Old Dominion University
ODU Digital Commons
"!"<+42,*;276: .6;.9/797*:;*4"1?:2,*4!,.*6709*81?
Variability of Surface Pigment Concentrations in
the South Atlantic Bight
Charles R. McClain
James A. Yoder
L. P. Atkinson
Old Dominion University4**;326:77-<.-<
J. O. Blanton
T. N. Lee
See next page for additional authors
7447>;12:*6-*--2;276*4>793:*; 1B8:-202;*4,75576:7-<.-<,,87)8<+:
"*9;7/;1. 425*;.75576:*6-;1. !,.*6709*81?75576:
A2:9;2,4.2:+97<01;;7?7</79/9..*6-78.6*,,.::+?;1..6;.9/797*:;*4"1?:2,*4!,.*6709*81?*;!&202;*475576:;1*:+..6
*,,.8;.-/7926,4<:27626"!"<+42,*;276:+?*6*<;1792@.-*-5262:;9*;797/!&202;*475576:79579.26/795*;27684.*:.,76;*,;
-202;*4,75576:7-<.-<
#.87:2;79?2;*;276
,4*261*94.:#(7-.9*5.:;326:76"4*6;76!..% $260.9*6-<44.9*90.99*63'*92*+242;?7/
$<9/*,."205.6;76,.6;9*;276:26;1.$7<;1;4*6;2,201; CCPO Publications
1B8:-202;*4,75576:7-<.-<,,87)8<+:
!92026*4"<+42,*;2762;*;276
,4*26#(7-.9;326:76"4*6;76!..% $260.9<44.93*90.9'*92*+242;?7/:<9/*,.
8205.6;,76,.6;9*;276:26;1.$7<;1;4*6;2,201; Journal of Geophysical Research: Oceans, 93-72
28
JOURNAL
OF
GEOPHYSICAL RESEARCH, VOL.
93,
NO.
C9,
PAGES 10,675-10,697, SEPTEMBER
15,
1988
variability
of
Surface Pigment
Concentrations
in the
South
Atlantic Bight
YODER,
2
L.
P.
ATKINSON,
3
J.
0.
BLANTON,2
T.
N.
LEE,
4
J. J.
SINGER,5
AND
FRANK
MULLER-KARGER
6
A 1-year
time
sequence (November
1978
through October
1979)
of surface pigment images
from
the
South Atlantic Bight
(SAB)
was
derived
from
the Nimbus 7 coastal zone color scanner. This data set
is
augmented with
in
situ observations of hydrographic parameters, freshwater discharge,
sea
level,
coastal
winds,
and currents
for
the purpose of examining the coupling between physical processes and the spatial
and temporal variability of the surface pigment
fields.
The
SAB
is
divided into three regions: the east
Florida
shelf,
the Georgia-South Carolina shelf and the Carolina Capes. Six-month "seasonal" mean
pigment
fields
and time series of mean values within subregions
were
generated. While the seasonal mean
isopleths
were
closely oriented along isobaths, significant differences between seasons
in
each region
were
found
to exist. These differences are explained
by
correlating the pigment time series with physical
parameters and processes known to
be
important
in
the
SAB.
Specifically, summertime concentrations
between
Cape Romain and Cape Canaveral
were
greater than those
in
winter, but the opposite
was
true
north of Cape Romain. It
is
suggested that during the abnormally high freshwater discharge
in
the
winter-spring of
1979,
Cape Romain and Cape Fear
were
the major sites of cross-shelf transport,
while
the
cross-shelf exchange during the
fall
of
1979
occurred just north of Cape Canaveral. Finally, the
alongshore band of
high
pigment concentrations increased
in
width throughout the year
iii
the vicinity of
Charleston, but near Jacksonville it exhibited a minimum width
in
the summer and a maximum width
in
the
fall
of
1979.
INTRODUCTION
ring
the past
15
years, the
South
Atlantic
Bight
(SAB),
in
Figure
1,
has been
the
site
of
an
extensive multidisci-
y oceanographic research effort (see Blanton et
al.
],
Atkinson et
al.
[1985],
and
papers
in the collection
nography
of
the
Southeast
U.S.
Continental
Shelf
and
cent
Gulf
Stream
(Journal
of
Geophysical Research,
e
88,
number CS, 1983)). As a result, it
has
been found
a
wide
variety
of
biological, physical,
and
chemical
pro-
interplay to form a complex system
that
had
largely
unexamined.
Much
of
the
work
in
the
SAB
prior
to
the
1970s
focused
on
the vast
estuarine
system
that
extends
the
Outer Banks
of
North
Carolina
to
Cape
Canaveral,
·
a,
because it was believed
that
the
estuarine
and
the
hore
regimes were
the
primary
sites
of
the
biological
uction
[Haines and Dunstan, 1975; Turner et al., 1979;
r,
1981].
Few
observations
had
been collected
beyond
nearshore regime,
and
sampling
strategies for offshore
ob-
lions
were designed after
those
applied in
other
areas
as
the
Mid-Atlantic Bight
(MAB)
where the
spring
bloom
'nates the
annual
cycle. As investigations in
the
SAB in-
'ficd
during the 1970s, it
became
clear
that
the
Gulf
has a
major
impact
on
the
oceanography
of
the
SAB
that
the SAB was quite different from
the
MAB
and
the
I Coast systems [Pietrafesa, 1983b]. Because
of
the epi-
. nature
of
Gulf
Stream-induced
events,
traditional
sam-
2$
ASA
Godda~d
Space Flight Center, Greenbelt, Maryland.
l
kidaway
Institute of Oceanography, Savannah, Georgia.
Department
of Oceanography, Old Dominion University, Nor-
•
Virginia.
:osenstiel School of Marine and Atmospheric Sciences, University
1
•ami,
M1am1,
Florida.
Science
Applications International Corporation, Raleigh, North
ohna
.
,,_'
_Hborn
Point
Environmental Laboratories, University of Maryland,
-....
ndge.
Copyright
1988
by
the American Geophysical Union.
~number
8C0317.
48--0
227
/
88
/
008C-03
l
7$05.00
piing
methodologies
proved
inadequate
for resolving
the
space
and
time scales
important
on
this shelf,
and
revisions
of
pre-
vious ideas
regarding
dynamical
processes,
nutrient
sources,
and
primary
production
were required [Stefansson et al., 1971;
Blanton, 1971; Dunstan and Atkinson, 1976; Atkinson, 1977;
Atkinson et al., 1978; Lee and Brooks, 1979; Blanton et al.,
1981; Lee et al., 1981; Yoder et al., 1981, 1983; Yoder, 1985].
Thus,
subsequent
field
programs
such
as
Georgia
Bight
Exper-
iments
GABEX-1
in 1980
and
GABEX-11 in 1981,
the
Spring
Exchange
Experiment
(SPREX)
in 1985,
and
the
Fall
Ex-
change
Experiment
(FLEX)
in
1987 were designed using
sam-
pling
concepts
which
incorporated
multiple ships,
mooring
arrays,
and
aircraft.
Concurrent
with these developments, satellite infrared
ob-
servations
were
proving
to
be
invaluable for observing
Gulf
Stream
filaments
[Stumpf
and Rao, 1975; Legeckis, 1975; Vuko-
vich and Crissman, 1975]
and
the
deflection
of
the
Gulf
Stream
by a
bathymetric
feature called
the
Charleston
'
Bump
and
for
quantifying
the statistical
behavior
of
the
Gulf
Stream
front
[Bane and Brooks, 1979; Olson et al., 1983].
The
primary
limi-
tation
of
infrared
observations
of
the
SAB
is
that
sea surface
temperatures
are
fairly
uniform
during
the
summer
months,
so
that
frontal
boundaries
cannot
be determined.
In
October
1978,
the
Nimbus
7 coastal
zone
color
scanner
(CZCS) was
launched
offering
synoptic
year-round
estimates
of
near-surface
chlorophyll
concentration
[Hovis et al., 1980]
of
reasonable
accuracy
[ Gordon et al., l 980, 1983a; Walters,
1985; Barale et al., 1986].
Color
imagery
has
been
combined
with field
data
from the SAB
to
examine
specific events
and
processes [McClain et al., 1984; McClain and Atkinson, 1985;
Yoder et al., 1987].
These
studies
showed
that
the
pigment
retrievals
are
quite
good
and
that
the
structures
in
the
surface
pigment
fields
are
associated with subsurface
structures
in
other
water
properties such
as
temperature,
salinity,
and
nutri-
ent
concentrations.
lh
this
paper
a
I-year
time series
of
imagery is
integrated
with a variety
of
field
observations
in
order
to
examine
the
influence
of
physical processes
on
the
temporal
variability
and
the
spatial
distribution
of
surface
pigment
concentrations
10,675
10,676
McCLAIN
ET
AL.:
VARIABILITY OF
SURFACE
PIGMENT
CONCENTRATIONS
35°
33°
32°
31°
Region 2
30°
Region 1
28°
--,
{ Bahamas,
27°-+'""'"-''--'--"
........
~'+--~-=+'-'--'-~-~-"'----r'---~---1~
Fig.
1.
Map
of
the
South
Atlantic Bight with regions
1,
2,
and
3
outlined
and
locations
of
hydrographic transects, current meter
s,
tide
gauges,
and
meteorological stations identified.
throughout
the SAB.
The
term "surface"
is
emphasized be-
cause the
CZCS
maps concentrations only in the first optical
depth [Clark, 1981]
and
does not sense subsurface blooms,
which can be extensive
during
highly stratified conditions over
the shelf in the SAB
[Yod
er et al., 1983; Atkinson et al., 1984].
Optical depths can vary from less than l m in the turbid
nearshore areas to more
than
20 m in the
Gulf
Stream. Partic-
ular topics to be addressed include the cross-shelf migration
of
the nearshore pigment front, the influence
of
subsurface intru-
sions
on
the surface pigment field, the pathways
of
transport
of
material across the shelf,
and
the differences in the seasonal
modulation
of surface pigment concentrations in the Georgia
Bight
and
the Carolina Capes regions.
ENVIRONMENTAL
FACTORS
AND
FOR
C
ING
MECHANISMS
Shelf geometry, the
Gulf
Stream, the winds, river discharge,
and stratification act either directly
or
indirectly to cause epi-
sodic
and
seasonal variability in the cross-shelf
and
along-
shore distributions
of
surface phytoplankton. These factors are
not always independent
(e
.g
.,
the
Gulf
Stream
and
stratifi-
cation),
and
while there
are
a variety
of
dynamical interactions
which couple them, each category
is
unique in certain ways.
Basin Geometry and
Bathymetry
The
SAl3
shelf has a crescent shape which
is
very narrow
near
Cape
Canaveral
and
Cape
Hatteras
and
relatively
broad
off
Georgia
(maximum width
is
200 km). Five capes (Cana-
veral, Romain,
Fear
,
Lookout,
and
Hatteras) paFtition the
shelf into four embayments called, from
south
to north, the
Georgia Bight,
Long
Bay, Onslow Bay,
and
Raleigh Bay. The
Georgia Bight
and
Long
Bay form one continuous shelf,
but
Onslow Bay
and
Raleigh Bay are dynamically isolated by the
capes
and
shallow shoals which act
to
restrict exchange be-
tween the bays while enhancing cross-shelf exchange [Blanton,
1971; Blanton and Pietrafesa, 1978;
Atki
nson et
1
1
k
d
,r.
a·• 9
i~son an Pi
et
ra
1
esa, 19~~; Blanton et
al.,
1981;
Jano
Pietrafesa, 1982].
In
add1tion, there exists a sharp
in the continental slope off the coast of Georgia
pr~
Charleston Bump. This feature deflects the Gulf
S~~
sho~e [~rooks and Bane, 1978; Chao and Janowitz,
19
79].
sultmg m enhanced
Gulf
Stream meandering off the C
[Legeckis, 1979]
and
a quasi-permanent cyclonic
edd
~
called the Charlest_
on
Gyre
[McCla_in and Atkinson, l~S].
The
shelf break
1s
near the 60-m 1sobath and
is
quite a
Between
Daytona
Beach
and
Cape
Lookout the shelf
is
sloping, so
that
a distinct middle shelf region
separatea
freshwater nearshore region
and
the Gulf
Stream--<to
shelf break regime. This separation limits the influence
or
Stream- induced intrusions
on
the nearshore
regime
nearshore processes such as freshwater discharge
on
the
shelf regime.
The
middle shelf regime
is
not as distinct
ar
Cape
Canaveral
and
in Raleigh Bay. Also, the degree to
the inner, middle,
and
outer
shelf regimes interact
varies
sonally with stratification
and
with the strength or
the
shore front [Blanton,
1981
; Blanton and Atkinson,
1983).
Gulf
Stream
The
Gulf
Stream plays a
major
role in the dynamics
biology of the middle
and
outer shelf regimes. The
nisms through which the
Gulf
Stream influences the
tra
of nutrients
into
the euphotic zone on the continental
shell'
frontal eddy upwelling
[Pietraf
esa and Janowitz,
1979;
C
1981
; Y oder et al., 1981; Bane et al.,
1981
; Lee
and
Atki
1983; Pie
traf
esa, 1983a; McClain et al
.,
1984],
subsurfa1:e
trusions
of
Gulf
Stream water [Blanton,
1971
; Atkinson, I
Hofmann et al.,
1981
; Leming and
Mo
oers, 1981; Blantontt
1981
; Janowitz and Pietrafesa, 1982], and frontal
and
break upwelling during the winter
[D
ey, 1986;
Oey
et
1987].
The
frequency
of
occurrence of frontal
eddy
e
(also known as filaments
and
shingles)
is
of the order
of
2
10
days
and
are the major source
of
new nutrients to
the
o
shelf. Subsurface nutrient-rich intrusions of Gulf Stream
onto
the shelf have been observed in the Georgia Bight
and
Onslow Bay
but
have not been found in Long
Bay.
Intrusi
are the result of
Gulf
Stream meandering and the
cu
interaction with the capes
and
can be reinforced
by
upwel
favorable winds.
That
intrusions have not been
reported
Long
Bay may be due to the lack
of
observatio
ns
but
also be due to the fact
that
the
Gulf
Stream
is
usually
offi
of
the shelf break
at
that
location.
As
was mentioned earlier, the
Gulf
Stream
is
deflected
shore by the Charleston Bump, resulting in a cyclonic
ciraa--
lation over the
outer
shelf off Charleston. The upwelling
sociated with this circulation has been documented
by
Sin(llt
et al. [1983]
and
McClain and Atkinson [1985]. The
intellSI
ul
of the upwelling
is
probably modulated by the degree
of
G
Stream deflection
and
by increased stratification
in
the
summer which inhibits the upwelling
of
nutrients to the
sur·
face.
Stratification
All
locations in the SAB experience a seasonal
cycle
in
water column stratification [Atkinson, 1985] which
is
die
I f I
• . . . d • ·
Jar
insolauon.
resu t o seasona vanat1ons m wm m1xmg, so
1
subsurface intrusion frequency
and
freshwater discharge. a
the
fall
and
winter, surface c~oling causes convective
overh;
. . . · ·
so
that t
turnmg,
and
higher wmd stress enhances
m1xmg,
MCCLAIN
ET AL. : VARIABILITY
OF
SURFA
CE
PI
GMENT
CONC
ENT
RATI
ONS
1
0,6
77
'79
F M A M J J A S O N
- .
·--
·--··-··--
··-
•-54
••
,
..
_
-·---··--
·-
·-···
52
...
1 - -
--
·-··
a
-·
••
a a 37
L
---
--
- Daily
------
-----<
~~-
-----
5-Day Means
-----
-~
L--
----
Daily Means
--------<
6-Hourly
~==
=
====
===
===
==
=
==~
-
----,
H
H
H H
H
._
_____
6-Hourly
_____
---i
0
50
100
150
200 250
300 350
400
DAY
Fig. 2.
Data
set synopsis.
ty
fie
ld
over the inner and middle sh~l'. has rel~tively
littl_e
I or horizontal structure.
The
trans1t10n to highly strat1-
con
ditions, at
le
ast in the Georgia Bight, begins in late
a
nd
early spring when freshwater runoff peaks.
In
the
r,
the
relaxation
of
the winds and the increase in solar
tio
n further the process. The enhanced stratification pro-
th
e s
ub
surface intrusion process by decreasing the sur-
laye
r density, making it easier to displace.
The
range
of
-bottom temperature difference
at
midshelf
is
0°C in
r a
nd
can be greater
than
10
°C in summer. However, in
sum
m
er,
stratification can be offset by upwelling favorable
wh
ich
can cause the subsurface water near the coast
to
to
the surface.
e
ff
ect
of increased stratification
on
intrusion processes
atic [Atkinson, 1977]. Bottom intrusions
of
subsurface
Stream
water are confined to the outer shelf in the winter
ca
n extend to the coast of northern
Florida
in the
er
. The residence time
of
stranded intrusions can be a
th
or longer and
bottom
chlorophyll concentrations ex-
. g 7 mg/m
3
have been observed in the Georgia Bight
oder,
1
985
; Yoder et
al.
, 1985]. Current-bathymetry interac-
c
an
cause onshore
bottom
flow
of
Gulf
Stream water in
Geo
r
gia
Bight
north
of
Cape
Canaveral where the iso-
di
ve
r
ge
and offshore flow off Savannah where the iso-
s
be
g
in
to converge [Blanton et al.,
198
I;
J anowitz and
afesa,
1982].
In
Onslow Bay, Blanton and Pietrafesa
8) al
so
found preferential onshore subsurface flow
at
the
th
em
en
d of the bay.
hwater
Discharge
Riv
er
discharge in the SAB is concentrated between
Cape
r
and
Jacksonville. The Altamaha, Pee Dee,
and
Cape
riv
ers are the major contributors. Normally, the runoff
s
in
early spring, with a secondary peak in the fall, caus-
l
arge
b
ut
localized cross-shelf
and
alongshelf density gradi-
.
The
transport
of
this water across the shelf
is
determined
Be
ly
by the strength
of
a nearshore front [Blanton,
1981
;
ton and Atkinson, 1983] which serves as a barrier
to
later-
nti
xin
g.
The position
of
the front
is
influenced primarily by
arge,
stratification,
and
wind forcing.
It
tends to be fur-
off
shore
in
the spring
and
summer when the winds are
predominantly from the sou
th.
The
fr
ont can break down
during upwelling favorable wind events, allowing parcels of
low-salinity nearshore water , to separate
fr
om the
fr
ont
and
drift offshore,
and
fingers
of
high-pigment water stretching
across the shelf were often observed in the CZCS image
ry.
The
Cape
Fear
River empties
on
the south side of
Cap
e
Fear
,
and
the Pee Dee
Ri
ver empties
on
the
north
side of
Cape
Romain.
During
high-discharge conditions, large plumes
extending eastward from these capes suggest that these lo-
cations can be preferential sites
of
cross-shelf
tra
nsport.
Winds
The
winds play an
important
role in the circulation, partic-
ularly over the inner and middle shelf. Analyses
of
climatologi-
cal wind fields by Weber and Blanton [1980]
and
Blanton et
al. [1985] show a seasonal reversal
of
the general wind pat-
terns, which
are
northerly in the winter
and
southerly in the
summer.
Thus
the winds tend to support coastal upwelling in
the summer
but
tend to confine the nearshore front nearer to
the coast in the winter. Also, Janowitz and Pietrafesa [1980]
found
that
wind-driven upwelling can occur
at
locations
where changes in
bottom
slope are
abrupt
, such as along the
north
Florida shelf break. Recent observations
and
theory
suggest
that
upwelling occurs over the shelfbreak
and
upper
slope during
strong
southward
wind events [Oey, 1986; Oey et
al
.,
1987]. Coastal winds are correlated with offshore winds
[Schwing and Blanton, 1984; Wesiberg and Pietrafesa [198~]
but are lower in magnitude by as much as a factor
of
2.
This
coherence
is
greatest in the summer.
METHODS
AND
DATA
SETS
The
data
. sets (Figure
2)
include
CZCS
imagery spanning
the period from November
2,
1978, to November
4,
1979;
coastal winds from six locations; hydrographic
and
biological
dat
a from
five
cruises which covered the shelf from
Cape
Ca-
naveral to
Cape
Fe
ar; biological
data
from seven
other
less
synoptic cruises in the Georgia Bight (not shown in Figure 2);
sea level
data
from seven coastal locations; river discharge
from five sources;
and
surface currents
at
two locations.
czcs
The
CZCS
data
set consists
of
143 level-3 (remapped
and
registered
to
the coastline) pigment scenes extracted from
71
orbits.
The
coverage in the SAB was divided into three regions
shown in Figure 1: the east Florida shelf, from West
Palm
Beach to Jacksonville (region
l);
the Georgia-
South
Carolina
V)
12
11
10
9
~
8
w
~
7
~
6
e:i
5
(Il
~
4
z
3
2
~
I
~
v
CZCS
COVERAGE
L,
I,
p
L,
~
~
~
I
I,
I,
I
~
O
NOV
DEC
JAN
FEB
MAR
A
PR
MAY
JUN JUL
AUG
SEP
OCT
-
REGION
1
MO
NT
H
~
REG
I
ON
2
[Z2l
R
EGION
3
Fig. 3.
Number
of
level-3 scenes per
month
and
region.