|
The Coastal Oceanography
group POC was formed in 2001 and is constituted of scientists from
LEGOS and LA. POC's primary objective is the study of ocean circulation
at regional and coastal scales. The approach is essentially based
on numerical modelling, by means of the in-house models Symphonie
and T-UGO. Substantial effort is focused on model development, such
as in the investigation of several grid-types (C-grid, hybrid coordinates
and Unstructured grids), as well as the application of different
hypotheses (hydrostatic, non-hydrostatic).
To date, the POC models
T-UGO and Symphonie have been used to study gravity waves, wind-induced
currents, river plumes, along-slope circulations, dense water formation.
Several
POC-members are involved in biogeochemical and sediment transport
modelling. This field is particularly challenging due to the complexity
of ecological processes and their strong dependence on physical
conditions. This requires close collaboration with the POC ocean
physics modellers, notably in coupling pelagic models with the POC
ocean circulation models. Furthermore, POC closely collaborates
with external groups involved in observational studies, which is
particularly important for model validation and physical and/or
ecological interpretation.
Data assimilation is also
an important research topic. In the first few years of POC's existence
data assimilation was mainly used for optimisation of the barotropic
tide model. At the present time, 3D regional applications are conducted.
Since it is based on EOF analysis of ensembles of runs, this branch
requires substantial computing resources provided by a cluster of
linux-pc's. Satellite altimetry data is an important potential source
of coastal observational data, but several methodological problems
specific to the coastal zone are to be overcome. Therefore, POC
develops new methods for satellite altimetry data treatment.
|
Modelling:
The
3D circulation models are based on finite difference and finite volume
models developed by the group. The vertical coordinates, lateral boundary
conditions and the formulation of air-sea fluxes are specific points
of detailed examination. Variational tools devoted to the initialisation
of models from observations and/or OGCM output are an originality
of our group.
The modelling strategy is based on grid-nesting from the regional
scale that is forced by OGCM data up to the chosen scale. The resolution
ratio of successive grid meshes varies between 2 and 5.
Model Outputs Analysis:
The model output is analysed
using various tools, involving from wavelet to EOF-analysis, and combinations
thereof. In practice, our approach is based on the analysis of ensembles
(or series) of simulations. In order to provide meaningful statistics,
an ensemble should contain on the order of 100 runs, for which a linux
cluster is used.
Coastal Altimetry:
Satellite altimetry data is generally used
the framework of offshore circulation studies. Indeed, in coastal
zones the use of altimetry data is subject to technical constraints
involving continental influence (making radiometric corrections tricky),
the resolution of geophysical corrections (high-frequency sea level
variations associated with tides and atmospheric pressure forcing),
uncertainties in mean sea level and, finally, the spatial and temporal
resolution, which is low with respect to the scale of variability
of coastal ocean processes. However, on the one hand post-processing
of altimetry data for coastal research purposes has seen recent improvements
and on the other, new satellite projects will provide an improved
spatial description of the sea surface.
Data Assimilation:
Data
assimilation is required for short-term forecasting but also to get
realistic simulations at annual or multi-annual time-scales. It is
a necessary condition to accurately quantify the budgets of carbon
and associated elements in the coastal zone. We develop ensemble methods
which allow for empirical characterisation of model errors, and their
association with errors in forcing terms. The application of these
methods to 2D and 3D models is well-adapted to the complexity of coastal
physics.
Operational Oceanography:
In
collaboration with the engineering company Noveltis, the POC has developed
an operational system which provides weekly predictions of the ocean
circulation in the North-western Mediterranean. |
Gravity
Waves:
We
feel it necessary to simultaneously take into account high frequency
(external gravity waves induced by tide and atmospheric pressure)
as well as low frequency dynamics (circulations induced by wind and
density gradients) in order to fully represent processes such as turbulent
mixing and the generation of internal gravity waves. Thus, we develop
a large-scale barotropic modelling approach which provides boundary
conditions to coastal models, which allows the representation in the
coastal zone of gravity waves formed far way.
An important effort is furthermore devoted to the accurate representation
of sea level variations in the Mediterranean and on the European Atlantic
shelf with the T-UGO-2D model. The non-linear tide is the object of
particular attention. Finally, the generation of internal waves by
the barotropic tide is studied in the Bay of Biscay using the Symphonie
3D model. A similar study starts in the Southwest Pacific with the
finite volume model T-UGO-3D.
3D Circulations:
Exchanges
between the coastal zone and open sea, wind-induced circulation on
the continental shelves, dilution of fluvial waters, interactions
with topography, formation and cascading of dense water are some of
the various applications of our circulation models. The regions under
study notably include the Northwest Mediterranean, the Gulf of Thermaikos
in the Aegean Sea, the Bay of Biscay and the Gulf of Lions.
Sediment transport, Ecology:
The
horizontal circulation and vertical motion are key to understanding
transport and transformation of continental input and the cycle of
major constituents in the coastal zone. The transport of sediment
from rivers to the open sea is one of our research topics, for which
associated models are developed. We also collaborate with colleagues
from other groups to couple biogeochemical models with physics.
|