Etudes et évaluation de processus océaniques par des hiérarchies ...
Etudes et évaluation de processus océaniques par des hiérarchies ... Etudes et évaluation de processus océaniques par des hiérarchies ...
118 CHAPITRE 4. ETUDES DE PROCESSUS OCÉANOGRAPHIQUES 4.6 Estimation of friction parameters and laws in 1.5D shallow-water gravity currents on the f-plane by data assimilation tel-00545911, version 1 - 13 Dec 2010
4.6. ESTIMATION OF FRICTION PARAMETERS AND LAWS IN 1.5D SHALLOW-WATER GRAVI Ocean Dynamics DOI 10.1007/s10236-008-0151-8 Estimation of friction parameters and laws in 1.5D shallow-water gravity currents on the f-plane, by data assimilation Achim Wirth · Jacques Verron Received: 11 June 2008 / Accepted: 15 September 2008 © Springer-Verlag 2008 tel-00545911, version 1 - 13 Dec 2010 Abstract A 1.5-dimensional, 1.5-layer shallow water model and an ensemble Kalman filter are used to evaluate the feasibility of estimating friction parameters and determining friction laws of oceanic gravity currents. The two friction laws implemented are a linear Rayleigh friction and a quadratic drag law. We demonstrate that the assimilation procedure rapidly estimates the total frictional force, whereas the distinction between the two laws is evolving on a slower time scale. We also demonstrate that parameter estimation can, in this way, choose between different parametrisations and help to discriminate between physical laws of nature by estimating the coefficients presented in such parametrisations. Keywords Ocean dynamics · Gravity current · Friction laws · Parameter estimation · Data assimilation 1 Introduction Buoyancy forces caused by density differences of fluids are a major source of fluid motion in nature. When these forces act adjacent to topography, gravity cur- Responsible Editor: Tal Ezer A. Wirth (B) · J. Verron LEGI / MEOM, CNRS, BP 53, 38041 Grenoble Cedex 9, France e-mail: achim.wirth@hmg.inpg.fr J. Verron e-mail: jacques.verron@hmg.inpg.fr rents are created. The major part of the deep and intermediate waters of the world ocean and marginal seas have done at least part of their voyage to the deep in the form of a gravity current. Oceanic gravity currents play a role of paramount importance in the formation of the water masses of the deep ocean and are thus key to understanding the oceanic component in the earth climate system. For the major part of oceanic gravity currents, the earth’s rotation plays an important role, as they evolve on a timescale much larger than the rotation period of the planet earth. The dominant force balance is, thus, between gravity and the Coriolis force, that is geostrophic. Indeed, when the turbulent fluxes of water masses and momentum are neglected, such gravity currents flow along the inclined ocean floor, changing neither depth nor composition. Turbulent fluxes of tracers, that is mixing, entrainment and/or detrainment, determine the change of the water mass. Together with the turbulent fluxes of momentum at the floor and interface of the gravity current, they determine the change in position of the gravity current. The answer to the question of the evolution of a gravity current lies, thus, in the determination of the laws and parameters of its turbulent fluxes. The inter-comparison study of several ocean general circulation models (OGCMs) (DYNAMO Group 1997; Willebrand et al. 2001) clearly determined the high sensitivity of the meridional heat transport (in the North Atlantic) to the representation of the (Denmark Strait and Faroe Bank Channel) overflows. The meridional heat transport is very sensitive to small-scale details and turbulence in the overflow regions not explicitly resolved in OGCMs and the numerical representation (the grid) of the overflow regions in the OGCMs. Dedicated research on these sub-grid-scale processes is key
- Page 73 and 74: 4.2. THE PARAMETRIZATION OF BAROCLI
- Page 75 and 76: 4.2. THE PARAMETRIZATION OF BAROCLI
- Page 77 and 78: 4.2. THE PARAMETRIZATION OF BAROCLI
- Page 79 and 80: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 81 and 82: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 83 and 84: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 85 and 86: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 87 and 88: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 89 and 90: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 91 and 92: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 93 and 94: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 95 and 96: 4.3. A NON-HYDROSTATIC FLAT-BOTTOM
- Page 97 and 98: 4.4. TILTED CONVECTIVE PLUMES IN NU
- Page 99 and 100: 4.4. TILTED CONVECTIVE PLUMES IN NU
- Page 101 and 102: 4.4. TILTED CONVECTIVE PLUMES IN NU
- Page 103 and 104: 4.4. TILTED CONVECTIVE PLUMES IN NU
- Page 105 and 106: 4.4. TILTED CONVECTIVE PLUMES IN NU
- Page 107 and 108: 4.4. TILTED CONVECTIVE PLUMES IN NU
- Page 109 and 110: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 111 and 112: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 113 and 114: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 115 and 116: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 117 and 118: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 119 and 120: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 121 and 122: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 123: 4.5. MEAN CIRCULATION AND STRUCTURE
- Page 127 and 128: 4.6. ESTIMATION OF FRICTION PARAMET
- Page 129 and 130: 4.6. ESTIMATION OF FRICTION PARAMET
- Page 131 and 132: 4.6. ESTIMATION OF FRICTION PARAMET
- Page 133 and 134: 4.6. ESTIMATION OF FRICTION PARAMET
- Page 135 and 136: 4.6. ESTIMATION OF FRICTION PARAMET
- Page 137 and 138: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 139 and 140: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 141 and 142: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 143 and 144: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 145 and 146: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 147 and 148: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 149 and 150: 4.7. ON THE BASIC STRUCTURE OF OCEA
- Page 151 and 152: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 153 and 154: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 155 and 156: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 157 and 158: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 159 and 160: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 161 and 162: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 163 and 164: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 165 and 166: 4.8. ESTIMATION OF FRICTION LAWS AN
- Page 167 and 168: 4.9. ON THE NUMERICAL RESOLUTION OF
- Page 169 and 170: 4.9. ON THE NUMERICAL RESOLUTION OF
- Page 171 and 172: 4.9. ON THE NUMERICAL RESOLUTION OF
- Page 173 and 174: 4.9. ON THE NUMERICAL RESOLUTION OF
4.6. ESTIMATION OF FRICTION PARAMETERS AND LAWS IN 1.5D SHALLOW-WATER GRAVI<br />
Ocean Dynamics<br />
DOI 10.1007/s10236-008-0151-8<br />
Estimation of friction <strong>par</strong>am<strong>et</strong>ers and laws in 1.5D<br />
shallow-water gravity currents on the f-plane,<br />
by data assimilation<br />
Achim Wirth · Jacques Verron<br />
Received: 11 June 2008 / Accepted: 15 September 2008<br />
© Springer-Verlag 2008<br />
tel-00545911, version 1 - 13 Dec 2010<br />
Abstract A 1.5-dimensional, 1.5-layer shallow water<br />
mo<strong>de</strong>l and an ensemble Kalman filter are used to evaluate<br />
the feasibility of estimating friction <strong>par</strong>am<strong>et</strong>ers<br />
and d<strong>et</strong>ermining friction laws of oceanic gravity currents.<br />
The two friction laws implemented are a linear<br />
Rayleigh friction and a quadratic drag law. We <strong>de</strong>monstrate<br />
that the assimilation procedure rapidly estimates<br />
the total frictional force, whereas the distinction b<strong>et</strong>ween<br />
the two laws is evolving on a slower time scale.<br />
We also <strong>de</strong>monstrate that <strong>par</strong>am<strong>et</strong>er estimation can,<br />
in this way, choose b<strong>et</strong>ween different <strong>par</strong>am<strong>et</strong>risations<br />
and help to discriminate b<strong>et</strong>ween physical laws of nature<br />
by estimating the coefficients presented in such<br />
<strong>par</strong>am<strong>et</strong>risations.<br />
Keywords Ocean dynamics · Gravity current ·<br />
Friction laws · Param<strong>et</strong>er estimation ·<br />
Data assimilation<br />
1 Introduction<br />
Buoyancy forces caused by <strong>de</strong>nsity differences of fluids<br />
are a major source of fluid motion in nature. When<br />
these forces act adjacent to topography, gravity cur-<br />
Responsible Editor: Tal Ezer<br />
A. Wirth (B) · J. Verron<br />
LEGI / MEOM, CNRS, BP 53,<br />
38041 Grenoble Ce<strong>de</strong>x 9, France<br />
e-mail: achim.wirth@hmg.inpg.fr<br />
J. Verron<br />
e-mail: jacques.verron@hmg.inpg.fr<br />
rents are created. The major <strong>par</strong>t of the <strong>de</strong>ep and<br />
intermediate waters of the world ocean and marginal<br />
seas have done at least <strong>par</strong>t of their voyage to the<br />
<strong>de</strong>ep in the form of a gravity current. Oceanic gravity<br />
currents play a role of <strong>par</strong>amount importance in the<br />
formation of the water masses of the <strong>de</strong>ep ocean and<br />
are thus key to un<strong>de</strong>rstanding the oceanic component<br />
in the earth climate system. For the major <strong>par</strong>t of<br />
oceanic gravity currents, the earth’s rotation plays an<br />
important role, as they evolve on a timescale much<br />
larger than the rotation period of the plan<strong>et</strong> earth. The<br />
dominant force balance is, thus, b<strong>et</strong>ween gravity and<br />
the Coriolis force, that is geostrophic. In<strong>de</strong>ed, when the<br />
turbulent fluxes of water masses and momentum are<br />
neglected, such gravity currents flow along the inclined<br />
ocean floor, changing neither <strong>de</strong>pth nor composition.<br />
Turbulent fluxes of tracers, that is mixing, entrainment<br />
and/or d<strong>et</strong>rainment, d<strong>et</strong>ermine the change of the water<br />
mass. Tog<strong>et</strong>her with the turbulent fluxes of momentum<br />
at the floor and interface of the gravity current, they<br />
d<strong>et</strong>ermine the change in position of the gravity current.<br />
The answer to the question of the evolution of a gravity<br />
current lies, thus, in the d<strong>et</strong>ermination of the laws and<br />
<strong>par</strong>am<strong>et</strong>ers of its turbulent fluxes.<br />
The inter-com<strong>par</strong>ison study of several ocean general<br />
circulation mo<strong>de</strong>ls (OGCMs) (DYNAMO Group 1997;<br />
Willebrand <strong>et</strong> al. 2001) clearly d<strong>et</strong>ermined the high sensitivity<br />
of the meridional heat transport (in the North<br />
Atlantic) to the representation of the (Denmark Strait<br />
and Faroe Bank Channel) overflows. The meridional<br />
heat transport is very sensitive to small-scale d<strong>et</strong>ails<br />
and turbulence in the overflow regions not explicitly<br />
resolved in OGCMs and the numerical representation<br />
(the grid) of the overflow regions in the OGCMs. Dedicated<br />
research on these sub-grid-scale processes is key