J. of Geophys. Res. Vol. 194, NO. C1, Pages 1431-1453,
January 15, 1999
The Connectivity of Eddy Variability in the Caribbean Sea,
the Gulf of Mexico, and the Atlantic Ocean
Sylvia J. Murphy
National Center for Atmospheric Research, Boulder, Colorado
Harley E. Hurlburt
Naval Research Laboratory, Stennis Space Center, Mississippi
James J. O'Brien
Center for Ocean Atmospheric Prediction Studies, Florida State
University, Tallahassee
A set of numerical simulations is used to investigate the connectivity of
mesoscale variability in the Atlantic Ocean, the Caribbean, and the
Gulf of Mexico. The primitive equation models used for these simulations
have a free surface and realistic coastline geometry including a detailed
representation of the Lesser Antilles island arc. Two simulations have
1/4 degree resolution and include a 5.5-layer reduced gravity and a 6-layer
model with realistic bottom topography. Both are wind forced and include
the global thermohaline circulation. The third simulation is from a 1/2
degree linear wind-driven model. In the two nonlinear numerical
simulations, potential vorticity from decaying rings shed by the North
Brazil Current retroflection can be advected through the Lesser Antilles.
This potential vorticity acts as a finite amplititude perturbation for mixed
barotropic and interla mode baroclinic instabilities, which amplify
mesoscale features in the Caribbean. The eddies associated with the
Caribbean Current are primarily anticyclonic and transit a narrow corridor
across the Caribbean basin along an axis at 14N to 15N with an average
speed of 0.15 m/s. It takes an average of 10 months to transit from the
Lesser Antilles to the Yucatan Channel. Along the way, many of the eddies
intensify greatly. The amount of intensification depends substantially on
the strength of the Caribbean Current and is greatest during a multiyear
period when the current is anomalously strong owing to interannual
variation in the wind forcing. some Caribbean eddies squeeze through the
Yucatan Channel into the Gulf of Mexico, where they can influence the timing
of the Loop Current eddy-shedding events. There is a significant
correlation of 0.45 between the Loop Current eddy shedding and the eddies
near the Lesser Antilles with a time lag of 11 months. However, Caribbean
eddies show no statistically significant net influence on the mean
eddy shedding period nor on the size and strength of shed eddies in the Gulf of Mexico. Additionally, no significant correlation is found between eddy
shedding in the Gulf of Mexico and transport variations in the Florida
Straits, although transport fluctuations in the Florida Straits at 27N and
the Yucatan Channel showed a correlation of about 0.7 with a lag of 15 days.
The linear solution exhibited a multiyear anomaly in the strength of the
Caribbean circulation that was concentrated in the central and eastern
Caribbean due to a multiyear anomaly in the wind field over the basin.
In the nonlinear simulation this anomaly extended into the western Caribbean
and across the entire Gulf of Mexico. This westward extension resulted
from the nonlinearity and instability of the Caribbean Current, the
westward propagation of the eddies, and the passage of Caribbean eddies
through the Yucatan Channel into the Gulf of Mexico.
Back to Publications List
Hongjun Zhang:
zhangho@ucar.edu