Combining waves and circulation, perspectives for
operational oceanography
web forum
(Updated 12 August 2004)
How
to participate:
Please send
e-mails to Fabrice, he will put your contributions
on this page, with a link to your e-mail address (unless you do not accept)
for direct replies.
Opening statements:
Klaus Hasselmann: I find your approach to wave
research and operational oceanography and wave forecasting and analysis very
encouraging. Indeed, they agree entirely with my own views, and so I wish
you all success in your endeavours, including the symposium you are organizing.
However, I am afraid I will not be able to attend myself.
Owen Phillips: I am certainly interested in
the topic and would be delighted to help in your considerations about the
best directions for research that will be needed to solve important problems
in civil and naval contexts. I have, however, been doing a lot of travelling
recently and have a couple more obligations this year that I cannot escape
[...]. The areas you mention are all interesting scientifically and important.
For example, in the topic of "pollutant drift", there are questions not only
of drift in currents and wind-induced surface drift but also of dispersal
by meso-scale ocean turbulence, partly wave-generated, and break-up by whitecaps
of various scales. There are considerations of the back reaction of slicks
on the waves, attenuation (not really well understood physically even now),
and wave dispersion. Also, questions of safety at sea include the occurrence
and effects of extreme wave events, which, like
extreme atmospheric events such as tornadoes whose occurrences in space and
time are not predictable in specific detail. It is probably feasible to recognize
scenarios where they are likely to occur and the characteristics of the events
themselves in time to issue warnings to mariners and oil drilling operators.
Another area between meteorology and oceanography is the question of small
boat icing in arctic storms: what are the precise factors (humidity, wind
speed, wave breaking density, scale and directional distribution, etc) that
influence the icing rate at various heights above sea level? There are lots
of other research areas of the kind you are seeking. [...]
This
note is to wish you well for a great meeting in June and to suggest a couple
of areas where, I think, fundamental and useful advances might be made. One
is, I think, is the area of whitecapping. Important advances have been made
by Mike Banner and his group on the inception of whitecaps, but measurements
on the distribution of breaking front length p.u. area with respect to speed
of advance by us (JPO) and by Ken Melville (Nature) show unresolved differences.
This function is central to upper ocean mixing
and dissipation (Terray et al jpo, 1996) and meso-scale dispersion. So far,
the observations have been analysed in terms of wind speed only and show
a lot of scatter, but the effect of wave age must be important also. Young
seas (short fetch or duration, rising winds) and old seas with falling or
distant winds (c/u*)>30 or so, must have very different wave breaking
patterns that have not been isolated. They should not be analysed using statistical
equilibrium ideas, but that's all that has been done so far, I think.
Related, I think, are
questions of the characteristic structure of the turbulence generated by whitecaps.
A breaking wave loses momentum and supplies a local impulse to the water
which produces a large structure something like half a vortex ring looping
down. from the surface. Vortex rings, even turbulent
ones, are generally long-lived structures. How do they influence upper ocean
dynamics and mixing? I don't think we know much quantitatvely about the break
up of slicks by incident breaking waves in terms of slick characteristics
(which are messy -- pardon the pun!) and the physical functions above describing
the breaking dynamics. How can the processes be retarded? or accelerated?
I
wish I could be with you, but I am sure you will have a stimulating and useful
meeting.
Fabrice
Ardhuin (ardhuin@shom.fr) : Looking back at the book "Directional
ocean wave spectra" edited in 1991 by Beal with proceedings from a symposium
in the wake of the LEWEX experiment, I read the panel discussion and Klaus Hasselmann's enlightening
banquet speech "Waves dreams and
visions". He envisionned an integrated observing and
modeling system for global monitoring, including ocean circulation models,
atmospheric models... and wave models (see Hasselmann's figure 1).
It struck
me that this was certainly a very good idea, not only for climate applications
but even more so for coastal matters, but it looked like we are a bit behind
schedule. Indeed the "wave model" that is today in use in this general framework
is ... the bulk formulae and drag coefficient for air-sea exchanges and bottom
friction plus some ad hoc tinkering with Stokes drift (the residual drift
due to wave motion) for surface currents. Besides this we have very good
(relatively speaking) wave models, able to forecast wave heights within 10
% on a global scale (almost the measurement error) 3 days ahead without any
need for data assimilation. These wave models should give accurate wind stresses,
air-sea fluxes, EM-bias for altimeters, wave set-up at the coasts, erosion
rates, and surface drift velocities. All these quantities are, by themselves,
very useful for society, but they are often better combined with currents
from ocean circulation models to provide all the information needed.
So what is wrong ? The oceanography community
is too scattered and we need to talk again across these new fences that are
the so-called "discipline boundaries". People like Walter Munk, Harald Sverdrup
or Klaus Hasselmann did not care about these differences and would work one
day on ocean waves, the next on ocean acoustics, general circulation or climate
change and its economics. What I would like to stress here, is that a coherent
use of today's model (ocean, waves, atmosphere) is possible, right now with existing tools and theories, and tomorrow
with a bit more research into mixing processes and boundary layers. This
will also require a novel way of using observation and may greatly benefit
from new techniques (in particular remote sensing) that are being developed
today. This coherent use should vastly improve the accuracy and relevence
of the products delivered by the "operational oceanography". "Operational
oceanography", at least for the French Navy is not "running an ocean circulation
model" but rather "providing information and decision aids about the ocean
to end users". I believe that for all users "operational oceanography" should
focus on its users' needs (governments, fisheries, Met Officies, Navies, mariners,
surfers ... and oil companies) and organize around these, this will naturally
oblige "operational people" to use whatever meteorology and oceanography
they need. Because the effort to run an ocean circulation observation and
modelling system is so large, resources have to be pooled together, and the
GOOS effort is really admirable. We now have to use make sure that the resources
put in place satisfy the user's need. We still poorly understand surface
mixing and surface drift currents, and we need more than higher resolution
OGCM and fancy data assimilation techniques. We need to take the ocean for
what it is: a complex environment with motions at all scales, surface films
over 10% of its surface, variable sediments (silt, sand, rocks, mud ...),
biology ...
Take the case of the oil pollution due to the sinking of the tanker Prestige
The questions I have are of the following order: In coastal areas, should
we use wave models for forcing circulation models, should we rather couple
two-ways waves and circulation ? How should the
KPP or Mellor-Yamada turbulence closure models be modified to account for
wave-induced stresses and Langmuir circulations ?
(I've just seen a paper by Jim McWilliams and Peter Sullivan
addressing this latter aspect: Spill Science and Technology Bulletin, vol
6, p 225-237, 2000). Should we rather couple waves and atmosphere
(as done at ECMWF) or waves and ocean, or the three together
? Has anybody tried
?
Your comments (specific
scientific issue, practical problem, interesting
papers ...):
Selected references (PDF)