|
| |
DETERMINATION OF
CIRCULATION IN NORTH ATLANTIC
BY INVERSION OF ARGO
FLOAT DATA
Herlé MERCIER, Carole GRIT
Laboratoire de Physique des Océans
IFREMER - Centre de Brest
ARGO was
designed to provide unique observations of the large-scale and low frequency
variability of the ocean thermal and haline contents. However, that signal is
aliased by the mesoscale motions that are only partly resolved by the observing
system. A challenge is thus to retrieve the large scale and low-frequency signal
of interest, which requires filtering the mesoscale "noise". We present here a
methodology that aims at estimating the North Atlantic circulation and heat
budget variability at seasonal time scales using the ARGO data set.
The methodology is based on the finite difference inverse model developed at LPO
(Mercier et al. J. Phys. Oceanogr. 1993). The model looks for a best estimate of
the geostrophic circulation that minimizes the weighted sum of the squared
distances to observed temperature and salinity fields and the squared residuals
of dynamical constraints (mass, heat and potential vorticity conservations). The
weights are error covariance matrices.
In the first part of this study, the model dynamics was stationary and annual
mean circulations were estimated. In a second part, time-dependent dynamics was
used to resolve the seasonal cycle of the circulation. A model run based on high
quality CTD data from 1995-1998 provided a reference to which the circulation
estimated using the 2001, 2002 and 2003 ARGO data were compared. When
observations of temperature and salinity were not available, the model was
restored to the Reynaud et al. (1998)'s climatology. This occurred in particular
for depths greater than 2000 m when using ARGO data. The model domain extended
from 10°N to 70°N with a horizontal resolution of 1° in latitude and 1.5° in
longitude. The inverse model was forced using wind stresses observed from space
and air-sea heat fluxes estimated by numerical weather prevision models.
See also:
|
