It is noteworthy that the same law is found in breaking internal gravity waves.

JSP49/L/02-B5 1500

MIXING IN DOWNSLOPE FLOWS INTO STRATIFIED ENVIRONMENTS

Peter G. BAINES, CSIRO Atmospheric Research, PMB No. 1, Aspendale, Australia 3195

The properties of downslope flows may be significantly affected by density variation in the environment. Oceanic examples are downslope flows at high latitudes in the North Atlantic and around Antarctica, driven by surface cooling and brine rejection due to freezing of sea water, and nocturnal flows in the atmosphere driven by radiative surface cooling on slopes.

This paper presents a description of laboratory experiments on downslope flows with stratified environmental fluid, with a dynamical interpretation of the observations, and with measurements of the relevant quantities to enable application of these results to larger scale models. The experiments were carried out in a glass-sided tank 80 cm high and 299 cm long, filled with uniformly salt-stratified water, in which two-dimensional downslope flows of width 23 cm were generated. For each run, a continuous two-dimensional source of dense salty water was introduced at the top of a sloping bottom, for a finite time that depended on the strength of the source and was typically of several minutes duration. Density profiles were measured before the downflow, and some time after it, when all residual motion had ceased. Taking the difference between these two profiles gave quantitative information about the variation of the downflow with downslope distance, the final disposition of the introduced fluid, and the redistribution of the ambient stratification.

A wide range of differing behaviour was observed, depending on the slope angle, the parameter M (which depends on the mass flux, buoyancy and stratification) and the Reynolds number. The thickness of the density current is generally uniform with downslope distance, but the current exchanges fluid with the environment by the processes of turbulent entrainment into it and more significantly, detrainment from it. The latter is generally the stronger process for small slope angles, where mixing mostly occurs at the upper boundary of the current, but for steeper slope angles (30 degrees and larger) the mixing penetrates deeper, and entrainment is dominant. These processes may be quantified by entrainment and detrainment coefficients for modelling purposes.

JSP49/W/02-B5 1520

DEPENDENCE OF MIXING EFFICIENCY ON INITIAL STRATIFICATION IN SHEAR LAYERS

C.P. CAULFIELD (Centre for Environmental and Geophysical Flows and School of Mathematics, University of Bristol, University Walk, Bristol, BS8 1TW, UK, Email: c.p.caulfield@bris.ac.uk) and W.R. Peltier (Department of Physics, University of Toronto, Toronto, ON M5S 1A7, Canada, Email: peltier@atmosp.physics.utoronto.ca)

We investigate the detailed nature of the "mixing transition'' through which intense turbulence may develop in stratified free shear layers. We explicitly quantify the time-evolving irreversible mixing which occurs within the flow. We consider in detail the variation in the character of the mixing processes with the initial ambient stratification.

Using the numerical data from a sequence of three-dimensional simulations with varying stratification, we accurately track the nonlinear amplification of streamwise vortex streaks, verifying that they are well-predicted by secondary stability analysis, and that they are due to a convective destabilization of the periphery of a Kelvin-Helmholtz billow. At all times we calculate the minimal potential energy of the system accessible by (notional) adiabatic rearrangement of fluid parcels, and so quantify continuously the irreversible "mixing''.

Vortex stretching leads eventually to a violent subcritical vortex-vortex collision which drives the dominant mixing process in the flow life cycle. An appropriate definition of the "mixing efficiency'' implies that the irreversible small-scale mixing of the density which is triggered by shear layer transition leads inevitably to a density "staircase'', with regions of well-mixed fluid separated by narrow regions of relatively strong density gradient.

JSP49/W/18-B5 1600

TURBULENT MIXING AND DISSIPATION ACCOMPANYING KELVIN-HELMHOLTZ INSTABILITY

DAVID C. FRITTS and Joseph A. Werne, Colorado Research Associates, 3380 Mitchell Lane, Boulder, CO 80301 USA

The Kelvin-Helmholtz (KH) instability is a major source of turbulence and mixing in stratified and sheared fluids. However, numerical simulations have only recently achieved sufficient resolution to describe the generation and mixing accompanying an extended inertial range of turbulence. Instabilities arise in the outer portions of the KH billows, penetrate into the billow cores thereafter, and lead ultimately to a layer of turbulence that becomes nearly homogeneous horizontally. Turbulence rapidly mixes the fluid initially within the billow cores, leading to a nearly adiabatic layer with large velocity and thermal gradients above and below. Energy and thermal dissipation profiles exhibit considerable spatial and temporal variability and non-Gaussian statistics, particularly at the edges of the mixing zone. As the turbulence layer restratifies, a dynamically stable shear flow is achieved.

JSP49/W/21-B5 1620

QUANTITATIVE RESULTS ON TURBULENT PATCHES IN THE STRATOSPHERE

J.-R. ALISSE and C. Sidi Service d'Aeronomie du CNRS BP 3 91371 Verrieres-le-Buisson France

The small-scale three-dimensional turbulent activity in the stably stratified atmosphere mainly dwells in isolated patches, with a larger horizontal extent than the vertical one. These structures are usually described as pancakes; they are believed to be completely mixed and hence should display sharp boundary transitions at their edges. Using high-resolution measurements of temperature and velocities profiles, obtained from a balloon-borne instrumentation, we present quantitative results on these patches. From the analysis of Richardson numbers, dissipation rates, Ozmidov, Thorpe and Kolmogorov scales within each layer, a typical pattern emerges, associating low Richardson numbers and significative Thorpe scales. It is noteworthy that the Ozmidov and Thorpe scales are usually quite smaller than the eyeballing thickness of the patches, as observed too in the ocean. Associated with the weak available potential energy dissipation rates, these elements suggest that in almost all the patches the mixing is imperfect. This may be related to the generating processes and life cycle of the patches and has important outcomes for the estimation of a vertical diffusivity on the large-scale tracer fields linked to this small scale turbulent activity.

JSP49/W/01-B5 1640

DISSIPATION IN SHEAR-FREE TURBULENCE NEAR BOUNDARIES

Miguel A.C.TEIXEIRA and Stephen E. Belcher (both at Department of Meteorology, University of Reading, Earley Gate, PO Box 243, Reading RG6 6BB, UK, email: miguel@met.reading.ac.uk)

Processes such as gas transfer across gas-liquid interfaces are controlled by small-scale turbulence near the interface. Modelling these flows is complex, partly because turbulence dissipation is notoriously difficult to represent in closure models, particularly near boundaries. Hence, we investigate the fundamental dynamics of turbulence near interfaces.

Most closure models are developed for solid walls, but recent DNS of shear-free turbulence near boundaries shows significant differences between flows bounded by solid walls and those bounded by free surfaces. Here, we develop a simple model for the rapid distortion of turbulence by such boundaries, which shows how viscous processes lead to the differences between solid walls and free surfaces. The new model produces profiles of Reynolds stresses and turbulence dissipation rates that are in good quantitative agreement with DNS data at short times. As in previous rapid distortion studies, this short time model is able to predict correctly qualitative features of the Reynolds stress and dissipation profiles at longer times.

We conclude that the new model captures the essential physics of the problem and that blocking and the dynamics of the viscous boundary layer are the essential factors determining the shape of the turbulent Reynolds stresses and dissipation profiles near the boundary. The differences between model results and DNS data at long times can be attributed both to the neglected nonlinear effects and to turbulence decay due to dissipation. This highlights the importance of correctly representing dissipation in near-wall closure modelling.

JSP49/W/03-B5 1700

A STUDY OF TURBULENT REGIME AS REVEALED BY VECTORIAL AS WELL AS SCALAR VARIANCE.

PROF.KPR Vittal Murty

This article presents a detailed study of he variances of wind components ,temperature and humidity estimated from the data obtained from three dimensional fast response sonic anemometer during May 1998 which is a part of inter-disciplinary Pantanal experiment(IPE-1).Pantanal is the largest wet land in the world.It is located in central -west Brazil It has ecological and socio- economic significance.There is large seasonal variation in meteorological and hydrological parameters in this region,marked by dry wintersand wet summers.The variances of wind components for pure mechanical turbulence are constantand agree fairly with the the values reported by Panofsky and Dutton.It shold be mentioned that the values of velocity variance lie between the values obtained from earlier expeditions for flat and rolling terrain. Thisindicate that the terriagn features of Pantanal are in between flatand rolling terraigns. The vertical varince is scaled with Monin -Obukhov length scale(L).It is minimum at nuetral condition increasing with stability as well as in-stability. An interesting feature is the relation between horizontal variance and Richardson number.The slope of the curve varies for aclear day and cloudy day thus is a function synoptic scale disturbance.BOth the Scalar variances of temperature and humidity showed a similar variation with Z/L (here Z is the height of observation and and L monin-obukhov length scale) Both the parameters showed a discontinuity at Z/L = 0.

JSP49/W/09-B5 1720

SST ANOMALIES, VERTICAL TURBULENT MIXING AND SUBMARINE EARTHQUAKES

M.A. NOSOV, M.V. (Lomonosov Moscow State University Faculty of Physics Moscow State University Russia) B.W Levin., S.N. Skachko

The appearance of large-scale (~500 km) cool SST anomalies were discovered in the epicenter regions of strong underwater earthquakes. In our opinion the mechanism of the SST anomaly origin was certainly connected with an export of cold lower water layers to the ocean surface as a result of intensive turbulent mixing generated by seismic bottom motions. These SST anomalies may lead to some obvious and at the same time important consequences and accompanying phenomena should be studied. Among all the phenomena we single out the following: 1)export of the nutrients toward upper water layers where there is an usual shortage of the substances causing an increase of the productivity (like in an upwelling zone), 2)generation of large amplitude internal waves (much larger than internal tsunami waves) as a result of evolution of the anomaly stratification zone; 3)the appearance of the cool SST anomaly should cause a response of the atmosphere. This study is aimed to reveal the mechanism of the turbulence generation above the moving bottom. The numerical model of the process is based on the equation of the balance of the turbulence energy in stratified fluid and equation of turbulent heat transfer. The results of this study gives us grounds to state that submarine earthquakes can produce large scale cool SST anomalies above the pleistoseismic zone by means of intensification of the vertical turbulent exchange.