(*Note on viewing the quicktime movies found here)
... And Errata. Please do email me if you catch another mistake, or something that might be a mistake...
(Skip to EquatorMix Cruise Report & HiPADS samples)
(Skip to Bi-PADS development)
Front evolution during AESOP (60s ave)
Front in AESOP with waves
Surface waves are among the first things anyone observing the ocean encounters. For many researchers, they are merely a nuisance, causing large motions that make instruments fail, overwhelm the delicate signals expected from the deeper waters, and (of course) make many researchers sea-sick and less able to think clearly. However, surface waves are still an active area of research in themselves, since they affect the exchanges of momentum (wind stress), moisture and aerosols, and gas-uptake between the surface layer of water (the oceanic mixed layer) and the air.
Surface waves can directly induce currents in a variety of ways. One is that wave "groups," where a few waves in a row are much larger than the surrounding waves, will induce a trapped "long-wave response," where a mean (Eulerian) flow is set up that runs counter to the direction of wave propagation. The orbital motion of the waves also gives rise to a slight net forward "Stokes drift," so the net motion of actual water parcels is a combination of the Eulerian mean and this drift. This "net parcel motion" is called the Lagrangian mean flow, and represent the actual displacement per unit time of the water itself. In the case of the group-forced long-wave response, the wave-forced Eulerian mean roughly cancels the Stokes drift, so the Lagrangian mean (parcel motion) is more nearly zero. For more details, see
Another way waves can generate currents is by breaking, and transferring the large momentum they carry to the underlying flow. This can be particularly effective in the nearshore region, where waves break due to shoaling (see "Nearshore Waves and Currents" below).
Waves also break out in the open ocean, and inject bubbles many meters into the water as well as sending spray up into the air. A major outstanding problem in oceanography lies in understanding the interface between the air and sea, and how this is affected by breaking waves. The challenge is to study the minute details of the motion that are relevant to bubble dynamics and gas exchange, without suffering instrument damage from the large breaking waves themselves. For example, here is a brief description of a recent attempt to probe upward into the crest of breaking waves using sound.
Surface waves also have a strong indirect influence on the mixed layer, by stimulating or augmenting a form of wind-induced motion known as "Langmuir circulation" (see the next section of my page, below). While this arises as a fairly complex instability problem, there are also more direct wave-current interactions that can be important.
Much of the confusion about the interaction of waves and currents arises from the conceptual division of the flow into "Eulerian" (fixed location) versus "Lagrangian" (fluid-following) frameworks. For example, the difference between the mean velocity at a fixed point versus that of a drifting particle is the "Stokes' drift," due to the presence of waves. Often the results of seemingly complex analyses from one viewpoint can be more simply interpreted from the other. In any event, the relation between (often intrinsically Lagrangian) dynamic constraints and the measurements (normally Eulerian) must be borne in mind. An attempt to write this up in an understandable way is found here:
For the sake of WISE attendees, here is a copy of the lecture I gave as a PDF with movies (44 MB big). The movies only work in Acrobat. You have to click on the (first) image to start each movie, and then click outside the movie before you can go to the next slide. Convenient? Be glad if it even works. Adobe(tm).
Mixing associated with Langmuir circulation (LC, described below) can be important in the long-term evolution of the mixed layer. The strength and depth of these structures influence the sea surface temperature, and are important to weather and climate as well as to biology and chemistry.
Langmuir circulation is a form of motion in the wind-mixed layer (rougly speaking, the top 100 m or so that "feels" the wind stress pretty directly). When the wind blows, the water in this layer tends to "spiral" downwind, converging along the surface towards downwelling "streaks" (often made visible by floating seaweed, foam, or oils that collect there but are too buoyant to sink). From there the water, which is moving downwind faster than the rest of the surface water, sinks to the base of the mixed layer, then diverges away to the sides, and gradually upwells between the streaks to start the next cycle.
This form of motion is reinforced by an interaction with the waves. First, the downwind current maximum at the downwelling streaks tends to refract the waves, "bending" the wave momentum away from the streak slightly. Since the waves carry momentum, and the total momentum is conserved, something must be accelerated towards the streak to keep the net momentum steady as the waves are refracted away. That something is the water itself: the water at the surface is pushed toward the streaks by the refracting waves, reinforcing the overall spiral pattern described above. As the water flows along the surface toward the convergence or "streak," it is accelerated by the wind, so by the time it gets there it is going faster than the surrounding water, reinforcing the downwind "jet." Mathematically, the initial stage of this instability can be expressed in a form that indicates exponential growth. Later, finite-amplitude motion effects must stabilize the flow pattern (under steady winds). A somewhat eclectic (some would say "quirky") review of the history of Langmuir circulation is given in
Next we are considering possible interactions between LC and internal waves, based on data from the "Hawaiian Ocean Mixing Experiment" or HOME. Some initial findings and considerations are described here: "Do internal waves and Langmuir circulation interact?" (from 1/23/04).
Recently, with J. MacKinnon, J. Polton, and A. Tejada-Martinez, we have been working on a numerical model to simulate LC over an initially still stratified layer (an "LES" type model). We find the combination of LC and the inertial current generated by the sudden onset of wind to generate very high-frequency internal waves: see Rapid generation of high-frequency internal waves beneath a wind and wave forced oceanic surface mixed layer. (GRL 2008, 2.4 MB PDF)
Results from our group over the past couple decades, mainly from observations made from the R/P FLIP, are summarized in Observations of Langmuir Circulation From FLIP. and also in "Long-Range Acoustic Doppler Array Measurements of Surface Velocities" (HTML).
Some findings related to mixing the surface layer of the ocean are described here, with more details in the paper "Evolution of Langmuir circulation during a storm" (Smith 1998; pdf, 620k). Another addendum (pdf, 108k) expands the scaling analysis to include additional data, and focusses more attention on the puzzling finding that the surface rms velocity scales more nearly with the waves alone than with the wind (although these are themselves strongly correlated).
A major field experiment called "SandyDuck" took place at the USACE's "Field Research Facility", located near Duck, NC, on the barrier islands off the East Coast of the USA. The focus time ran from September through November, 1997. Many people and many measurements were involved. The goal is to understand the dynamics of flows near shore. These flows are forced by a combination of wave breaking, winds, and topographic effects. Of particular interest is the occurrence, form, and dynamics of rip currents. (Here, "rip currents" are loosely defined as narrow, offshore-directed flows extending some distance seaward from the shore through the surf zone.)
Some recent work has focused on separating the flow into a part that is horizontally non-divergent (but can have vorticity) and irrotationsl (but can be divergent) parts. The technique is based on the FFT, so it's quite fast. It's applied to two examples: in one, a plume of fresher, lighter water sweeps through the area (likely an out-flow from Chesapeake Bay after rain); in the other, a possible "vortex pair" is seen to propagate through the study region. See Smith (2008), Vorticity and divergence of surface velocities near shore. (pdf 2.6MB)
Surface waves are important in many of today's concerns. For example, in planning shipping routes, estimating storm damage risks, and driving coastal erosion. Global models of directional wave spectra are routinely run to assist in these endeavors; however, a model is only as good as the information going into it. The better the observations, the better the results.
Ocean waves can be tricky to measure- the instruments must survive the storms that make the biggest waves. Most modern measurements are fairly simple, giving us estimates of the size of the waves and (perhaps) the mean direction of wave propagation. But to model waves from several storms at once, it would be better to have more detailed directional information. Satellite pictures have great potential in helping with this; however, the "snapshot" pictures of the waves arising from satellites have a 180-degree ambiguity: they can show the orientation of wave crests quite accurately, but cannot tell which direction they are moving. To do that, one needs arrays that resolve the waves in both space and time- for example, Doppler sonar beams extending hundreds of meters along the surface, sampling every second or so.
Of course, little of this would be possible without the Doppler sonar technology needed to make measurements over large ranges with such precision and resolution. The "Ocean Physics Group" (OPG) here at SIO has a group of world-class engineers and technicians (M. Goldin, M. Bui, A. Madduri, T. Hugens, A. Aja) dedicated to the design, construction, testing, deploying, trouble-shooting, and analyzing data from (whew!) Doppler sonar systems (and any other hare-brained measurement ideas we come up with). They are supported by two PIs, Rob Pinkel (who, frankly, provides most of the support) and myself. We also have several graduate students and undergraduate interns.
Some of the relevant technology was described above, in the "Some earlier findings about LC" section, and some of the considerations about the effective use of these instruments are outlined in these papers:
We are currently developing a bi-static phased-array Doppler sonar system ("Bi-PADS") with the hope of getting a 2-D, 2-component map of vertical and horizontal velocities on a plane extending right up into the crests of shoaling and breaking waves (sponsored by NSF). Initially we'll try for the 2 in-plane components; subsequently we plan to try a combination of radial and cross-plane components, as the next step towards getting all 3 velocity components over the whole 2-D planar area. There are many technological obstacles to be dealt with (aside from the sheer complexity of the system itself) - including the effects of strong advection on the two acoustic paths, and the substantial variations in the speed of sound in the presence of bubbles. Optimizing the match between the images from the distinct acoustic paths will require some level of estimation of the sound speed anomolies - but this too is valuable information, as it related directly to the bubble content of the fluid mixture.
And a glimpse of data - even & odd pings alternate transmits from either side. The surface is up near 6m, and you can see some "lumps" in the water sloshing (very slowly!) with the waves. It is reassuring that at least some of the spots are in the same place no matter which side the sound comes from. Proper location of intensity features is sensitive to the detailed geometry of deployment. Playback is about 1/2 real time speed: