Gulf of Mexico Circulation Modeling Study: Annual Progress, 7 D

OCS STUDY
MMS 85-0025
. i
~« ~,
~ .
GULF OF MEXICO
CIRCULATION MODELING STUDY
Annual Progress Repo

Year I
~o
,7~
~~ ~ D
D
AZ
U .S . DEPARTMENT OF THE INTERIOR/MINERALS MANAGEMENT SERVICE

OCS STUDY
MMS 85-0025
GULF OF MEXICO
CIRCULATION MODELING STUDY
ArT'UAL PRAGRESS REPORT
YEAR I
United States Department of the Interior
Minerals Management Service
Prepared under Contract 14-12-0001-30073
Alan J . Wallcraft, Ph .D
JAYCOR
November 1984
Prepared by the
Environmental Studies Section
Gulf of Mexico OCS Regional Office
Metairie, LA
DISCLAIMER
This report has been reviewed by the Minerals Management Service and approved
for publication . Approval does not signify that the contents necessarily
reflect the views and policies of the Bureaii, nor does mention of trade names
or commercial products constitute endorsement or recommendation for use .
Request Copies From : United States Department of the Interior

Gulf of Mexico OCS Regional Office
P .O . Box 7944
Metairie, Louisiana 70010
Attention : Public Information Unit

Telephone (504)838-0519
I
TABLE OF CONTENTS
I. INTRODUCTION 1
II . EXPERIMENT 0 9 3
III . EXPERIMENT 34 6
IV . EXPERIMENT 40 8
V. EXPERIMENT 60 12
VI . EXPERIMENT 68 18
VII . FIGURES 24
VIII . REFERENCES 102
II
I . INTRODUCTION
This final report gives details of the first year

of a four year numerical ocean circulation modeling program
for the Gulf of Mexico . The aim of the program is to pro-
gressively upgrade, in modest increments, an existing numer-
ical ocean circulation model of the Gulf so that the final
model has a horizontal resolution of about 10 km and verti-
cal resolution approaching 1 to 10 m in the mixed layer, 10
m at tae thermocline and 100 m in the deep water . Through-
out the four year period, the validity of the upgraded model
will be continuously tested, and velocity field time series
delivered periodically based on the most realistic simula-
tion of Gulf circulation available .
Experiments in the first year were with the exist-
ing NORDA/JAYCOR two layer hydrodynamic primitive equation
ocean circulation model of the Gulf on a 0 .2 degree grid .

They concentrated on correctly specifying the coastline and
bottom topography for maximum realism in circulation simula-
tion, and on how best to include wind forcing . Details of
selected experiments are presented in this report .

Experiment 9 represents the best (compared to our
incomplete knowledge of the real Gulf) simulation available
at the beginning of the project . It is forced by flow

through the Yucatan Straits only (no wind forcing), and
exhibits many of the flow features observed in the Gulf .
Simulated surface currents sampled every ten days for three
Loop Current eddy cycles (1140 days) were delivered to MMS

at the start of the contract period as an 'early simulation
run' . Experiment 34 is similar to Experiment 9, but with

the addition of seasonal wind forcing . The basic circula
tion patterns show far more variability in this case . Ex-
periment 40 has no wind forcing and its total inflow trans
port is identical to that in Experiment 9, but the distribu
tion of transport between the model's two layers has been
changed (upper layer transport reduced) . It exhibits Loop
Current eddies that are nearer to the size observed in the
Gulf (Experiment 9 has rather large eddies) . The increased
lower layer flow helps prevent intrusion onto shelf areas,
and its sea surface variability is remarkably similar to
that obtained from satellite altimeter crossovers for the
Gulf . Experiment 60 is identical to Experiment 40 except
that the horizontal eddy viscosity has been reduced . Some
of the flow features, seen in Experiment 9 were less obvious
,
in Experiment 40, but the latter's lower velocities allowed
the reduction in eddy viscosity and Experiment 60 exhibits
these features plus some new circulation patterns . Experi-
ment 68 is identical to Experiment 40 with the addition of
wind forcing from the Navy Corrected Geostrophic Wind data

set for the Gulf . This wind set has wind stresses every 12
hours from 1967 to 1982 . The addition of winds increases
the velocities encountered, and attempts to add this wind
forcing to Experiment 60 were not successful . Simulated
surface currents sampled every three days for more than 10
years were delivered to MMS from Experiment 68, as repre-
senting the best simulation available from the first year
effort .
2
II . EXPERIMENT 9
Since the ocean model contained many innovative

features it was discussed in detail in Hurlburt and Thompson
(1980) . In particular Section 2 (pp 1613-1614) gives the
model equations and Appendix B (pp 1647-1650) describes the
numerical design of the model . Since that time the capabil-
ity to handle general basin geometry has been added but this
does not affect the description in any major way . Wind
forcing is treated identically to interfacial and bottom
stress terms, i .e ., wind stress appears directly as an addi-
tive term in the momentum equation [see p 1614 of Hurlburt
and Thompson (1980)] .
In terms of 'realism' Experiment 9 was the most
successful Gulf of Mexico numerical simulation prior to the
start of this project . The model was driven from rest to

statistical equilibrium solely by a steady inflow through
the Yucatan Straits which was compensated by outflow through

the Florida Straits . The model parameters were :
o upper/lower layer inflow transport - 26/4 Sv,

o horizontal eddy viscosity, A - 300 m 2 /sec,
o Coriolis parameter at the southern boundary f - 5 x
10-5 sec-1,
o gravitational acceleration, g - 9 .8 m/sec2,
o reduced gravity, g' - .03(H1+H2)/H2 m/sec2,
o reference layer thicknesses, H1 - 200m and H2 -
3400m,
o minimum depth of bottom topography - 500m, grid
spacing, 25 by 25 Km,
3
o beta, df/dy - 2 x 10-11 m-1 sec-1,
o wind stress - 0,
o interfacial stress - 0,
o coefficient of quadratic bottom stress - .002 ; and
o time step - 1 hour .
Figure 1 compares 'instantaneous' upper ocean flow
patterns just before an eddy is shed by the Loop Current (a)

from the numerical model and (b) from observations by Leip-
._ . (1970) . The ability of the model to simulate observed
features is clearly demonstrated by this comparison, which
is remarkable given the simplicity of the model forcing .
However some discrepancies remain, for example the eddy has
not penetrated as far into the Gulf and is more intense than
that shown in the observations . Waves can be seen moving

around the wall of the Loop Current in both the model and
the observations, but in the model they are at the limit of
resolution and therefore unrealistically large . Moreover in
the Gulf the waves are more pronounced on the eastern wall
of the Loop and can form strong cold intrusions that may
contribute to the eddy shedding process (Vukovitch and Maul,
1984) . This is an example of a feature that would benefit
greatly from 10km model grid resolution . As shed eddies

propagate westward (Figure 2a) the model spontaneously de-
velops a counter rotating vortex pair (Figure 2b), a struc-
ture repeatedly observed in the Western Gulf (Figure 3) .
The roles of the wind and the Loop Current eddies in the
formation of this structure have been a matter of some con-
troversy (Merrell and Morrison, 1981) . Although wind forc-
ing was not present in this simulation a major role for
winds has not yet been ruled out . After spin up the
4
experiment sheds an eddy once every 390 days and the eddy
shedding cycles are very similar .
5
III . EXPERIMENT 34
Experiment 34 is similar to Experiment 9, but with

the addition of wind forcing based on a seasonal climatology
from ship observations (Elliot, 1979) . Linear interpolation
in time was used between the seasonal fields to produce the

wind stress at each time step . The model parameters were :

o horizontal ddy viscosity, A - 300 m2 /sec,
o Coriolis parameter at the southern boundary f - 4 .5
x 10-5 sec-1,
o gravitational acceleration, g- 9 .8 m/sec2,
o reduced gravity, g' - .03(H1+H2)/H2 m/sec 2 ,
o reference layer thicknesses, H1 - 200 m and H2 -
3300 m,
o minimum depth of bottom topography - 500m,
o grid spacing, 20 by 22 Km (0 .2 by 0 .2 degrees),
o beta, df/dy - 2 x 10-11 m-1 sec- 1,
o wind stress from seasonal climatology based on ship
observation s,
o time step - 1 .5 hours .
Seasonal climatological winds obviously cannot

represent most of the wind variability, and Experiment 31
which was driven solely by these winds attains a steady
yearly cycle (JAYCOR, 1983) . Even so the addition of wind
forcing increases the variability of the Loop Current
6
system, including the eddy shedding period and eddy path .
For example Figure 4 compares 360 model days from Experi-
ments 28 and 34 (which are identical except that Experiment
28 has no wind forcing) . From these snapshots, taken every
90 days, there is little difference between the two experi-
ments . But if Experiment 34 is sampled every 20 days, as in

Figure 5, it is apparent that eddies were shed in the space
of about one year . Figure 5 also shows that the circulation
pattern in the western Gulf can change very rapidly at
times . Figure 6 shows the mean interface deviation and its

variability for experiments with wind forcing only, port
forcing only and with wind plus port forcing . This demon-
strates that even in the mean the interaction of wind and
port forcing is not linear, i .e ., the mean of the dual forc-
ing experiment is not the sum of the other two means . The
variability is increased in the dual forcing case, particu-
larly in the central and western Gulf .

Figures 4, 5 and 6 also clearly demonstrate that
ocean circulation climatologies are inappropriate for use in
oil spill risk analysis in the Gulf of Mexico . Figures 4
and 5 show the highly dynamic nature of flow in the Gulf,
and that there is no obvious averaging period (since the
dominant Loop Current eddy shedding cycle can take anywhere
from 6 to 18 months) . Figure 6 shows that variability is as
strong as the mean signal, and that the mean is quite dif-
ferent from circulation at any given point In time .
7
IV . EXPERIMENT 40
Experiment 9 exhibits many of the circulation feal-
tures found in the Gulf, but discrepancies remain . Some

problems, such as the correct simulation of the waves on the
wall of the Loop Current, can only be solved by upgrades to

the model proposed for years 2, 3 and 4 of this project .
Others, such as the correct simulation of circulation on
shelf areas, cannot be completely s :,' .--3d without a break-
through in model design which is outside the original scope

of the project . However the major aim of the first year
effort was to investigate just how realistic the simulation

could be made without major changes to the model .
Loop Current intrusion onto the continental shelf

was identified as a major problem area . This is caused by
the fact that the model's bottom topography is confined to
the lower layer, so the minimum topography depth is taken to
be 500 m and there is a flat shelf in the model between the
500 m isobath and the coast (Figure 7) . Note that intrusion
of strong Loop Current related flows onto the continental
slope/shelf does occur in the Gulf, so the problem is how to
control such intrusions in the model and how to have confi~
dence in the results given the apparent deficiencies in
model formulation .
Coastal areas are so important for oil spill risk
analysis that the use of a layered circulation model might
be carefully examined . But the Loop Current and its associ-
ated eddies dominate the overall Gulf circulation, and have
a major impact on shelf circulation in both the eastern and
western Gulf . Given the state of the art in supercomputers
8
it is simply not practical to produce a circulation model of
the Gulf that has 10 Km (or even 25 Km) horizontal grid
resolution Gulf-wide to simulate the Loop Current system,
and high vertical resolution to give improved simulations
over the shelf . Therefore the choice was between a layered
model such as the one used here (possibly coupled to local
shelf models), or a level model not significantly better
Gulf-wide than the existing geostrophic climatology . In
year four of the project, the layered model will be coupled
to a one dimensional mixed layer model with high vertical
resolution . This will improve simulation accuracy over
shelf areas, but will not solve the intersection problem .
Further details on the question of model design can be found
in the original proposal (JAYCOR, 1983) .
Several ad hoc methods were tried to control the
flow over the shelf areas . For example the Yucatan coastp,
line was extended to cover most of the Campeche Bank in an

attempt to prevent early westward bending of the Loop CurF
rent, and interfacial friction was applied over shelf areas
only to try to control intrusion . But none of these atr
tempts were very successful, and the best solution came from
simulations addressing the fact that the Loop Current eddies
in Experiment 9 are large and have high maximum currents

(they are at the very edge of the acceptable range of eddy
sizes) .
Eddy radius is dependent on the upper layer velocir

ty at the core of the Loop Current (Hurlburt and Thompson,
1980) . It can be controlled by three parameters, (a) upper
layer rest thickness, (b) the density contrast between lay -
ers (i .e ., g'), and (c) inflow transport and its distribul-
tion between layers . Upper layer thickness and g' were
9
carefully chosen in the original experiments to give the
best representation possible of the Gulf in a two layer
hydrodynamic model . Data on the actual inflow transport
through the Yucatan Straits is not plentiful, but the figure
of 30 Sv for the total average transport is consistent with
what data is available . Data on the distribution of that
transport in the vertical is almost nonPexistent, indeed
even the direction of deep flow is not entirely certain .
Therefore the original distribution of 26 Sv in the upper
layer and 4 Sv in the lower layer was somewhat arbitrary,
and the upper layer transport can be lowered to pro : :!ce
smaller eddies . Exactly what range of eddy sizes is realis~
tic is hard to quantify, but there is one source of Gulf
wide data that can be used as a guide . Maps of sea surface
variability for the Gulf have been produced from all hydro,
graphic, STD and XBT data (Maul and Herman, 1984), and from
satellite altimeter cross,overs (Marsh, et al ., 1984) . The
20 Sv upper and 10 Sv lower layer distribution of inflow
transport in Experiment 40 gives rise to'a variability map
very similar to that obtained from the satellite (Figure 8),
these maps agree more closely with each other than with the
map from hydrographic data (Figure 9) . Based on the agree=
ment of variability maps, the mean sea surface from Experi~
ment 40 may well be the best mean available for the Gulf
(Figure 10) . The model parameters were :

2
o horizontal eddy viscosity, A - 300 m /sec,
x 10" 5 sec~1,
o gravitational acceleration, g - 9 .8 m/sec 2 ,
10
o reduced gravity, g' -' .03(H1 ;H2)/H2 m/sec 2 ,
3300 m,
o minimum depth of bottom topography - 500 m,
o beta, df/dy - 2 x 10i~ 11 mr1 secM1,
o wind stress - 0,
As a side effect of increased flow in the lower
layer intrusion onto the Florida Shelf has been reduced, as
can be seen from a comparison of free surface snapshots from
Experiments 28 (with 26/4 transport distribution) and 40

(Figure 11) . This is because lower layer flow tends to fol-
low the bottom topography contours (or more exactly f/H con~
tours), and the increased deep transport allowed the upper
layer flow to feel the continental slope more strongly .
11
V. EXPERIMENT 60
Horizontal eddy viscosity is used in the model to

parameterize sub-grid scale processes . As a general rule
eddy viscosity should be chosen as low as possible in high
resolution models, although if it is too low circulation
features can be produced that are not adequately resolved by
the model grid . These features will not necessarily be
simulated accurately . For example, the waves moving around
the Loop Current eddy in Experiment 9 fall into this cate ;,
gory . There is no substitute for high horizontal resolu-
tion .
Experiment 40 does not exhibit the smaller scale
features, such as the meanders on the wall of the Loop Cur ;-
rent and the countertrotating vortex pair as dramatically as
earlier experiments . However since this experiment has
lower maximum speeds the horizontal eddy viscosity can be
lowered . Experiment 60 therefore is identical to Experiment
40 except that the eddy viscosity is 100 rather than 300
m2/sec . The full model parameters were :

o horizontal eddy viscosity, A - 100 m2 /sec,
x 10 , 5 secrl,
o reduced gravity, g' - .03(H1+H2)/H2 m/sec2,
3300 m,
12
*,
o beta, df/dy - 2_x 10~11 m 1 sec~1,
o wind stress - 0,

Figures 12 to 19 show upper layer velocities from

Experiment 60 covering 300 days at irregular intervals .
Figures 12, 13 and 14 are 40 days apart and show a Loop
Current eddy moving into the western Gulf . In Figure 12
there is a pair of eddies in the northeast Gulf . In Figure
13 there are six or more small eddies in the northeastern
Gulf that have been spun off the main Loop Current eddy,
this may be the feature observed as a meandering current in
Brookes and Legeckis (1984) . Forty days later the currents

in this area have again changed and most of the flow is to
the east (Figure 14) . Figure 15 shows the situation one

hundred days later . The flow patterns in the entire western
half of the Gulf are extremely complex as the Loop Current
eddy dissipates on the coast of Mexico at about 24N . Note
also the current along the continental slope off Florida .
Ninety days later an eddy is about to break off from the
Loop Current, Figures 16 to 19 are ten days apart . Cyclonic
rings are moving around the wall of the Loop, and one api-
pears to pass right through the space between the Loop and
the shed eddy (Figure 19) . During this time the eddy is
intruding significantly onto the Campeche Bank .

On 19 November 1980, the NOAA Buoy Office (NDBO)
deployed three experimental TzD drifting buoys, at approxi=
mately 24 .5N and 92W, in an eddy that had just been shed by
13
the Loop Current (Kirwan et al ., 1984) . All three drifters
stayed in the eddy for at least five months as it propagated
westward to the Mexican coast, see Figure 20 . The buoys

were undrogued but had 200 m thermistor cables, which clearr
ly coupled the drifters to the deeper circulation . Figure
21 shows simulated drifter tracks for 160 days from Experio-
ment 60 . They start earlier in the eddy cycle than the NBDO
buoy tracks . The simulated drifter moves in response to the
upper layer velocity from the ocean model, which represents
the mean velocity above the thermocline . Along the drifter
tracks the upper layer thickness is between 250 and 350 m .
The simulated drifter position is calculated every 45 mink
utes based on the velocities at the four previous positions

using a Adams-Bashforth prediction method . Velocities are
linearly interpolated in space and time between the archived

model velocity fields that are available once every ten
model days . Interpolation in time is not an ideal way to
calculate velocities since the eddy moves west about 30 Km
in ten days . The slight elongation of the eddy's east-west
axis this causes is acceptable, but it also filters out any
short time scale velocity fluctuations . One advantage of
simulated drifters is that the ocean model data also gives a
view of the entire Gulf, Figures 22 and 23 show snapshots of

the model free surface deviation every 30 days from the
simulated deployment date (model day 1680) until model day
2010 .
The simulated drifters are in good general agreeo

ment with the actual drifter data . Both follow approximate-
ly the same path into the southwest Gulf . Experience with
the model suggest that this is the preferred eddy trajectory
although they can also track due west to arrive at the coast
14
further north . The observed average rotation period is
between 14 and 1 7 days, with a westward translation speed of
5 to 10 cm/sec and velocity component speeds of on the order
of 50 cm/sec (Kirwan et al ., 1984) . The simulated eddy has
a rotation period of 15 to 16 days, a westward translation
speed of 3 to 6 cm/sec and velocity component speeds of on
the order of 50 cm/sec . The lower model eddy translation
speeds may be due to the absence of wind forcing, or to the
model's idealized vertical structure . The addition of a
third layer and thermodynamics to the model in year three of
this project will improve the vertical density structure,
and may lead to different translation speeds .
The simulated drifters exhibit very regular loops
and appear to remain at approximately the same relative
position within the eddy for long periods . The actual
drifters on the other hand are much more variable, with
paths indicative of changes of the drifter location relative
to the ring center . These changes are probably primarily
due to windage effects on the drifters, and might be mini^
mized by adding a drogue at the end of the thermistor line
(i .e ., at 200 m) on future buoys . In principle windage
could be accounted for in drifter simulations ; model experi~-
ments with wind forcing automatically account for layer
averaged Ekman effects, but allowing for winds acting dip
rectly on the buoy would require knowledge of the the rela-
tive effectiveness of such forcing .
Figure 24 follows the path of simulated drifter
number 3 for 300 days . Remarkably it is still tracking the
eddy after all this time, even though the eddy has almost
totally dissipated by model day 1980 (see Figure 23) .
Drifter 1599 tracked the eddy remnants until mid ;-June 1981,
15
but drifters 1598 and 1600 left the eddy in midmApril and
earlyriMay respectively . The model eddy probably dissipated
too slowly, because its interaction with the continental

shelf cannot be modeled accurately . But equally it is not
necessarily the case that the actual eddy had entirely disr
appeared by midFJune 1981, it is possible that wind effects
caused drifter 1599 to leave the eddy at that time . In any
case the model accurately simulates the northward motion of
the eddy once it reaches the coast of Mexico . However the

pa*.' :s of 1598 and 1600 once they leave the eddy suggest that
the remnant of a previous Loop Current eddy that persists
off the Texas coast in the model simulation throughout this
time period was not present in the summer of 1981 . Similar
features occur in almost every model simulation, even in

simulations with wind forcing only, where it is a wind in-
duced gyre rather than a Loop Current eddy . Their presence
is explainable by the northward migration of antircyclonic
rings along the coast until they reach Texas, where the
continental slope turns east and they can go no further .
But in the Gulf the rings probably dissipate more quickly
against the continental shelf than they do in the model,
since the latter cannot include topography shallower than
500 m .
Figure 25 contains time series of velocity for

drifters 1598, 1599 and 1600 . The strong high frequency
contribution is unusual and is largely due to the 30 hour
basin tidal resonance and possibly inertial oscillations,
along with diurnal and semidiurnal tides and a 7-hour free
gravity mode (Kirwan et al ., 1984) . Figure 26 shows time
series of velocity for simulated drifters 3 and 4 . These
show no high frequency components, as is to be expected
16
given that the simulation only has access to new model velo ;
city fields once every ten days . High frequency components
might appear in simulated drifters that were calculated "on
the fly" within the ocean model, since new velocity fields
would then be available every 90 minutes . The low frequency
velocity components of the actual buoys agree well with the
simulated drifters . Both show periodicity associated with
the eddy circulation and velocities of about 50 cm/sec . The
simulated time series are far more regular, as is expected
from the com ;: ` .son of drifter tracks .
17
VI . EXPERIMENT 68
Simulations forced by winds based on a seasonal

climatology from ship observations have already been de-
scribed . Such wind fields are not ideal for driving ocean
models since they contain very little of the total wind
variability and mean wind strengths are in general far weak,-
er than instantaneous winds . Recognizing this deficiency
NORDA funded JAYCOR to produce a wind set for the Gulf
based on the Navy's twelve hourly surface pressure analysis,
which is available from 1967 to 1982 (Rhodes et al ., 1984) .
The geostrophic winds, corrected geostrophic winds, and wind
stresses (all on a one degree grid covering the Gulf) every
12 hours from 1967 to 1982 are on magnetic tape . These will
be made available through the MMS Gulf of Mexico regional
office .
Figure 27 shows the wind stress and wind stress
curl from this data set for 0000 and 1200 GMT on 14 January
and 0000 GMT on 15 January 1976 . There is large temporal
variability of the wind field during this period, as general

easterly flow gives way to strong northerly flow after a
frontal passage in just a 24 hour period . The wind stress
curl field also shows the rapid change, from a relatively
weak field to a very strong field with strong horizontal

gradients . Figure 28 shows similar plots for 14 and 15 July
1976 . Even in the summer, when flow is generally weaker,

very significant differences can be seen in a short time
period . These strong variations and very rapid changes in
the wind field indicate why the modeling of Gulf circulation
requires wind data on short temporal scales .
18
Figures 29 to 32 show the seasonal climatologies
averaged over the period 1967 ;1982 . The wind stress and
wind stress curls are much stronger in the winter season
than the summer season as would be expected . There are
persistent areas of positive curl over the Yucatan and negao
tive curl in the southwest Gulf that are present for all
seasons, but were not seen in any previous study of Gulf
wind stresses . Although not present at all time periods
(Figures 27 and 28), these are also the dominant features of
the instantaneous curl fields .
There have been no putlished accounts of driving an
ocean circulation model for long time periods with winds
sampled as frequently as those available in this wind set .

Previously a monthly climatology would have been considered
an exceptionally good data set for such an application . One
of the initial goals of this first year effort was to deter~

mine how best to use this data set to drive the ocean model .
The model simulates the layer averaged circulation above and
below the thermocline, and so only includes the longer term
effects of winds on ocean currents . But it is clearly
preferable for the ocean model to integrate the effect of
short term variability on these currents, rather than for

the winds to be averaged before input to the model . Before
testing began it was expected that inertial oscillations and
gravity waves generated by the highly variable forcing would
make using the 12 hourly winds directly impractical . So the
plan was to test the model with increasingly long wind aver~-
ages until these problems became manageable . In all cases
the wind stresses are linearly interpolated in time between
inputs, so the wind forcing is slightly different at each
time step .
19
Experiment 68 is identical to Experiment 40 except
for the addition of wind forcing after the port forced cir~-
culation has fully spun up . Wind input is every 12 hours,
at first 1967 winds were used repeatedly to spin up the
wind driven flow and then winds from 1967 to 1977 were ap-
plied in sequence . The expected difficulties with frequent
wind input did not arise, although attempts to add these
winds to Experiment 60 (with lower eddy viscosity) were
unsuccessful . The model parameters were :

o horizontal eddy viscosity, A - 300 m2/sec,
x 10~5 sec~l ,
o reduced gravity, g' - .03(H1+H2)/H2 m/sec ' ,
3300 m,
11 -1
o beta, df/dy - 2 x 10^ m . sec
o wind stress from 12 hourly Navy Corrected Geostroi-
phic Wind set,
Figures 33 to 39 show upper layer currents (i .e .,

vertically averaged currents above the thermocline) every 60
days for 360 days . Vectors are only drawn at every second
point (i .e ., every 0 .4 degrees) to improve readability .
20
Figure 34 shows the furthest northward penetration of the
Loop Current ever attained by the ocean model, this configu4
ration is often seen in the Gulf . After the eddy breaks off
the Loop Current intrudes onto the Florida Shelf and some of
the flow splits off to the north for a brief time (Figures
35 and 36) . Similar intrusions have been observed in the
Gulf, but the models inadequate representation of shelf
topography make it likely that the simulated currents in
shallow areas (say less than 100 m) are too high . A persis-
tent antir-cyclonic gyre in the north west Gul .f : :as been a
feature of almost all Gulf simulations performed to date .
The addition of wind forcing in Experiment 68 has increased
its average size and its effect on incoming Loop Current
eddies (Figures 37, 38 and 39) . The presence of a gyre in
this position is explainable by the northward migration of

antipcyclonic eddies along the coast of Mexico until the
continental slope bends eastward and they can go no further .
However In the Gulf the gyre probably dissipates relatively
rapidly against the shallow shelf area . The ocean model
cannot include a shallow shelf, but it may be possible for
it to parameterize the effect of such a region on strong
currents . Ad hoc patches to the model of this kind can only
be justified if (a) there is clearly a problem with the

existing model simulation and (b) there is sufficient obser-
vational data available to verify that the patch is indeed
"correcting" the simulation towards more realistic flow
patterns . Modifying the simulation towards the general

perception of what the circulation should be, without supi-- .
porting data, is dangerous . Our understanding of Gulf cirr~-
culation has changed radically in the last ten years, and
given the sparsity of the existing observational data base
21
it is likely to continue changing . In the case of the gyre
in the northwest Gulf condition (a) probably can be satisfii,
ed, but obtaining the data to satisfy condition (b) will be
difficult . MMS's planned observational program in the westft
ern Gulf will be of assistance in this area . Finally it
should be noted that problems of this kind can sometimes be
solved by less drastic measures . For example the Loop Cur^
rent Intrusions onto the Florida Shelf were eventually rer
duced by re-evaluating the inflow transport distribution .
Similarly the north western gyre was least obvious in Experr
iment 60 which has the lowest eddy viscosity . It is possi-
ble therefore that next years simulations with a 10 km grid
and correspondingly lower eddy viscosity will resolve this
question .
Simulated surface currents from Experiment 68 have

been delivered to MMS as representing the best simulation
data available to date . They consist of velocity component
(u and v) fields on a 0 .2 degree rectangular grid covering
the Gulf area, sampled every three days for 3780 model days
(10 .3 years) . Velocity plots taken once every 30 days (i .e .
plots of every tenth set of fields) for the entire ten year
period were also delivered . Figures 33 to 39 represent a
very small subset of these plots, note that a current vector

is only drawn at every second point (i .e ., on a 0 .4 degree
grid) to improve readability .
The velocities represent mean currents above the
thermocline and are therefore essentially independent of
local winds . To a good approximation they can be treated as
geostrophic surface currents . The Oil Spill Trajectory
Analysis (OSTA) model can therefore include the simulated
surface currents in exactly the same way as the previous
22
climatological geostrophic surface currents . Except that
now new current data is available more frequently, and the
OSTA model can include variation in surface currents (over a
ten year period) as well as variation in wind driving .
However the use of ten years of surface currents, rather
than climatology (which is effectively one year of data),
means that the OSTA model will require more computer re~-
sources to complete its statistical analysis of risk .
23
FIGURES 1-39
FIGURE 1 : (a) Instantaneous view of the interface deviation

in a two-layer simulation of the Gulf of Mexico
driven from rest to statistical equilibrium
solely by inflow through the Yucatan Straits
(Experiment 9) . The contor interval is 25 m,
with solid contours representing downward devia-
tions . (b) Depth of the 22 degree isothermal
surface, 4-18 August 1966 (Alaminos cruise 66-A-
11), from Leipper (19T0) . The contour interval
is 25 m .
24
a ) 1300 ----~
t_
L`. 1
'•_ ~---~
--
~~ftb
v '/~
F/ =i \
o ;• {
((... ")
( o / L 27 L
O ` 1 L H
H
% / JAYC OR COM MODEL

Arlm Ow~~ort
s m ooroou .
0 th
0 (wn) 11W
b) ' ' I '

300
25•
~l °N .,
., / ` Sr
20 • `, ~... .~r ,
o .n+ or a~c eonwr p+.rp
-
Nun
95• 90• a5° t~n
25
FIGURE 2 : (a) Interface deviation from the Gulf of Mexico
simulation at model day 1970 after an eddy has
separated from the Loop Current and propagated
westward . (b) Ninety days later the major anti-
cyclonic eddy at day 1970 has developed into a
counter-rotating vortex pair in the western Gulf .
The cyclonic vortex is to the north and the anti-
cyclonic to the south .
26
(a)
1300
r -~~.
NOW
.-- . ~- ~~
.~ ,:
, :' ~`.
.~
~
%
%
c`m') ~ (
.0~:
- , .001%
(fqc, !
0 ' (km ) 1750
0
1300
(b)
. ti
"0 --
.~-~~~~ •
zz
C `.•~
.~ .j
~-!•'
.
~;.,
~
~
(km )
I,,` e ; /
;`, \
-
; '~
~\ O
0 (km 1750
o )
27
FIGURE 3 : Counter-rotating vortex pair in the western Gulf
of Mexico as shown by the depth of the 15 degree
isotherm (in meters), observed in April 1978 .
The cyclonic vortex is to the north and the anti-
cyclonic to the south (from Merrell and Morrison,
1981) .
28
N
~
:;. . ., .;; .: . . .,
.... • G
i
ts
t \
~~ 4tb
FIGURE 4 : Instantaneous view of the inter face deviation
every 90 days, from day 90 of model year 9 to day
0 of model year 10, for Experiment 28 (left) and
Experiment 34 (right) . Experiment 34 is identi-A
cal to 28 except for the addition of wind forci-
ing . The contour interv al is 25 m .
30
. .
. - • . / ." \
;.. .. / /- 1 + \. ~
• .% \ • •
+ \ \ .
~ /
/
I
,
- `-'
\_/
/
C\~~O
'+ I
II
\
l~ _~~
. ~
\~I 1 1
,- ear 9 • - Year 9
' ~ Day 90
`/ #28 Da ~340
.9
.
--\ ~-- .
\
. .-\`~,
\
1
,
--
C O
•
• - Year 9 -
-
~
. ~ ., d . . ..,,.\,
`~
:.
`\ ~/ 1 \\+~~
Year 9
. .
.
Day 180 - ; Day 180
#2$ ' #34
-\
~i /
;~ ~,
. . _.
-- I
•
1
- - -
_ .
.2 . I - _ _
0
- ~ - I 1 \
1 ~ . ~i \\ O
+ O
Year 9 Year 9
% ` ` Day 270 Day 270
. : . . ~_= Y,123 #34
. `\
`
\\+. ~~ `==== .
;+ O ~ "~ ~
1 ~ ' I/ 1 +
+ \\ ~~ u' ~ `,``\\\\
•~/ ~~ --_-
Year 10 Year 10
--'" Day 0 Day 0
#28 #34
31
FIGURE 5 : Instantaneous view of the interface deviation
every 20 days, from day 260 of model year 9 to
day 0 of model year 10, for Experiment 34 .
32
~
. :.
. ~ .
. ~~'- • .
• -~- ~
''-
\ / \ \ ~ ///~~\ \ \ \
(`!-
/ / / ~ J \ \
/ / / / I \\ O ~_ . ~ ~ \
\ \ 1 \ i"
I1 ~ 1//j ZE +` 1-+~-i 1
II 11 \ ~`~ • , I\ \ \~-+\ ~=~ 1
\\ • •1111` / ~\\
1\\ i ;\\~~/
\\ ~ ~ \- ~+~ 11\
\ / . . . .. .
Year 9 Year 9
Day 260 Day 2$0
.: . ., . 'L34
. ". , . . #34
~ .
, , ..
. ~_
-~ \
/~ 1 1 , .-~ \\\\ 1 ' . . : .

~~- . 1~ \ \\
\\ ~
\ '/ ~\\
•\\~ 1
. +\~~- ~ I ~ / \\\ 1 . :.
\1 ~' ~ ~ /1 \1 ..r~'
O /
1\1 rr // O \\\ ~
/
1 /~ \
\ \
// 1 \ 11
~ -----
.
. \ ~ l . . . .. .
- r - .
\ . . Year 9 -~ Year 9
.~ / . /--
, Day 300 / •-, Day 320
-,'~'34 -~-
, ~ .
. . . . . . . . .
, ' .
. .. , .. . \\ += :1
.' . -+
~- ~•'
/i~'-+~_--~~+- \
\ 1
\ \
\ \ ~-` \ ~~`\\ \ ,, .
/ \\ \ •
\ /~\ ~ \
~\ 1 1
~ O /\ 1 \~\~-~:= f 1 / I~ /\ ~ n \1\I\\\~ - -
I O / II ~J ~ \-__-
I // 11 1
I O :~~ 1\
. . .--_--
i ' \
+ ;~ _ 1 \ \ + '
Year 9 Year 1(1

ray 340 Day 0
.' I - .--- ;~34 -,'~34
33
FIGURE 6 : Interface deviation mean and variability, for the
Gulf of Mexico from ocean model . (a) Experiment
31, wind forcing only ; (b) Experiment 28, port
forcing only ; (c) Experiment 34 wind plus port
forcing . The contour interval is 12 .5 m .
34
/ .
~-
\\`\~. .
ll ~
-- . . `~-/ _ . . . .- . . .
35
FIGURE 7 : Bottom topography and coastline geometry for Gulf
of Mexico model on 0 .2 degree grid . The contour
interval is 250 m and the shallowest depth is 500
m . The section of the Caribbean shown is treated
as land by the model, the position of the inflow
port is marked by the termination of contour
lines in the Yucatan Strait .
36
B0TT0M T0P0GRAPHY G. 0F MEXICO
DX .DY = 0.200.0 .200 (DEG) OBT = 250 .0 (M )
31N
-- ;-
~ - ~
,
r
.
(DEGI
~
~ ,
~ - .--~-
, ,
. .
. , .
--~--,-
. . . , . ,
'8N
1
98W (DEG) BOW
37
FIGURE 8 : Sea surface height variability for the Gulf of
Mexico . ( a) Based on about 16,000 GEOS-3 and
SEASAT cross overs, spanning nearly four years
(from Marsh, Cheney and McCarthy, 1984) . (b)
Based on an ocean model simulation' with port
forcing only ( Experiment 40), measured over three
eddy cycles at statistical equilibrium with the
free surface sampled every ten days for a total
of over 300,000 "observations" .
38
SEA SURFACE VARIABILITY FROM
GEOS-3 AND SEASAT CROSS OVERS (CM)
30
25
20
100°W 90°W 80°W
SEA SURFACE VARIABILITY (CM)

FROM NORDA MODEL
31N
( DEC )
18N
8W (DEG) , , 80W
39
FIGURE 9 : Sea surface height (a) variability and (b) mean,
for the Gulf of Mexico . Based on all available
hydrographic, STD and XBT data at over 16,000
stations, with substant lal filtering ( from Maul
and Herman, 1984) .
40
STANDARD DEVIATION OF MEAN DYNAMIC TOPOGRAPHY
30° N
25°
20•
95° W 90° 850 80 ° 75 °
MEAN DYNAMIC TOPOGRAPHY AT 25 km RESOLUTION
30° N
250
200
95° W 90° 85° 80° 75°
41
FIGURE 10 : Mean sea surface height for the Gulf of Mexico .
Based on an ocean model simulation with port
forcing only (Experiment 40), measured over
three eddy cycles at statistical equilibrium .
The contour interval is 5 cm .
42
MERN FREE SUR . DEV . G . 0F MEXICO 0. 40
DX .OY = 0 .2 0 .2 (DEG) OBT = 5 .0(CM)
31N
, ,
--,-- ;-
- -~--~• -I
i~~ ~ , =
,- ~
dOop
- I ~ %•
\ \ \ \ ~ ~ ' I
(DEG) /
r 0
i
~
.
--
,~
~-
-~•-•--
\\\1 0
, ,
--,--, ,-- - ;--, .

18N
-98E (DEG) -80E
43
FIGURE 11 : Instantaneous view of the free surface deviation
(a) from Experiment 28 (with 26 upper and 4 Sv
lower layer inflow transport), and (b) Experi-
ment 40 (with 20 upper and 10 Sv lower layer
inflow transport) . In similar phase of eddy
cycle, Experiment 40 shows less intrusion of the
Loop Current onto the Florida Shelf . The
contour interval is 10 cm .
44
FREE SURFACE DEV . 0 . aF M:XICa 0 . 28
OFtY = 1890 OH = 10 .0 ( CM)
31N
i
...
~-~
/ . :
\
I =_~
fD1EG1
- _•-- :-
i\ , .
/ / .
- ;
~ ,
~-- ;-
i8N
98N IDEGI 80W
M1N =-5 .06E 01 MRX : 6 .13E 01
FREE SURFACE DEV . G . 0F MEXICO 0 . 40

ORY = 1980 OH : 10 .0(CM)
31N
~
/1 ~~- ~
•
.
IDEG1 ~ ~-
of oo .
:
. -~--~-
18N
98N cDEG) eoW
MIN =-1 .75E 01 MAX = 4 .49E 01
45
FIGURE 12 : Instantaneous view of upper layer averaged velo-
cities from Experiment 60 on model day 2130,
velocities above 50 cm/sec are not shown .
46
GEOSTR . CURRENTS G . 0F MEXICO 0 . 60
, MODEL DAY = 2130
31N
:DEGI
18N
98W (DEG) 80W
VECTORS UP T® 0 .50 M/S PLOTTED ( MAX = 0 .90 M/S)
47
48
GEaSTR . CURRENTS G . 0F MEXICO 0 . 60
MODEL OAY = 2170
31N
(DEG)
18N
98W tDEG) 80W
VECTORS UP T0 0 .50 M/S PLOTTED ( MAX = 0 .91 M/S)
49
50
GEaSTR . CURRENTS G . 0F MEXICO 0, 60
MODEL DAY = 2210
31N
~CE=G)
18N
98W fDEG) 80W
VECTORS UP T0 0 .50 M/S PLOTTED ( MAX = 0 .88 M/S)
51
cities from Experiment 6G on model day 2320,
52
GE©S TR . CU RRE NTS G . 0F MEXICO 0 . 60
MODEL DAY = 2320
31N
I DEG
18N
98W (DEG) 80W
VECT©RS UP T0 0 .50 M/S PLOTTED (MAX = 1 .2 3 M/Sl
53
FTr,URE 16 : Instantaneous view of upper layer averaged velo=
54
GE©STR . CURRENTS G. O F MEXICO 0, 60
MODEL DAY = 2410
31N
DEG)
18N
98W (DEG) 80W
VECTORS UP TQ1 0 .50 M/S PL0TTE0 (MAX = 1 .06 M/S)
55
citiez :~ om Experiment 60 on model day 2420,
56
GEOSTR . CURRENTS G . 0F MEXICO 0 . 60
MODEL Df)Y _ 2 420
31N
)EG l
18N
98W (DEG) eLJW
VECTORS UP TO 0 .50 M/S PLOTTED (MAX = 1 .09 M/Sl
57
58
GEOSTR . CURRENTS G . 0F MEXICO 0, 60
MODEL DAY = 2430
31N
:Gl
8N
5BW (DEG) 80W
VECTORS UP TQl 0 .50 M/S PLOTTED (MAX = 0 .99 M/S )
59
FIGURE 19 : Instantaneous view of upper layer averaged vP] .o-
cities from Experiment 60 on model day i~L40,
60
GEOSTR . CURRENTS G . 0F MEX I C0 0 . 60
MODEL OAY = 2440
11N
:G I
8N
98W f DEG ) 8 0W
VECTORS UP T0 0 .50 M/S PLOTTED (MAX = 1 .01 M/Sl
61
FIGURE 20 : Paths of NDBO drifters (a) 1598, (b) 1599, and
(c) 1600 from November 20, 1980, through May 11,
1981 . The numbers 0 through 6 give the posi-
tions on November 20, December 20, January 20,
February 20, March 20, April 20, May 11, respec-
tively (from Kirwan et al . . 1984) .
62
~
.. ∎r
loe.a...
w
Y.waew ..av. .rw
. . ~
r• ..
~ .
~ !r •
. . . .'' !~'
.. !• ~~ .. . .. . . . . . ... . .'r. .
...iT' •~
CoM1N, .'
trra - C w lrn
t ' ...
PMt .
1MM\
w.,y,~y0 ~ 1
~.
~
'
/~
!o - H•
qHLCO
POa
~
,
i
'..
7-/ : :
t .• 1• ~~ 1 ! ~+L . .t 1.•
.• ~ Y
~• 0• • ~
, , ,4 •.
..rme
tt•
: r.fe.
tw ~ ...~ .
o o
. . . . ''
.
~
^ .. se• s• u• •o• . .• . :• .o•
a
b
~
MlRsTm
F Z 0'
COA•U~ , _ _
f.a~~n I
POIR .
• . .a • 0
.. ; \
MliO ; I
to
~~ j_~2i
..~ •• .
. -~-- F :t•
~'~- . . . .,.~
I
°•~ a .~
I
I . . .~~ ~
.a• .r
63
FIGURE 21 : Paths of simulated drifters (a) 3, and (b) 4
from model day 1680 to model day 1840 of Gulf
model Experiment 60 (which is forced solely by
inflow through the Yucatan Straits) . The tracks
start earlier in the eddy cycle than those in
Figure 20 . The simulated drifter moves in re-
sponse to the upper layer velocity from the
ocean model, which represents the mean velocity
above the thermocline . Along the drifter tracks
the upper layer thickness is between 250 and 350
m . The track is drawn as a solid line for 20
days, then dashed for 20 days, and so on . There
is a dot every 5 days .
64
DR I FTER TRAJECTORY G. eF MEX IC0 60
DRIFTER N0 . 3 FROM 1680 TQ 18+0 (ORYS )
30N
-----
i0EG1
~ -- - .
.
. •,, ,.,
. ;
.
.J-
1 :
\ •\
20N
, .
98W (OEGI 85W
DR I FTER TRAJECTORY G. 0F MEXICO 60

DRIFTER N0 . 4 FROM 1680 T0 1840 (ORYS )
30N
.
. .' . ' I I
.--
. -~-
(OEGI
~ -.
•.-- «-
. .
20N
98W (DEG) 85W
65
every 30 days, from model day 1680 to model day
1830, for model Experiment number 60 . The con-
tour interval is 10 cm, and solid contours re=
present upward deViations with respect to the
sea surface height at rest .
66
FRE E SUqFaCE OEV . G . 0f rEx1C0 0 . 60 FREE SURFPCE U_V .
G . Of MExiCO 0. 60
DAY = 1680 OM = 10 .0(CMl CRr : 1770 Cti = 10 .31CM )
31N 31N ,
'
10EG1 (OfG) ~___ / ( I f \
~---
1 1 . . 1
. _~ . .
18N 9 , 18N
6w ( 0EG ) , sow 98W ( 0EG ) ' 80W
DAY = 1710 OM = 10 .0(CM) DAY = 1800 oM - 10 .01Cm l

3IN 31N
\ i-
\ ~i ~ / f \ \ \ ~ ~•
(0fG1 ~ ICIEGI
~
/ 1 O
/, . : .
., . / 1 .~ . /~.~ .~ .
. . ~ / . ,
l8y 98N (0EG
) 80a 1~ 98W (0.G) 801a
DAY = 1740 OM = 10 .0 ( CM) DAY = 1830 oy = IO .OfCMI
31N 31N
~ .
-~• ~- . .:. '

--\\ ~1`----~~ ; `
i ~ \\ ~ O I
\
O
I'-'EGI • ~ ( DE G1
20
._~_1\ Oo // \I O
~; / .. . . .. . .
18N 98W (0EG1 801a 18N S9u ,' (OE GI 80N
67 .
every 30 days, from model day 1860 to model day
2010 (Experiment 60) .
68
FREE SURFACE DEV• c . oF MExlco 0 . s o FREE SURFACE OEV . G. OF MExICO 0. 60
DAY : Ie60 01•I : 10 .0(CM! OpY = 1950 OM = 10 .0cCM)
31N 3IN
i
.:. _ .:. - ..
• i1 -`-
` `-
\ \
\ . ~ ...
.
Z GI t~~ IDEGI
(\\ ~ ( i~ ~\
~
. -~
.~ \ O
~ r ~ .. .
`\ . .
\ , .
r• . .
I8N 18N
98N IDEGI BOW 98u IOEG1 80Lt
DAY - 1890 ON = 10 .01CM) DAY c 1980 OH = 10 .0(CM )
3! N 31N
- , ' ~-
-
~ \ \ \ .~
.
\\
., 1
i :. Gl
l \\I~~ r / `
IOEGi
` ~~
~- ..
rl. . . . . . . .
.
~~ . . . I
I8N IeN
~ law IoECa eow 98u IGEGI 80W
DAY = 1920 0M : 10 .01CM1 DAY = 2010 DM : I0 .0ICM1
31N 3114 . . , . . , , , r-
.
. . '. . . . .
. .
O 1
VVVI
\ ~
~ ~ \ ~ Q \
\\~ 1 \ ( , \ \
10E61 ~~` - 10EG1
.',- ~ i L . . . . : .
. .. ~ • ~ ~ -•-
`
_ . •__ . ; .. . .
18N ~ 18N ~
98w (DEG ) lOEGI >3CW
69
FIGURE 24 : Path of simulated drifter r : .m:ber 3 from model
day 1680 to model day 1980 (EXperiment 60) .
70
DR I FTER TRAJECTORY G . OF MEXICO 60
DRIFTER NO . 3 FROM 1680 T0 1980 (DAYS)
31N
. .
- r -
( DEG)
, -.
.~ .
.
. •. --
. ~
---~--«--
f f
.
~ . f
~ , .
- ;-- , ,-- ---,---,-
1
18N
98W (DEG) 85W
71
FIGURE 25 : Time series of velocity for drifters (a) 159 .'j
( b) 1599, and (c) 1600 . From Kirwan et al .,
(19sa) .
72
1or 1a
s o sc
~ .
~.
E _N E
Y
V
~
~
W
W si ~ Oc
d 0)
q
c
•s -s 4
•100L ...., . .,
-1ee
20 40 s0 s0 100 120 20 40 •0 s0 100 120
TIME (DAYS) TIME (DAYS)
100
so
0
..
E -so
V
W s0
0.
N
0
-s0
-/00
!0 40 so •0 100
TIME (DAYS)
73
FIGURE 26 : Time series of velocity for simulated drifters
(a) 3 and (b) 4 from model day 1680 to model day
1880 (Experiment 60) .
74
VELOCITY VS TIME G . OF ME X ICO 60
PRIFTER NO . 3 FROM 1880 TO 1880 (DAYS)
100
cn
~
~
c. .~
~
-100 L
0 V COMPONENT (DAYS)
100
~
cn
~
S
c.)
v
-100 ,-
O U COMPONENT (DAYS) 200
VELOCITY VS TIME G . OF MEXICO GO

DRIFTER NO . 4 FROM 1680 TO 1880 (DAYS)
100 r- 7
N
~
~
V
v
-100 L
0 V COMPONENT (DAYS)
100
.-.
f/7
~
U
v
- 100 L.
0 U COMPONENT (DAYS) ~ -w
75
FIGURE 27 : Instantaneous wind stress and wind stress curl
from the Navy Corrected Geostrophic Wind data
set, for 0000 and 1200 GMT on 14 January and
0000 GMT on 15 January 1976 .
76
LL w-~t•r . us sw .
w-aa~r . •rs .r~
LO-349 a- Xm 90-594't- - PIIR .1W/SiIxAQ ZZ'C - SS3KLS aMZ1 t1tUa7fr}I
309- (03Q) 196- I09- (O8Q)
r r x9t
/ .• ~ ~. ~. ~ . . . /
. '
~~ -
O
Ota)
~
/ 00
( l i
RTC xtc
GO-1o -z - Ja iH9 0 .Lr 9L6t/4to •'i7- dA'J 0 ir 9t.6t/4t0
"TZi11D SSmS QNIbI SSZ2i.LS QNIIII
..-3n-" Lec ..r. ,.-». -c . LU .d..
90-Z[t't - XYR 90-8£0't- - xq1 s14~/S9lLla Zf C - SS=fS axtl Rt1Rn(911
209- ('Jia) z96- 209- (Jia) 3GS-![Ot
x9t
c
1
\
(93Q) 'J3
y . . . i w .. .• ~ .. .. ~. ~ I' loo, /
. . ,. . . i i .• I /
oo'
. r . . . . 1 I~ /
. . . ~ / ~~
RTC RTC
L0-Z0•z - Ja 1JlJ zt Sr 9L6t/rt0 •it" 1J[9 zt Lr 9bst/fto
axn0 ssaILcs axleL ssZ2i•LS aNIAII
"-JI7h1, fii . .0
1M-3mi6 i8 VM
LO-ISO'6 - XrR L0-8zZ'4- - xIA fRa/S iNm Of't - SSuts aKa Rnp=R
209- • ('JI(l ) 396- 109- ('JY(1) z 9s-
tmt xot
/~a•,~~•\ ~ ` . . . v . . .
t
(via) (oi
RTC Ntt
L0-30'Z - JQ iR0 0 t1' 9LEt/ft0 •T SRJ 0 S9 9k6t/ft0
Tdt13 SSm .LS QN[dl ssa2i.IS QNldll
FIGURr ?8 : Instantaneous wind stress and wind stress curl
from the Navy Corrected Geostrophic Wind data
set, for 0000 and 1200 GMT on 14 July and 0000
GMT on 15 July 1976 .
78
WIND STRESS WIND STRESS CURL
196/1976 AT 0 G1fT J3 . 196/1976 AT 0 GUT DC - 2 .0E-07
IN 31N
. f / r . . . ~
t t . . . . . -, -. ~ ~ ( ^J ( \~ _
1 l \ .. . . . . . . . . ~ -. ~
_ ~----~\ -\ -
\`\\
\ J~ /\ ~ Q \\ `\\ \ ` .
(DEG)
~ \~-
.\
DN +- 18N
-98E (DEG) -80E -98E ' (DEG) -80E
IIA33aNN WIND STRESS - 2 .51 DYNES/CY' m w I MIN ~ -1 .25E-06 HAX , 1 .30E-06
rIY f 7-KC-M MY. 777 1}KC-N
wIND STRESS WIND STRESS CURL

196/ 1976 AT 12 GUT '~- 196/1976 AT 12 GUT DC - 2 .0E-07
1N 31N
t 1 1 . . . . . ~ ..
l 1 .. _ . . ' . . . .. .. -r _ .- --- ~
_~ I>
~- -~
-)1\\\\\; ~.
, ~ ~, . . . . .
__~. ~\ .. IN
: G) (DEG) C
Os
. ~ \ \ ~ ~irr ~ ~r
,,,\
I5N +-
-98F (DEG) -80E -98E (DEG) -8oE
IlA)QYUIi WIND STRESS - 3.16 DYNES/CYt NIN - -2 .27E-06 MAX - 2 77E-06
Mq .77f 11-KC-M MIIY.IT If-KC-Y
'WIND STRESS WIND STRESS CURL

197/1976 AT 0 GNT 1. 197/1976 AT 0 GUT DC - 2 .0E-07
IN 31N
\ \ 1 . + . ~ ~. 11 ~.
\ \ • ~ . . - .. .. ~. ~ ~.
.
\ ~-
ti! ~ \ \^ ' 1 1+ ~ \-- ~ \\ \
~ •~ . . . , . :
:-,~ ~\
(DEG)
1 J
LI )(>
." ~ \\
. :
ci; ~_ . j _ P ;
\ ..-----
1_
8N J-- ,_
-98E (DEG) -80E 1 -08E (DEG) -80E
HA70YUY WIND STRESS ~ 2 .23 DYNES/G/=..2. MIN - -1 . 3 7E-08 YAX - 2 .33E-06 "
1 t1-KC-M M/Y m iFKC-L
79
FIGURE 29 : Climatologi,sl (1967R1982) wind stress and wind
stress curl from-the *Navy Corrected Geostrophic
Wind data set, for winter (December, January,
February) .
80
WIND STRESS
W II1TEB 1967-1982 ~...
31N
• r r r r r J, -,w r t-NeeeN, • 9
r ~. r r r r`-~ .~ r r•- ~- ~- .~ i
o
~' r/00 le Ie 1 A/ A/ le r r r r r d Ir
-ff"
./ .l I0, s-1 ./ a A"' .e v .- r r .• t .r
1 Il Il I -1 1 ./ . -- s-- r r . ., r r r .r
(DEG) •-' ./ .oo' . oo, .eo, ./r 00 1a' .ir r r r r v .
~ f l . 1*1 . oe . -01 s-le 000, ./ •~' •~ •~ r r r r r-
~-~
'000- l'
././ .,~ . loe p2i!;

, 00~
„/,/ . f -''r'~'
//,~ r
~- 000
0.. r/ '
-08E (DEG) -80E

MAMDcvY T rrD sTDESS - 2.79 DnKES/ck*
W" os ~s-.woc-.~
WIND STRESS CURL
W~ ITSB 1067-108Z DC - 1 .0E-07 M
3114 ~- ~
~
/ 1
~ ~1 I
1
\I
I
(DEG) 1 /
~~------- ( -
_0 ~-
c~
///l)~ ~
7
1 ` J \ ` \ _
/
1W\
~
..r
~ r 1 1
q
- 95E (DbG) -80E
YIId - -1 .42E-08 MAX - 3.15E-07
NauoA st! t t-occ-"
81
FIGURE 30 : Climatological (1967-1982) wind stress and wind
stress curl from the Navy Corrected Geostrophic
Wind data set, for spring (March, April, May) .
82
WIND STRESS
9PRIIJG 1967-1982 - .
31N
• • • A A
~~•~ w w w w w R a a a • • •
~ `~ ~. w w w w w w w w w w .6
(DSG)
18N L-
-98E (DEG) -80E
1LAXnlUY WII1D S'TRE3S - 1 .33 DYNES/CHI
WMA us 1 s-ncc-.a
WIND STRESS CURL

SPRIING 1967-1982 DC = 1 .OE-07 ]QM
31N
.01
(DEG)
~ `~1 \ ` r ~' \ \ r
~
~ `~~~\ \ ! ,~ \
\~ill~~~,^
1 I \
18N
-98E (DEG) -80E
l@T = -7 .86E-07 MAX - b.85E-07
WWA Mm 1 =-vcc-4 a
83
stress curl from'the Navy Corrected Geostrophic
Wind data set, for summer (June, July, August) .
84
W1N 1) S'1'KESS
8U]O= 1967-1982 _= .
31N
A
1 1 • A p ~ p
~ R R . • . . . . ~ ~ ~ . . . .
~ ~ ~ ~ •~ .~ ..~ .~ . .~ w. w w •. .. .. • ..
(DEG)
V%s,
18N L--
-98E (DEG) -80E
MAX LUM TM BTRESS - 1.29 DYNES/W
NOW ais 1 Z -occ-"
WIND STRESS CURL
9UNLER 1967-1982 DC - 1 .OE-07 l K9
31N
i
~
~---
%J
(
~
(DEG)
~
~ ~---
\ ~j
d~ **-
. ,
\
18N '--
-98E (DEG) -80E
l@i - -8 .84E-07 MAX - 4 .31E-07
MoMA sts t 2-occ-"
85
stress curl from the Navy Corrected Geostrophic
Wind data set, for fall (September, October,
November) .
86
WIND STRESS
F1LL 1967-1982 -L-S .-O
31N
• ~ • r r r r r r r Ar,,,I w ~' 1
f ~- •- r r .r r r r r• r .~ ~
f+~ r r r r .~- .r r r r .- r r r r
.r r r r r r r r r
r- .*, ,, *0 .i .i a a a .oll r r r r r r
r .-10, .i 0/ &/ r f,,r .--I r r r r r r r

(DEG)
,,/ j*/
r r~ . 10,
/
0,00, a
'
0
/
fi .-e fir
~ .~ .... . ..~ .r~
r rr r ~
J~e
18N L--
-98E (DEG) -80E
1Wm[VI[ VIIdD 8TRE39 - 1.84 DYNE.9/Cl[B
NOo A it! 11--0[C-M
WIND STRESS CURL

FALL 1967-1982 DC - 1 .0E-07 MK3
31N
/
~( 1
/T
\ ~~ 1
J
`
(DEG)
+ J
1
b
~
C::
. O~
-
\\
r ~ M \ ~-
18N 1.. ...~~
~
I ' I t
-98E (DEG) -80E
lQN - -9.53E-07 MAX = 3 .22E-07
MOIOA in 12-KC-M
87
cities from Experiment 68 on model day 3858 .
Vectors are only plotted at every second model
grid point, i .e ., every 0 .4 degrees .
88
GEOSTR . CURRENTS c . 0F MEXICO o, 68
MODEL OAY = 3858 WIND DAY = 1967/239
31N
<L, >~ J y 1 -
1 1 vyq9)1. ai .1»sa 1
---~-- I - t~fr 7 • .0 +.s~~ia~>>s~ -~ -- -
1 9 iras l 4r!- V` c f. q I JI w 7 - s L 4 • aa~.sy 36 1'~ 1
. .? .•~R'k !lV~vr aa~7^f> Vv
-- ~
c V J - ~ -- 1- -
e! 7 -I V. r 4
a 1 TV 3 .a .i a a~ -.%- Y .4
~r ti~ ~/M a V VY A d 1,1y> Y V Y
.a .4 .A q v ~ V V ~ 1
. . /
.` ~s• • Y ~
t jr --o
` '
4 >• ,r „
)EG l it
• \ 'k \
IV
/ ^s
' ~ .-
r •r a s • .•1
A '4 ft VAI F~
~~R ~ -~~ r t e ( ~ R t R 1 1.
~ftesR ~S ~ ..VL
1~ R A• t•~ A ~ 1
r~fR %I
A
II
rr ~ ~ 40
• !' A j%
< 1 A \/
-f ~R R
A `1- v 1
v *.e t % ,. ,,
/Rw*R %, < lk .~ .. 1 1
-~-
1 1 R t P ~ 1 1
1 1 1 1 1 1
18N '-
98W tDEG) 8oW
MAX PLOTED VECTOR = 1 .46 (M/SEC)
89
FIGURE 34 : Instantaneous view of upper layer averaged veloi-
90
GEOSTR . CURRENTS G . 0F MEXICO 0, 68
MODEL DAY = 3918 WIND DAY = 1967/299
UN
1 1 1 , ~ A f ' 1
LA ! *' Q V ~
~ 1 1
wrr A eRt ~ cs .r ~
--
~ ~
..rl r~.~~
.~ ~r~v ~~ ~-- ;-
4
R~l y~ -Ir
*-
f
:Gl
•
.. v c ~ ~
d.l'~"r-~-
Iii'"L
R 'l ~ .
4 4 ~
•
t A pl. ~
i .1w I~ r ~ O
- R s ~ ~as . !t --l--
~ ~~ !aA, y . .~l~ ~ i
.e . .• . v 0 s 1~.
~ - .r 0rs % `r~ ~ ~
--'-- sr~, ~ -- --~- -'-

i
~ ~ 40
8N
98W (OEG) 80W
91
92
GEOSTR . CURRENTS o.
0F MEXICO o . 68
MODEL DRY = 3978 WIND DAY = 196?/359
31N
> ~ 1
1 1 1 1 , y<< ~'~ Y ,
1 ' / 1 V ib « f f<<<< r / '

1 1
s<e r r• P 1~ A A - --
, 1
V r < t<< d f f l~ t i 4•r 4! • -b a• V ~ f
1 ,r r a ti <s ~iAtr'~-R! p~„~~R .i rR R• A I A A 1
A . --/-
~r'! r .F R f. < A 1 1
.t~ 4 a< A +~ t f ! I• `
~~` +~ r* A 4 A
low • ! fRl. R .I Il h (• p / - - T -
A 1
1 r . • . • • 46
DEG )
~
j
~ l',.
f•
/ T- 1 A
~sy
4f 1
~\ \~~~~
y A
~
e%L ! V !~c
a w T, t~
*~.~
~~j
1l R
I , Rr R. ~~
/
~ < s
je
f< t• i . s q A A R~ ~, '
V r i ~
Y v ~ a s 7 I A f'~' • ~
11 ~qy rrt * A, .v~s a O
~ --1__!
, T A ; 1
11 1
TI;.
'I,
, ! `! < ~ 1 1 1~
-/-
1 1 ~l1 1 1
I t 1 / 1 1
18N . 1 1 1 ..1 1 1 ~ 1 L
I -
98W . (DEG) 80W

93
FIGURE 36 : Instantaneous view of upper layer averaged velor
94
GEOSTR . CURRENTS G. OF MEXICO o . 6s
31N
/ / 1 1 1 > > 1 ' /
1
1 / t 11
t t! A 7 > r 1
1 1 ~ r L r A A < t ~>>» 1
/ a V t . 4- f +.w -1.-~- r c v ~ 1
-1-
rvrr .~ r r~~~f
< < < (. n 1
! Lr ir
A .4 s c y J V ~ a~ ~-- T-
- Y v
~ ~<'~1~~1~J' ~ ~~c 't 4~
6% A A <aaac~^ , 1
~ `e ~~~~ 7> v t ~
0-0 2L
~I r
4 AM
4
; DEG l
t, .
-4 ~ ~.~
A *R'\~~~r~~~'` !f ! i ~' d J r ! y -& 1
> a . s 1 1 -
*r r+ra .~ ~f ~
. 11 A 1
a o. n~ x
/aw
..a w aaa JL . 6 a a a '1 c ~
7/ Jr~ry~ ~ A& 'f 1 1

'P
< 1A
1 `~~~ Y~ ~ V> 7 ~ 1 1
1 1 ~~t `•
1 1 1 / t 1
18N
98W (DEG) 80W
95
96
GE O STR . CURRENTS G . 0F MEXICO 0 . 68
31N
, 1 1 1 ,
~
. ~/ i 1 1 U.~ 1
~ t L( K j K K< . f R R . L L Y d,t q> t J 1

r y L I. • t t< t i 4 s. R R f< J>'I IY ~ , ~
4
' ~ , -~-
a sl .• sI- -C as jor >>>> 7 • 4
._
T
v
~~ ~ a ~1 a a >
Ir•t•.sYV ~ 0
~KS-~l~l~l\ A>aa a,4 r
14 4 t d
I l ~ ~
V y Y i1 ~ V •~ y d,~
r + 7 , /
J
~,a v 41 -r-
~ w
`-
iEG )
1 so%
(1 /~ t •
a~ s
iii1w
V> 4 40 ~C R
< ? . y ~' K 1 & ^ A
R
*y
tRt4 r
~ A / 11 V a f -W
a . s s A rv ~ ~ P. ~ ~
*r ~ Y a•~~>s i a 4 •
~ 1 lk
, R1ti ~t 't~ff K ~~o
i
-~-
18N
98W ( DEG) BOW
MAX PLOTED VECTOR = 1 .04 ( M/SEC)
97
98
GEOS TR . CURRE NTS G . 0F MEXICO 0 . 68
M©DEL DAY = 4158 WIND DAY = 1968/174
31N
Ir
1 1
..a 7 > a ~ ~.~ It
A bV
---~---- ~
~ c 7 1~AAs i i sa>saY -~--
~ sa 1~ s~ ~ s> s-s> 4r ~ •~lR s ra a j. .> > .t ,
~
i
vas>VkR ~ .~ ,aaaivY -~ -
V Y i I1 1l p c ~1 ~ r V Y ~ -r
Y y
T-
f„"y~ii~~~i >>r•r~~ v a 'VV .YY ~ i
yy .> ~ r . .~ ~a v y v v v . ~
~ T .LI~ ~[ y A~1~~4 7 t~ ~~ v v V~1 Y A - T-
ta~ 111~~~ I~R R

~' fi r y~ ~ y* 1 ~ A~ \i i f f J .A V ~
~•i+~Y• RR>y~ `rvaar

DEGI Y 1'~'f
~
,+
14 4
s1 T"'Rttt~ ~~1 j,l / f f~> 14
~ ar e f fa Ya 0
A rq i--~
t~ to ~tst~-r~r s t• 4
~`~ t i b Y i 1/ y I 1
~~vA .4 •is 4, f>Y

~ ~ V q .• . .a va .c
--~-- R . rr vr XA- -- --L--~-
~ ~ « .
.s~i
. .
18N
98W (OEG) 80W
MAX PLOTED VECTOR = 1 .00 (M/SEC )
99
FIGURE 39 : Instantaneous view of upper layer averaged veloi
100
GEOSTR . CURRENTS G . OF MEXICO 0 . 68
31 N ~--r~
I
s• a,, A 4 f< ~ 1
--~-- ; - I L 7 A - ; - -
~ay~f e~~rr~- A 1~ Jr~Jt~~, ~<< V ~ ~ 3L
. a ~K 1f A s r~Gi r ra •V ~-
--' , ~ - ~ - ;-
f . .r.-rr,-~ \14 ~ R* v~ v i•
--Op7 Y w + 4 4 r + ' - - T -
a w ~ahr~ .4 vt .4
•4+4A r~~ ~ V ~'a > -t4 vice V Y y~v ~
k R'k s sf 1~~ i<r cv r! Yr T
.
/14 Y ~.~~.~ „ 40 . . 1% y - .
DEG) , ! ' ` S ! ~ •
~.,. +. . er
.a-r • 4 ~. ~. a ~ q
-4,w
~ ti-~-
N
~
- ; ,,
--=--! --'
, so'
. A
* etaa -- --~--'-
i
18N L1 1 1
L98W (DEG) 80W
101
REFERENCES :
Brooks, D .A . and Legeckis, R .V . 1982 . A ship and satellite

view of hydrographic features in the western Gulf
of Mexico . J . Geophys . Res . 87 : 4195#-4206 .
Elliott, B .A . 1979 . Anticyclonic rings and the energetics
of the circulation in the Gulf of Mexico . Ph .D .
Dissertation, Texas A&M Univ ., 188 pp .

Hurlbur-L,, H .E . and Thompson, J .D . 1980 . A numerical study
of Loop Current intrusions and eddy shedding . J .
Phys . Oceanogr . 10 : 1611F1651 .
JAYCOR 1983 . A proposal for a Gulf of Mexico circulation
modeling study . JAYCOR Proposal Number 8206m83 .
Kirwan, A .D ., Merrell W .J ., Lewis, J .K . and Whitaker, R .E .
1984 . Lagrangian observations of an aniticyclonic
ring in the western Gulf of Mexico . J . Geophys .
Res . 89 : 3417r3424 .
Kirwan, A .D ., Merrell W .J ., Lewis, J .K ., Whitaker, R .E . and
Legeckis, R . 1984 . A model for the analysis of
drifter data with an application to a warm core
ring in the Gulf of Mexico . J . Geophys . Res . 89 :
3425-3438•
Leipper, D .F . 1970 . A sequence of current patterns in the
Gulf of Mexico . J . Geosphys . Res . 75 : 637 - 657 .
Marsh J .G ., Cheney R .E ., McCarthy, J .J . and Martin T .V .
1984 . Regional mean sea surfaces based on GEOS ^ 3
and SEASAT altimeter data . Marine Geodesy 8 : 385^
402 .
Maul, G .A . and Herman, A . 1984 . Mean dynamic topography of
the Gulf of Mexico with application to satellite
altimetry . Marine Geodesy (to appear) .
102
Merrell, W .J . and Morrison, J .M . 1981 . On the circulation
of the western Gulf of Mexico, with observations
from April 1978 . J . Geophys . Res . 86 : 4181h4185 .
Rhodes, R .C ., Thompson, J .D . and Wallcraft A .J . 1984 . The
Navy Corrected Geostrophic Wind data set for the
Gulf of Mexico . NORDA tech . rep . (to appear) .
Vukovitch F .M . and Maul G .A . 1984 . Cyclonic eddies in the
eastern Gulf of Mexico . J . Phys . Oceanogr . (in
press) .
103
APPENDIX A
MODEL PARAMETERS
104
REFERENCE PARAMETERS (EXPERIMENT 40) :
• upper layer inflow transport - 20 x 10` m' sect'

(20 Sverdrup),
• lower layer inflow transport - 10 x 10s m' sect'
(10 Sverdrup),
• wind stress - 0,
• horizontal eddy viscosity, A- 3 00 m 2 /sec,
• grid spacing - 20 by 22 km (0 .2 by 0 .2 degrees),
• upper layer reference thickness, H1 - 200 m,
• lower layer reference thickness, H2 - 3300 m,
• minimum depth of bot~om t~pograjhy - 500 m,
• beta, df/dy - 2 x 10 " m' sec ',
• Coriolis partmeter at the southe rn boundary,
f - 4 .5 x 10-s sect',
• gravitational acceleration, g - 9 .8 m/sec2,
• reduced gravity, g' - .03 (H1 + H2) /H2 m/sec2 ,
• interfacial stress - 0,
• coefficient of quadratic bottom str ess - .003, and
• time step - 1 .5 hours .
EXPERIMENT 9 :
• upper layer inflow transport - 26 x 106 m' sect'

(26 Sverdrup),
• lower layer inflow transport - 4 x 10s m' sec
(4 Sverdrup),
• grid spacing, 25 by 25 km,
• lower layer reference thickness, H2 - 3400 m,
• Coriolis p1ramete~r at the southern boundary,
f - 5 x 1 0 S sec ',
• coefficient of quadratic bottom stress - .002 ; and
• time step - 1 hour .
• All other parameters as in the reference experi-
ment .
EXPERIMENT 28 :
• upper layer inflow transport - 26 x 106 m3 sect1

(26 Sverdrup),
• lower layer inflow transport - 4 x 10° m' sect'
(4 Sverdrup),
• wind stress - 0,
• coefficient of quadratic bottom stress - .002 .
ment .
105
EXPERIMENT 31 :
• upper layer inflow transport - 0,

• lower layer inflow transport - 0,
• wind stress from seasonal climatology based on ship
observations,
ment .
EXPERIMENT 34 :
• upper layer inflow transport - 26 x 10` m' sect'

(26 Sverdrup),
• lower layer inflow transport - 4 x 10s m' sectl
(4 Sverdrup),
• wind stress from seasonal climatology based on ship
observations,
ment .
EXPERIMENT 40 :
• All parameters as in the reference experiment .
EXPERIMENT 60 :
• horizontal eddy viscosity, A- 100 m2/sec,

ment .
EXPERIMENT 68 :
• wind stress from 12 hourly Navy Corrected Geostro-

phic Wind set,
ment .
106
The Department of the Interior Mission
As the Nation's principal conservation agency, the Department of the Interior has responsibility
for most of our nationally owned public lands and natural resources. This includes fostering
sound use of our land and water resources; protecting our fish, wildlife, and biological diversity;
preserving the environmental and cultural values of our national parks and historical places;
and providing for the enjoyment of life through outdoor recreation. The Department assesses
our energy and mineral resources and works to ensure that their development is in the best
interests of all our people by encouraging stewardship and citizen participation in their care.
The Department also has a major responsibility for American Indian reservation communities
and for people who live in island territories under U.S. administration.
The Minerals Management Service Mission

As a bureau of the Department of the Interior, the Minerals Management Service's (MMS)
primary responsibilities are to manage the mineral resources located on the Nation's Outer
Continental Shelf (OCS), collect revenue from the Federal OCS and onshore Federal and Indian
lands, and distribute those revenues.
Moreover, in working to meet its responsibilities, the Offshore Minerals Management Program
administers the OCS competitive leasing program and oversees the safe and environmentally
sound exploration and production of our Nation's offshore natural gas, oil and other mineral
resources. The MMS Minerals Revenue Management meets its responsibilities by ensuring the
efficient, timely and accurate collection and disbursement of revenue from mineral leasing and
production due to Indian tribes and allottees, States and the U.S. Treasury.
The MMS strives to fulfill its responsibilities through the general guiding principles of: (1) being
responsive to the public's concerns and interests by maintaining a dialogue with all potentially
affected parties and (2) carrying out its programs with an emphasis on working to enhance the
quality of life for all Americans by lending MMS assistance and expertise to economic
development and environmental protection.

Gulf of Mexico Circulation Modeling Study: Annual Progress, 7 D

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gulf of Mexico Circulation Modeling Study: Annual Progress, 7 D

Uploaded by

Copyright:

Available Formats

OCS STUDY

Annual Progress Repo

U .S . DEPARTMENT OF THE INTERIOR/MINERALS MANAGEMENT SERVICE

ArT'UAL PRAGRESS REPORT

United States Department of the Interior

Minerals Management Service

Prepared under Contract 14-12-0001-30073

Request Copies From : United States Department of the Interior

Attention : Public Information Unit

VIII . REFERENCES 102

This final report gives details of the first year

ocean circulation model of the Gulf on a 0 .2 degree grid .

tion, and on how best to include wind forcing . Details of

selected experiments are presented in this report .

incomplete knowledge of the real Gulf) simulation available

at the beginning of the project . It is forced by flow

exhibits many of the flow features observed in the Gulf .

Simulated surface currents sampled every ten days for three

Loop Current eddy cycles (1140 days) were delivered to MMS

run' . Experiment 34 is similar to Experiment 9, but with

the reduction in eddy viscosity and Experiment 60 exhibits

these features plus some new circulation patterns . Experi-

ment 68 is identical to Experiment 40 with the addition of

wind forcing from the Navy Corrected Geostrophic Wind data

hours from 1967 to 1982 . The addition of winds increases

the velocities encountered, and attempts to add this wind

forcing to Experiment 60 were not successful . Simulated

surface currents sampled every three days for more than 10

years were delivered to MMS from Experiment 68, as repre-

senting the best simulation available from the first year

Since the ocean model contained many innovative

successful Gulf of Mexico numerical simulation prior to the

start of this project . The model was driven from rest to

the Yucatan Straits which was compensated by outflow through

o upper/lower layer inflow transport - 26/4 Sv,

Figure 1 compares 'instantaneous' upper ocean flow

patterns just before an eddy is shed by the Loop Current (a)

._ . (1970) . The ability of the model to simulate observed

features is clearly demonstrated by this comparison, which

is remarkable given the simplicity of the model forcing .

However some discrepancies remain, for example the eddy has

that shown in the observations . Waves can be seen moving

the observations, but in the model they are at the limit of

resolution and therefore unrealistically large . Moreover in

contribute to the eddy shedding process (Vukovitch and Maul,

1984) . This is an example of a feature that would benefit

greatly from 10km model grid resolution . As shed eddies

velops a counter rotating vortex pair (Figure 2b), a struc-

ture repeatedly observed in the Western Gulf (Figure 3) .

formation of this structure have been a matter of some con-

troversy (Merrell and Morrison, 1981) . Although wind forc-

ing was not present in this simulation a major role for

Experiment 34 is similar to Experiment 9, but with

from ship observations (Elliot, 1979) . Linear interpolation

in time was used between the seasonal fields to produce the

o upper/lower layer inflow transport - 26/4 Sv,

Seasonal climatological winds obviously cannot

ments 28 and 34 (which are identical except that Experiment

28 has no wind forcing) . From these snapshots, taken every

90 days, there is little difference between the two experi-

ments . But if Experiment 34 is sampled every 20 days, as in

of about one year . Figure 5 also shows that the circulation