Averaging procedures for flow within vegetation canopies

Most one-dimensional models of flow within vegetation canopies are based on horizontally averaged flow variables. This paper formalizes the horizontal averaging operation. Two averaging schemes are considered: pure horizontal averaging at a single instant, and time averaging followed by horizontal averaging. These schemes produce different forms for the mean and turbulent kinetic energy balances, and especially for the wake production term describing the transfer of energy from large-scale motion to wake turbulence by form drag. The differences are primarily due to the appearance, in the covariances produced by the second scheme, of dispersive components arising from the spatial correlation of time-averaged flow variables. The two schemes are shown to coincide if these dispersive fluxes vanish.     

The Effect of Wind-Turbine Wakes on Summertime US Midwest Atmospheric Wind Profiles as Observed with Ground-Based Doppler Lidar

We examine the influence of a modern multi-megawatt wind turbine on wind and turbulence profiles three rotor diameters ( $D$ D ) downwind of the turbine. Light detection and ranging (lidar) wind-profile observations were collected during summer 2011 in an operating wind farm in central Iowa at 20-m vertical intervals from 40 to 220 m above the surface. After a calibration period during which two lidars were operated next to each other, one lidar was located approximately $2D$ 2 D directly south of a wind turbine; the other lidar was moved approximately $3D$ 3 D north of the same wind turbine. Data from the two lidars during southerly flow conditions enabled the simultaneous capture of inflow and wake conditions. The inflow wind and turbulence profiles exhibit strong variability with atmospheric stability: daytime profiles are well-mixed with little shear and strong turbulence, while nighttime profiles exhibit minimal turbulence and considerable shear across the rotor disk region and above. Consistent with the observations available from other studies and with wind-tunnel and large-eddy simulation studies, measurable reductions in wake wind-speeds occur at heights spanning the wind turbine rotor (43–117 m), and turbulent quantities increase in the wake. In generalizing these results as a function of inflow wind speed, we find the wind-speed deficit in the wake is largest at hub height or just above, and the maximum deficit occurs when wind speeds are below the rated speed for the turbine. Similarly, the maximum enhancement of turbulence kinetic energy and turbulence intensity occurs at hub height, although observations at the top of the rotor disk do not allow assessment of turbulence in that region. The wind shear below turbine hub height (quantified here with the power-law coefficient) is found to be a useful parameter to identify whether a downwind lidar observes turbine wake or free-flow conditions. These field observations provide data for validating turbine-wake models and wind-tunnel observations, and for guiding assessments of the impacts of wakes on surface turbulent fluxes or surface temperatures downwind of turbines.     

Numerical Simulation of Strongly Stratified Flow Over a Three-Dimensional Hill

A numerical study of stably stratified flow over a three-dimensional hill is presented. Large-eddy simulation is used here to examine in detail the laboratory experimental flows described in the landmark work of Hunt and Snyder about stratified flow over a hill. The flow is linearly stratified and U/Nh is varied from 0.2 to 1.0. Here N and U are the buoyancy frequency and freestream velocity respectively, and h is the height of the hill. The Reynolds number based on the hill height is varied from 365 to 2968. The characteristic flow patterns at various values of U/Nh have been obtained and they are in good agreement with earlier theoretical and experimental results. It is shown that the flow field cannot be predicted by Drazin's theory when recirculation exists at the leeside of the hill even at UNh 1. The wake structure agrees well with a two-dimensional wake assumption when U/Nh 1 but lee waves start to influence the wake structure as U/Nh increases. The dividing-streamline heights obtained in the simulation are in accordance with experimental results and Sheppard's formula. The energy loss along the dividing streamline due to friction/turbulence approximately offsets the energy gained from pressure field. When lee waves are present, linear theory always underestimates the amplitude and overestimates the wavelength of three-dimensional lee waves. The simulated variations of drag coefficients with the parameterK (=ND/ U) are qualitatively consistent with experimental data and linear theory. Here D is the depth of the tank.     

Impact of Neutral Boundary-Layer Turbulence on Wind-Turbine Wakes: A Numerical Modelling Study

The wake characteristics of a wind turbine in a turbulent boundary layer under neutral stratification are investigated systematically by means of large-eddy simulations. A methodology to maintain the turbulence of the background flow for simulations with open horizontal boundaries, without the necessity of the permanent import of turbulence data from a precursor simulation, was implemented in the geophysical flow solver EULAG. These requirements are fulfilled by applying the spectral energy distribution of a neutral boundary layer in the wind-turbine simulations. A detailed analysis of the wake response towards different turbulence levels of the background flow results in a more rapid recovery of the wake for a higher level of turbulence. A modified version of the Rankine–Froude actuator disc model and the blade element momentum method are tested as wind-turbine parametrizations resulting in a strong dependence of the near-wake wind field on the parametrization, whereas the far-wake flow is fairly insensitive to it. The wake characteristics are influenced by the two considered airfoils in the blade element momentum method up to a streamwise distance of 14D (D = rotor diameter). In addition, the swirl induced by the rotation has an impact on the velocity field of the wind turbine even in the far wake. Further, a wake response study reveals a considerable effect of different subgrid-scale closure models on the streamwise turbulent intensity.     

Flow visualization experiments on stably stratified flow over ridges and valleys

Summary Flow visualization experiments on stably stratified flow over ridges and valleys formed by a pair of ridges were conducted in a large towing tank filled with stratified salt water. The ambient flow was normal to the axis of the ridges. Three experimental parameters were varied during the study: the steepness of the ridges, the separation distance between the ridges and the Froude number. The flow field in the valley was strongly influenced by flow separation from the lee side of the upstream ridge. For the gentle and intermediate ridges with maximum slopes of 13° and 27°, this separation was controlled by the lee waves when their wavelength was less than or equal to the width of the ridge. The flow field in the valley was similar to that downstream of a single ridge. For the steep ridge with a maximum slope of 40°, separation from the lee side of the upstream ridge and the flow field in the valley were influenced significantly by the presence of the downstream ridge.With 13 FiguresOn assignment from the National Oceanic and Atmospheric Administration, U.S. Department of Commerce.     

Interaction between atmospheric flows and obstacles: Experiments in a rotating channel

We performed an experimental study using scale models in a hydrodynamic rotating channel, concerning the interactions between fluid flows and obstacles of different shapes. The study was meant to analyze the characteristics of the wakes observed on the lee side of quasi-bidimensional obstacles, in a neutral atmosphere.The obstacles were half-cylinders (with aspect ratio 0.87), placed transversally on the channel bottom and totally submerged in the fluid. We call them quasi-bidimensional since their width was a little smaller than the channel width, thus allowing the flow to partially go round their edges.The simulations were performed in the rotating hydraulic channel of ICG-CNR in Turin, and included various conditions of rotation period and flow speed. An interesting behaviour of the wakes was found on the lee side of subsynoptic-scale obstacles, modelled in conditions of Reynolds-Rossby similitude. More precisely, if a given threshold of flow velocity is exceeded, wake size is constant and is fully determined by the height of the obstacle.     

Application of theE-ε turbulence closure model to separated atmospheric surface-layer flows

Neutrally buoyant atmospheric surface-layer flow over a thin vertical wall has been studied using a turbulence closure scheme designed specifically to address flow problems containing high shears. The turbulent flow model consists of a general solution of the time averaged, steady state, twodimensional Navier-Stokes equations, where theE- turbulence model has been used to close the system of equations. Model output compares favorably with measurements made in both a full-scale field study and in an atmospheric wind tunnel. In the simulation of flow over a solid wall, two recirculation eddies are produced. The smallest eddy is located windward of the wall with a separation point located atx/h=–0.8, and the largest is located in the lee of the wall atx/h=5.8. Immediately downwind of the wall top, the turbulent kinetic energy, the energy dissipation rate, and the momentum flux all reach a local maximum. These peak values generally maintain their height positionz/h=1.0, but decrease progressively downwind. The turbulent viscosity is strongly modified under the influence of the wall, with a local maximum forming in the lee of the wall top, and a local minimum forming at a heightz/h=2.0 above the lee recirculation eddy. The surface momentum flux reduction due to the presence of the wall begins atx/h=–10.0. Minimum zero fluxes occur at the surface separation points, and a local peak in momentum flux is produced at the centers of each recirculation eddy. Downwind of the wall, the modeled surface flux reaches an equilibrium at roughlyx/h=30.0.     

Stratified flow over three-dimensional topography

In order to investigate flows over topography in an atmospheric context, we have studied experimentally the wake structure of axi-symmetric Gaussian obstacles towed through a linearly stratified fluid. Three dimensionless parameters govern the flow dynamics: F, the Froude number based on the topography height h; Re, the Reynolds number and the aspect ratio r = h/L, where L is the topography horizontal scale. Two-dimensional (2-D), saturated lee wave (SLW) and three-dimensional (3-D) regimes, as defined in Chomaz et al. (1993), are found to be functions of F and r only (Fig. 1) as soon as Re is larger than Re_c ≈ 2000. For F < 0.7 the flow goes around the obstacle and the motion in the wake is quasi-two-dimensional. This 2-D layer is topped by a region affected by lee wave motions with amplitude increasing with r and F. For 0.7 < F < 1/r, the flow is entirely dominated by a lee wave of saturated amplitude which suppresses the separation of the boundary layer from the obstacle. Above the critical value 1/r, the lee wave amplitude decreases with F and a recirculating zone appears behind the obstacle. Simultaneously, coherent large-scale vortices start to be shed periodically from the wake at a Strouhal number which decreases as 1/F until it reaches its neutral asymptotic value.     

Experimental Study on the Wake Meandering Within a Scale Model Wind Farm Subject to a Wind-Tunnel Flow Simulating an Atmospheric Boundary Layer

The phenomenon of meandering of the wind-turbine wake comprises the motion of the wake as a whole in both horizontal and vertical directions as it is advected downstream. The oscillatory motion of the wake is a crucial factor in wind farms, because it increases the fatigue loads, and, in particular, the yaw loads on downstream turbines. To address this phenomenon, experimental investigations are carried out in a wind-tunnel flow simulating an atmospheric boundary layer with the Coriolis effect neglected. A \(3 \times 3\) scaled wind farm composed of three-bladed rotating wind-turbine models is subject to a neutral boundary layer over a slightly-rough surface, i.e. corresponding to offshore conditions. Particle-image-velocimetry measurements are performed in a horizontal plane at hub height in the wakes of the three wind turbines occupying the wind-farm centreline. These measurements allow determination of the wake centrelines, with spectral analysis indicating the characteristic wavelength of the wake-meandering phenomenon. In addition, measurements with hot-wire anemometry are performed along a vertical line in the wakes of the same wind turbines, with both techniques revealing the presence of wake meandering behind all three turbines. The spectral analysis performed with the spatial and temporal signals obtained from these two measurement techniques indicates a Strouhal number of \(\approx 0.20 - 0.22\) based on the characteristic wake-meandering frequency, the rotor diameter and the flow speed at hub height.     

Effects of Thermal Stability and Incoming Boundary-Layer Flow Characteristics on Wind-Turbine Wakes: A Wind-Tunnel Study

Wind-tunnel experiments were carried out to study turbulence statistics in the wake of a model wind turbine placed in a boundary-layer flow under both neutral and stably stratified conditions. High-resolution velocity and temperature measurements, obtained using a customized triple wire (cross-wire and cold wire) anemometer, were used to characterize the mean velocity, turbulence intensity, turbulent fluxes, and spectra at different locations in the wake. The effect of the wake on the turbulence statistics is found to extend as far as 20 rotor diameters downwind of the turbine. The velocity deficit has a nearly axisymmetric shape, which can be approximated by a Gaussian distribution and a power-law decay with distance. This decay in the near-wake region is found to be faster in the stable case. Turbulence intensity distribution is clearly non-axisymmetric due to the non-uniform distribution of the incoming velocity in the boundary layer. In the neutral case, the maximum turbulence intensity is located above the hub height, around the rotor tip location and at a distance of about 4–5.5 rotor diameters, which are common separations between wind turbines in wind farms. The enhancement of turbulence intensity is associated with strong shear and turbulent kinetic energy production in that region. In the stable case, the stronger shear in the incoming flow leads to a slightly stronger and larger region of enhanced turbulence intensity, which extends between 3 and 6 rotor diameters downwind of the turbine location. Power spectra of the streamwise and vertical velocities show a strong signature of the turbine blade tip vortices at the top tip height up to a distance of about 1–2 rotor diameters. This spectral signature is stronger in the vertical velocity component. At longer downwind distances, tip vortices are not evident and the von Kármán formulation agrees well with the measured velocity spectra.     

Observations of Boundary-Layer Wind-Tunnel Flow over Isolated Ridges of Varying Steepness and Roughness

Boundary-layer wind-tunnel flow is measured over isolated ridges of varyingsteepness and roughness. The steepness/roughness parameter space is chosento produce flows that range from fully attached to strongly separated. Measurementsshow that maximum speedup at the hill crest is significantly lower than predictedby linear theory and that recovery in the lee of the hill is much slower for stronglyseparated flow over steep terrain. The measurements also show that behaviour ofthe mean and turbulent components of the flow on the downwind side of the ridgeis fundamentally different between separated and non-separated flows. This suggeststhe dominance of much increased turbulence time and length scales in the lee of thehill in association with a production mechanism that scales with the hill length ratherthan the proximity to the surface as on the windward side of the hill crest.     

A laboratory simulation of mesoscale flow interaction with the Alps

A series of laboratory experiments, aimed at the simulation of some aspects of Alpine lee cyclogenesis has been carried out in the rotating tank of the Coriolis Laboratory of LEGI-IMG in Grenoble. Dynamic and thermodynamic processes, typical of baroclinic development triggered by the orography, were simulated. The background flow simulating the basic state of the atmosphere consisted of a stream of intermediate density fluid introduced at the interface between two fluid layers. The structure of the intermediate current was established by mixing fluid obtained from the upper layer of fresh water with fluid removed from the heavier salty layer below.The dynamical similarity parameters are the Rossby (Ro), Burger (Bu) and Ekman (Ek) numbers, although this last, owing to its small values, need not be matched between model and prototype, since viscous effects are not important for small time scales. The flow in both the prototype and laboratory simulation is characterized by hydrostatics; this requires (Ro²δ²/Bu)1 (where δ=H/L is the aspect ratio of the obstacle) which is clearly satisfied, in the atmosphere and oceans, and for the laboratory experiment.A range of experiments for various Rossby and Burger numbers were conducted which delimited the region of parameter space for which background flows akin to that found to the northwest of the Alps prior to baroclinic cyclogenesis events, were observed.One such experiment was carried out by placing a model of the Alps at the appropriate place in the flow field. The subsequent motion in the laboratory was observed and dye tracer motions were used to obtain the approximate particle trajectories. The density field was also analyzed to provide the geopotential field of the simulated atmosphere. Using standard transformations from the similarity analysis, the laboratory observations were related to the prototype atmosphere. The flow and the geopotential fields gave results compatible with the particular atmospheric event presented.     

Blocking in periodic valleys

Results are presented from a study of blocked flow (practically stagnant or recirculating light winds) in periodic valleys in thermally stably stratified ambient conditions. Inviscid and turbulent diffusion cases were modelled numerically to clarify the effects of turbulence on the blocking. The reflection of gravity waves from the top boundary of the hydrostatic model atmosphere was avoided by employing the radiation condition given by Klemp and Durran (1983). The dissipative numerical results are compared with new laboratory experiments which utilized the technique of Baines and Hoinka (1985) to simulate a semi-infinitely deep region.A criterion for the occurrence of blocked flow cannot be defined for the inviscid case except when the Froude number, Fr, based on the peak-to-trough ridge amplitude is less than about 0.4: then blocking is clearly identifiable before wave-breaking occurs. Breaking of waves is evident for Fr as large as 0.75, in agreement with analytical results given by Lilly and Klemp (1979).At small Froude number (Fr 0.5) in the dissipative flow simulations, blocked flow (stagnation) is present in the valleys, but a lee rotor (complete stagnation) is not evident. For order unity Froude numbers, blocking is a wave phenomenon, resulting from wave steepening and overturning or turbulent mixing. A finite thickness is brought to rest or participates in a recirculating flow when it first appears. A strong upward flow appears ahead of the rotor in the valleys, and the downslope wind over the windward side of the valleys is strengthened. Thus the present study shows that conditions for the onset of a rotor, and of stagnant flow, in periodic valleys are different.When blocked flow exists, the amplitudes of gravity waves in the upper layer are only 15% (Fr = 0.3) to 80% (Fr = 1.5) of those given by linear theory; this is supported by observations.     

A Wind-Tunnel Investigation of Wind-Turbine Wakes: Boundary-Layer Turbulence Effects

Wind-tunnel experiments were performed to study turbulence in the wake of a model wind turbine placed in a boundary layer developed over rough and smooth surfaces. Hot-wire anemometry was used to characterize the cross-sectional distribution of mean velocity, turbulence intensity and kinematic shear stress at different locations downwind of the turbine for both surface roughness cases. Special emphasis was placed on the spatial distribution of the velocity deficit and the turbulence intensity, which are important factors affecting turbine power generation and fatigue loads in wind energy parks. Non-axisymmetric behaviour of the wake is observed over both roughness types in response to the non-uniform incoming boundary-layer flow and the effect of the surface. Nonetheless, the velocity deficit with respect to the incoming velocity profile is nearly axisymmetric, except near the ground in the far wake where the wake interacts with the surface. It is found that the wind turbine induces a large enhancement of turbulence levels (positive added turbulence intensity) in the upper part of the wake. This is due to the effect of relatively large velocity fluctuations associated with helicoidal tip vortices near the wake edge, where the mean shear is strong. In the lower part of the wake, the mean shear and turbulence intensity are reduced with respect to the incoming flow. The non-axisymmetry of the turbulence intensity distribution of the wake is found to be stronger over the rough surface, where the incoming flow is less uniform at the turbine level. In the far wake the added turbulent intensity, its positive and negative contributions and its local maximum decay as a power law of downwind distance (with an exponent ranging from −0.3 to −0.5 for the rough surface, and with a wider variation for the smooth surface). Nevertheless, the effect of the turbine on the velocity defect and added turbulence intensity is not negligible even in the very far wake, at a distance of fifteen times the rotor diameter.     

Turbulence Structure in the Wake Region of a Meteorological Tower

A meteorological tower significantly modifies the air flow, the mean windspeed and wind direction as well as the turbulencestructure of the air. Suchchanges can be noticed in particular in the wake region of the tower.Measurementson the 200 m tower ofForschungszentrum Karlsruhewere carried outusing Solent sonic anemometers in the lee of the towerand cup anemometers on both sides.In the wake region, spectral energydensity is increased in the high-frequency range. Superposition of this disturbance spectrum on the undisturbedspectrum yields a `knee' in the resulting spectrum. In the case of low turbulence intensity with stable stratification,a plateau with a constant energy content is observed in front of the knee.This effect is caused by the new production of turbulence energy from the mean flow as well as by an energy transfer fromlarger to smaller vortices. Power spectra in strongly stable conditionsshow a more rapid decrease of intensity in the region where the inertialsubrange is expected.The relevant scales of wake turbulence are derived from the maximum of the disturbance spectrum.Locations of the high-frequency peak do not depend on atmospheric stability,but are controlled mainly by mean wind speed.Apart from the reduction of the mean wind speed, the spectra and cospectra exhibit a strong anisotropy for such cases.The results demonstrate the significant influence of a tower on turbulence spectra in the wake region.     

地形影响的飞机颠簸及其数值仿真实验

首先利用中尺度数值模式ARPS模拟气流过山生成飞行数值仿真所需要的风场，然后利用飞机载荷因数变量方程进行飞机飞行的数值仿真试验。研究结果表明，气流过山产生的山脉重力波由于风切变临界层破碎，一方面能在对流层产生较强的湍流引起晴空飞机颠簸，另一方面也能在山脉背风面产生强烈下坡风，背风转子环流及低空湍流，影响飞机的起飞和着陆。揭示了飞机在过山脉地形背风面所产生的大气湍流中飞行时引起飞机颠簸的物理机制，有助于增强飞机颠簸的预测能力和飞行气象保障能力。     

Turbulence Characteristics Of The Stable Boundary Layer Over A Mid-Latitude Glacier. Part Ii: Pure Katabatic Forcing Conditions

Observations obtained over a glacier surface in a predominantlykatabatic flow and with a distinctwind maximum below 13-m height are presented. The data werecollected using a 13-m high profilemast and two sonic anemometers (at about 2.5-m and 10-m heights).The spectra at frequencies belowthat of the turbulence range appear to deviate considerably fromthe curves obtained by Kaimal andco-workers during the 1968 Kansas experiment. The characteristicsof these deviations are compared tothe observations of others in surface-layers disturbed by anykind of large-scale outer-layer (orinactive) turbulence. In our case the disturbances arelikely to be induced by the highmountain ridges that surround the glacier. Moreover, the deviationsobserved in the cospectra seemto result from an, as yet, unspecified interaction between theinactive outer-layer turbulenceand the local surface-layer turbulence. Near the distinctwind maximum turbulence production ceasedwhile turbulence itself did not, probably the result ofturbulence transport from other levels. Consequently, we studied thelocal similarity relations using _w instead of u_* as an alternative velocity scale. Wellbelow the wind maximum, and for relatively low stability(0< Ri_g <0.2), the flow behaves accordingto well established local-scaling similarity relationshipsin the stable boundary layer. For higherstability (Ri_g > 0.2), and near or above the wind maximum, the boundary-layer structure conforms tothat of z-less stratification suggesting that the eddy sizeis restricted by the local stability ofthe flow. In line with this we observed that the sensibleheat fluxes relate remarkably well to thelocal flow parameters.     

Organized structures in developing turbulent flow within and above a plant canopy,using a Large Eddy Simulation

A Large Eddy Simulation (LES) model representing the air flow within and above a plant canopy layer has been completed. Using this model, the organized structures of turbulent flow in the early developmental stages of a crop are simulated and discussed in detail.The effect of the drag due to vegetation is expressed by a term added to the three-dimensional Navier-Stokes equation averaged over the grid scale. For the formulation of sub-grid turbulence processes, the equations for the time-dependent SGS (Sub-Grid-Scale) turbulence energy equation is used, which includes the effects of dissipation (both by viscosity and leaf drag), shear production and diffusion.The organized structure of turbulent flow at the air-plant interface, obtained numerically by the model, yields its contribution to momentum transfer. The three-dimensional large eddy structures, which are composed of spanwise vortices (rolls) and streamwise vortices (ribs), are simulated near the air-plant interface. They are induced by the shear instability at inflection points of the velocity profile. The structure clearly has a life cycle. The instantaneous image of the structure is similar to those observed in the field observations of Gaoet al. (1989) and in the laboratory flume experiments of Ikeda and Ota (1992). These organized structures also account for the well known fact that the sweep motion of turbulence dominates momentum transport within and just above a plant canopy, and the motion of ejection prevails in the higher regions.     

Measurements of internal waves and turbulence in two-dimensional stratified shear flows

A turbulent stratified shear flow is generated in a towing tank by towing a grid or a circular cylinder through a tank of stratified salt water. The internal waves and turbulence generated in these flows are visualized with shadowgraphs and measured with quartz-coated hot-film probes (up to four probes for velocity fluctuations) and single-electrode conductivity probes (up to four probes for salinity fluctuations) which are towed at the same speed as the obstacle. The velocity and salinity signals are recorded on magnetic tapes. A portion of these signals is processed directly-on-line with a digital computer. From these shadowgraphs and probe measurements, we observe that

1. Far downstream of the obstacle where the turbulence has already subsided, the stratified fluid always has a layered structure. This layered structure persists for a long time, and is a result of the convection of turbulently mixed layers by the mean flow. These results indicate that in the regions of a stably stratified atmosphere and ocean where the turbulence has subsided, one could often find layered structure.
2. There are spectral peaks and valleys in the measured velocity and salinity autospectra when the stratifications are sufficiently strong. Under certain conditions, these spectral peaks tend to lift up the spectral curves to show substantialf ^?5/3 subranges, although the turbulence Reynolds numbers are too low for the flows to have recognizable inertial subranges. This anomalousf ^?5/3 subrange demonstrates the pitfalls of using spectral measurements in thef ^?5/3 subrange to predict the turbulent energy dissipation rate through the Kolmogorov hypothesis.
3. A diagnostic method is developed for distinguishing internal waves from turbulence, utilizing their phase characteristics. The phase characteristics can be conveniently examined from the cospectra and quadrature spectra measurements of: (a), two vertically separated velocity probes; (b), two vertically separated density probes; and (c), a velocity probe and a density probe. This method is demonstrated to be useful in the laboratory and can be applied directly to atmospheric and oceanic measurements to distinguish internal waves from turbulence.
4. From the coherency measurements, it is found that the entire turbulent stratified wake is actually whipping up and down at a frequency corresponding to the Brunt-Väisälä frequency. This indicates that similar stratified shear flows in the atmosphere and in the ocean, such as the jet streams in the atmosphere and the Cromwell current in the ocean, may oscillate vertically, which in turn can induce horizontal oscillation and meandering.