A coherent framework for forecasting currents, waves and drift: 1. What we do at SHOM 2. Hydrodynamics theory 3. First results 4. Perspectives on remote.
Post on 19-Dec-2015
216 Views
Preview:
Transcript
A coherent framework for forecasting currents, waves and drift:
1. What we do at SHOM
2. Hydrodynamics theory
3. First results
4. Perspectives on remote sensing
Dr. Fabrice ARDHUIN, Nicolas Rascle
(SHOM/CMO, Brest, France)Dr. Alastair Jenkins,
(Bjerknes Center, Bergen, Norway)Dr. Bertrand Chapron (Ifremer/DRO/OS)
Funding from: Aurora program (Norway-France), and project MERCATOR
Plateaucontinental
Açores
Madère
EspagnePortugal
France
GrandeBretagne Plateau
continental
1998: first forecast of the SOPRANE system
Espagne
Portugal
France
GrandeBretagne
Climatology (1990)
Barotropic streamfunction
Ardhuin and others - 3Wave effects in the upper ocean – STW-SAR, Brest 2004
Center for Military Oceanographyfrom the early 1990s to today
Ardhuin and others - 3Wave effects in the upper ocean – STW-SAR, Brest 2004
Ocean circulation modelling today
HYCOM2004: Mercator, and …
2006: coastal model demonstrator : 1-2 km resolution from Dover Straits to Gibraltar(this is a 10 M€ program: field data, computer, HF radars, contribution to Mercator …)
Ardhuin and others – 4Wave effects in the upper ocean – STW-SAR, Brest 2004
One basic problem of today’s current models:
-Surface drift is about 2-3% of wind speed
-Turbulence is very strong near the surface -> uniform profiles -> small surface velocity (0.5 – 1% of wind)
(Agrawal et al. 1992, Craig and Banner 1994, Mellor and Blumberg 2004)
But what velocity do we talk about ?
-> ad hoc empirical fix (e.g. MOTHY …)
Ardhuin and others – 6Wave effects in the upper ocean – STW-SAR, Brest 2004
Motions at the air-sea interface
Ingredients of surface drift:
Ardhuin and others – 7Wave effects in the upper ocean – STW-SAR, Brest 2004
Motions of air and water (no oil, no ship …) Wave breaking
Langmuir circulations
eddy mixing
(internal wave breaking ...)Eddy viscosity
Kz
z = 0
z = - Hs ~ z0
total wind stress (momentum flux between atmosphere and ocean)
wave induced stress (wind input): in tangentialwind stress a - in
z = - <h>z = - <h> +
Sediments
Drift velocityue+ust
net momentum uptake for waves : in - dis
(growth-dissipation)
wave dissipation stress ("virtual wave stress" + breaking) : dis
Mixed boundary layerwaves+current
z = - <h> +
Mixing processes
thermocline
radiationstress
radiationstress
I
satellite
Ardhuin and others – 8Wave effects in the upper ocean – STW-SAR, Brest 2004
A general 3D formalism (Ardhuin & Jenkins submitted to JFM 2004,
extension of Mellor, JPO 2003): Mellor used a vertical coordinate transform from z to :
with due to waves
This can be re-derived from the GLM of Andrews & McIntyre 1978):
Ardhuin and others – 9Wave effects in the upper ocean – STW-SAR, Brest 2004
Mixing parameterization: a GLM-average TKE equation
TKE production by waves:
TKE production due to « Stokes drift shear »
Used by Tolman & Chalikov 1996
Ardhuin and others – 10Wave effects in the upper ocean – STW-SAR, Brest 2004
Application to swell dissipation(Ardhuin et Jenkins, submitted to JPO, 2004)
Ardhuin and others – 11Wave effects in the upper ocean – STW-SAR, Brest 2004
First application: the surface mixed layer
What is the vertical profile of Tdis ? (we are working on this, paper in preparation for J. Geophys. Res.)
Tdis is the momentum flux from waves to currents due to wave dissipation (viscosity, breaking, wave-turbulence interaction).
+ b.c. on momentum: - in Wind stress – wave-induced stress
+ b.c. on TKE flux(e.g. Mellor and Blumberg 2004, Janssen & al. 2004)
Coriolis force: waves and mean flow Vertical mixingWave dissipation stress
- Tdis (z)
Ardhuin and others - 9Ardhuin and others – 12Wave effects in the upper ocean – STW-SAR, Brest 2004
1D Mixed layer. No stratification (Craig and Banner, 1994)Kz = Sm q l, l=max[z0 ,0.41(z0-z)] based on Mellor-Yamada 2.5
q = sqrt(b) from TKE equation : db/dt = production + transport - dissipation
With P-M wave spectrum, z0= Hs
(Mellor and Blumberg 2004) U10=10 m/s
uL = U = û + Us = ue + ust
“Classical”, no waves z0=0.1 m. U10=10 m/s
Wind stress
Ardhuin and others – 13Wave effects in the upper ocean – STW-SAR, Brest 2004
HYCOM with wave forcing
1st realistic application : Prestige oil spill hindcast
HYCOM 1/3 degré ATL+MED, assimilating altimetry, forcing: ARPEGE winds + WAM (ECMWF) waves
Standard HYCOM
Ardhuin and others – 14Wave effects in the upper ocean – STW-SAR, Brest 2004
Wave heights from same image(tiled “imagettes” processed as level 2)
New observation methods: Doppler signal from Synthetic Aperture Radars, ATI and/or Doppler centroïd: here Envisat images in VV polarization
processing: Ifremer – Boost Technologies
Verfication of theory: perspectives on remote sensing
Ardhuin and others - 12Wave effects in the upper ocean – STW-SAR, Brest 2004
Ardhuin and others – 15Wave effects in the upper ocean – STW-SAR, Brest 2004
Doppler velocity UD (m/s)
EN
VIS
AT
AS
AR
, wid
e sw
ath,
VV
com
plex
, Dop
pler
cen
troï
d an
alys
is©
ES
A, B
oost
Tec
hnol
ogie
s (
http
://w
ww
.boo
st-t
echn
olog
ies.
com
)
HF radars measure « surface drift »: Sea trials with Uni. Toulon, 2003 (VHF)
Deployment of a HF-radar system, 2005 (funding: DGA research programs)
Ardhuin and others – 17Wave effects in the upper ocean – STW-SAR, Brest 2004
ardhuin@shom.frhttp://www.shom.fr/fr_page/fr_act_oceano/vagues/vagues.htm
http://www.shom.fr/fr_page/fr_act_oceano/vagues/
PLUS/PUBLIS/index_f.html http://surfouest.free.fr/WOO2003/
Conclusions:
- Surface velocities are not fully understood (work under
way)
- Today’s models are not coherent
- New remote sensing data (that we hope to further validate in the next 2 years)
APPENDIX: example of coastal wave forecast validation at Blancs Sablons Beach, just south of RCC Corsen (measurements: March-April 2004)
Observations
NB: none of the models include tidesas yet (that will be “Surfouest V2”).
CREST (ray-tracing), Initialized with WW3(“surfouest V1”)
REF-DIF initialized with WW3(“surfouest V0”)
Offshore waves
top related