BIO-OPTICAL VARIABILITY OF SURFACE WATERS IN
THE NORTHEASTERN GULF OF MEXICO
by
BISMAN NABABAN
A dissertation submitted in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Department of Marine Science
College of Marine Science
University of South Florida
Major Professor: Frank E. Muller-Karger, Ph.D.
Kendall L. Carder, Ph.D.
Robert H. Weisberg, Ph.D.
Chuanmin Hu, Ph.D.
Douglas C. Biggs, Ph.D.
Date of Approval:
Month date, 2004
Keywords:
© Copyright 2004, Bisman Nababan
Acknowledgements
[Bisman: this table of contents (and list of tables, figures) should be generated
automatically if you use the right fonts for the headings and subheadings. See an
example at
http://imars.marine.usf.edu/~hu/Piney_Point/reports/report_mid_term_USF_IMaRS.doc
It will take you forever to generate this type of tables manually, especially when
you’ll have to change the whole thing a few times to the least, according to our
comments. You need to learn how to work more efficiently]
TABLE OF CONTENTS
List of Tables……………………………………………………………………………..iii
List of Figures ……………………………………………………………………………..v
Abstract…………………………………………………………………………………...ix
INTRODUCTION………………………….……………………………………………..1
CHAPTER I: THE IMPACT OF RIVERS AND OCEANIC PROCESSES ON
CHLOROPHYLL-A VARIABILITY OF SURFACE WATERS
OF THE NORTHEASTERN GULF OF MEXICO………………………4
Abstract …………………………………………………………………..5
1.1. Introduction ………………………………………………………….6
1.2. Methodology ………………………………………………………...9
1.2.1. Study Area …………………………………………………..9
1.2.2. In-situ Data Collection, Processing and Analysis……………9
1.2.3. Water Depth of Influenced by Fresh Water Discharge……..12
1.2.4. Mixed Layer Depth…………………………………………13
1.2.5. Wind Field and Coastal Upwelling Index…………………..13
1.2.6. Satellite Data Collection, Processing and Analysis ………..17
1.3. Result and Discussion ……………………………………………..19
1.3.1. Spatial and Temporal Distribution of
Chlorophyll-a Concentration ………………………………19
1.3.2. Offshore dispersal of the Mississippi Plume in the
NEGOM ……………………………………………………21
1.3.3. Factors Affecting for Chlorophyll-a Concentration
Variability…………………………………………………..26
iii
1.3.4. In situ Observation for Offshore NEGOM Plume …………41
1.3.5. Nutrients Sources for Offshore NEGOM plume …………..47
1.3.6. Comparison between in situ and SeaWiFS chlorophyll-a
Concentration ………………………………………………55
1.4. Conclusion …………………………………………………………….59
CHAPTER II: COLORED DISSOLVED ORGANIC MATTER IN THE SURFACE
WATERS OF THE NORTHEASTERN GULF OF MEXICO ………..61
Abstract ………………………………………………………………..62
2.1. Introduction ………………………………………………………63
2.2. Methodology.……………………………………………………..66
2.2.1. Study Area ………………………………………………..66
2.2.2. Data Collection, Processing and Analysis ………………..66
2.3. Result and Discussion ……………………………………………69
2.3.1. Spatial and Temporal Distribution of CDOM and
Salinity……………………………………………………69
2.3.2. CDOM versus Salinity …………………………………...75
2.3.3. CDOM versus Chlorophyll ………………………………78
2.3.4. Estimating Salinity from CDOM ………………………...79
2.4. Conclusions ………………………………………………………83
CHAPTER III: VARIABILITY IN THE PARTICULATE ABSORPTION
COEFFICIENTS OF THE NORTHEASTERN GULF OF MEXICO
SURFACE WATERS ……………………………………..…………..85
Abstract ……………………………………………………..………....86
3.1. Introduction ………………………………………………………87
3.2. Methodology ……………………………………………………..90
3.2.1. Sample Collection ………………………………………..90
3.2.2. Absorption Measurements ………………………………..91
3.3. Result and Discussion …………………………………………….93
3.3.1. Total Particulate, Detritus and Phytoplankton Absorption
Coefficients ……………………………………………….93
3.3.2. Variation in Spectral Shape of the Phytoplankton
Absorption Coefficient …………………………………..101
3.3.3. Phytoplankton Absorption Coefficient vs. Chlorophyll-a
Concentration ……………………………………………117
3.3.4. Variation in Chlorophyll-Specific Absorption
Coefficient ……………………………………………….118
3.3.5. Sources of Variability in Chlorophyll-Specific
Absorption Coefficient …………………………………..123
3.4. Conclusions ……………………………………………………...126
iv
REFERENCES ………………………………………………………………………...128
LIST OF TABLES
Table 1.1. Cruise identifiers and dates ………………………………………………….10
Table 1.2. Starting and ending period (in week number of 52) of eastward wind
component measured at south of Mississippi delta (BURL1) and east of
Mississippi delta (42040) in relation to sea surface height and well-defined
Mississippi plume period. Sea surface height was extracted from the
location of 87.5W and 28.2N ……………………………………………..32
Table 2.1. Regression coefficient of determination (r2), intercept, and slope, with
respective standard error (se), between salinity and ag443 (S=a + b*ag443),
and mean percentage error (MPE) for salinity prediction based on ag443
data within 95% prediction interval of each region in different cruises
(seasons)……………………………………………………………………...77
Table 2.2. Mean percentage error of observed vs. predicted salinity employing
seasonal global NEGOM salinity-ag443 relationships to predict salinities
in other seasons………………………………………………………………81
Table 3.1. Mean values of a *ph (440) for all seven NEGOM cruises data corresponding
with the three major size cell markers of phytoplankton ………………….124
v
LIST OF FIGURES
Figure 1.1. Study area: the NEGOM region, encompassing the area between
27.3 - 30.7 N and 82.6 - 89.6W. The map shows the 10, 20, 100, 200,
500, and 1000 m bathymetric contours. A continuous line among closed
triangles is a cruise transect line during which surface chlorophyll-a
fluorescence, temperature and salinity observations were collected
continuously using a flow-through system. Closed triangles are stations
where CTD cast were employed with the cruise transect lines of L1-L11.
Numbered open squares show stations at which time series of
chlorophyll-a concentration and salinity were analyzed. Rectangle area
in the Gulf of Mexico inset is the area of the study ……………………….11
Figure 1.2. Examples of temperature, density, salinity, and fluorescence vertical
profiles from three different seasons showing mixed layer depth
estimate. Within river plumes, MLD estimate results could be
different (a) or similar (b) based on criterion used i.e., MLD1 based on
critical density gradient criteria and MLD2 based on critical temperature
gradient criteria. Outside river plumes both MLD estimate results are
normally consistent (c) …………………………………………………….14
Figure 1.3.
Spatial and temporal distribution of near-surface chlorophyll-a
concentration in the NEGOM region based on SeaWiFS data during
the nine separate NEGOM cruise periods. White and purple contours
are positive and negative total sea surface height, respectively …………...22
Figure 1.4.
Spatial and temporal distribution of near-surface chlorophyll-a
concentration in the NEGOM region based on in situ observation data.
Thin white line indicates the cruise transect during which continuous
flow-through chlorophyll fluorescence observations were collected.
vi
See Table 1.1 for cruise identification …………………………………....23
Figure 1.5.
Time series of near-surface monthly mean chlorophyll-a concentration
based on satellite SeaWiFS data for eight selected regions ………………24
Figure 1.6.
Near-surface mean chlorophyll-a concentration based on in situ
observation for eight selected regions. See Table 1.1 for cruise
identification and Figure 1.1 for region identification …………………….25
Figure 1.7.
Initial and subsequent position of the Mississippi River plume during
summer of 1998, 1999, and 2000. White and purple contour lines indicate
positive and negative sea surface height. White arrow indicates the
extent of the plume ………………………………………………………...27
Figure 1.8. Weekly mean time series of wind, upwelling index, river discharge, and
chlorophyll-a concentration off Mississippi region. A negative value for
upwelling index indicates an upwelling …………………………………...33
Figure 1.9. Lagged cross-correlation analyses for weekly mean river discharge and
upwelling index vs. chlorophyll-a concentration, with lags measured in
weeks ………………………………………………………………………34
Figure 1.10. Same as Figure 1.8 but off Alabama region ………………………………35
Figure 1.11. Same as Figure 1.8 but off Escambia region ……………………………...36
Figure 1.12. Same as Figure 1.8 but off Choctawhatchee region ………………………37
Figure 1.13. Same as Figure 1.8 but off Apalachicola region …………………………38
Figure 1.14. Same as Figure 1.8 but off Suwannee region …………………………….39
Figure 1.15. Same as Figure 1.8 but for central NEGOM region ………………………40
Figure 1.16. Near-surface salinity distribution in the NEGOM. Thin white line
indicates the cruise transect during which continuous flow-through
salinity observations were collected. See Table 1.1 for cruise
identification ………………………………………………………………44
Figure 1.17. Sea surface height (SSH) dynamic and ADCP-generated (10-14 m-deep)
currents in the NEGOM. Note: There are no ADCP data from cruise N4
east of 87W because of mechanical damage to ADCP equipment during
the cruise N4. See Table 1.1 for cruise identification ……………………45
vii
Figure 1.18. Average surface wind field in the NEGOM during the period of 9 cruise
observations (data obtained from Texas A&M University). Diamonds
and their identification indicate stations where wind speed and direction
were collected. Several offshore NDBC stations used to generate wind
gridded average were not shown in the picture because they were
located south of 27N …………………………………………………….46
Figure 1.19. Mixed layer depth distribution in the NEGOM. White dots indicate
CTD deployment stations. See Table 1.1 for cruise identification ……….51
Figure 1.20. Depth of salinity 34.5 distribution as a boundary of fresh water
influence from oceanic water in the NEGOM. See Table 1.1 for
cruise identification ………………………………………………………..52
Figure 1.21. Nutrients (NO3+NO2+NH4+Urea) mean concentration for 0-20 m-deep
of water column in the NEGOM. See Table 1.1 for cruise identification...53
Figure 1.22. Nutrients (NO3+NO2+NH4+Urea) mean concentration for 0-20 m-deep
of water column in the NEGOM. See Table 1.1 for cruise identification...54
Figure 1.23. Comparison between in situ and SeaWiFS estimates (OC4v4) of
chlorophyll-a concentration along ship track lines for nine NEGOM
cruises. MRE1 is negative MRE and MRE2 is positive MRE
(see section 1.2.6 for the formula) ………………………………………...57
Figure 1.24. Same as Figure 1.23, but the SeaWiFS chlorophyll-a concentration was
estimated with the MODIS algorithm (Carder et al., 1999) ………………58
Figure 2.1.
Near-surface ag443 (m-1) distribution in the NEGOM. Thin white line
indicates the cruise transect during which continuous flow-through
CDOM fluorescence observations were collected. See Table 1.1 for
cruise identification ……………………………………………………….73
Figure 2.2.
Mean and standard deviation of near-surface ag443 (m-1) (top panel),
salinity (middle panel) and chlorophyll concentration (bottom panel)
values for each subregion and for different time periods sampled. See
Table 1.1 for cruise identification …………………………………………74
Figure 2.3.
Scatter plots of salinity vs ag443 for each region in different seasons
viii
(cruises). See Table 1 for cruise identification. Yellow line is the global
NEGOM regression line for Winter 1998 (N4), yielding the lowest
error of observed vs. predicted salinities for spring and winter seasons
within 8 selected regions and global NEGOM region …………………….76
Figure 2.4.
Example of the error analysis of salinity predictions based on ag443
data. Mean percentage error (MPE) equals to average of absolute values
of 95% predicted interval values minus regression line values divided by
regression line value multiplied by 100%. MPE for (A)=0.04% and
(B)=8.12% (see also Table2). Note axis scales are different for better
display purpose ……………………………………………………………82
Figure 2.5.
Some example of observed and predicted global NEGOM salinity
employing global NEGOM salinity-ag443 relationship of other season.
MPE equals to average of absolute values of estimate salinity minus
observed salinity divided by observed salinity multiplied by 100%.
These two cases represent the best and the worst scenarios ………………82
Figure 3.1.
Near-surface total particulate absorption spectra in the NEGOM between
1998 and 2000. ‘N’ indentifies cruise number (Table 3.1). Note different
scales on y-axis ……………………………………………………………96
Figure 3.2.
Near-surface detritus absorption spectra in the NEGOM between 1998
and 2000. ‘N’ indentifies cruise number (Table 3.1). Note different
scales on y-axis ……………………………………………………………97
Figure 3.3.
Near-surface phytoplankton absorption spectra in the NEGOM between
1998 and 2000. ‘N’ indentifies cruise number (Table 3.1). Note different
scales on y-axis ……………………………………………………………98
Figure 3.4.
Proportion of detrital to total particulate absorption coefficient at
440 nm, as a function of chlorophyll-a concentration for each cruise.
See Table 1.1 for cruise identification …………………………………….99
Figure 3.5.
Time series of mean ap(440), ad(440), aph(440) and chlorophyll-a
concentration of each cruise, indicating possible concurrent seasonal
variability in the NEGOM. Right y-axis is for chlorophyll-a
concentration. See Table 1.1 for detail cruise identification ……………100
Figure 3.6.
Phytoplankton absorption spectra normalized to 440 nm from seven
NEGOM cruises. Individual normalized spectra are shown in black and
the average normalized spectra in blue. Panel H shows the average
normalized spectra for spring, summer and fall seasons. ‘N’ followed
by a number identifies each cruise (Table 1.1) …………………………..106
ix
Figure 3.7.
Near-surface distribution of the aph545/440 in the NEGOM. White dots
show location of water sample collections. See Table 1.1 for cruise
identification ……………………………………………………………..107
Figure 3.8.
Near-surface distribution of the aph625/440 in the NEGOM. White dots
show location of water sample collections. See Table 1.1 for cruise
identification ……………………………………………………………..108
Figure 3.9.
Near-surface distribution of the aph673/440 in the NEGOM. White dots
show location of water sample collections. See Table 1.1 for cruise
identification ……………………………………………………………..109
Figure 3.10. Near-surface distribution of the of mean aph545/440+aph625/440+
aph673/440 in the NEGOM. White dots show location of water sample
collections. See Table 1.1 for cruise identification ………………………110
Figure 3.11. Spatial distribution of biomass proportion of cell size marker for
microphytoplankton. See Table 1.1 for cruise identification …………….111
Figure 3.12. Spatial distribution of biomass proportion of cell size marker for
nanophytoplankton. See Table 1.1 for cruise identification ……………..112
Figure 3.13. Spatial distribution of biomass proportion of cell size marker for
pycophytoplankton. See Table 1.1 for cruise identification ……………..113
Figure 3.14. Relationship between the average of aph545/440+aph625/440+aph673/440
and microphytoplankton biomass proportion for each cruises. NA
indicates all NEGOM data were used. See Table 1.1 for detail of
cruise identification ………………………………………………………114
Figure 3.15. Relationship between average of aph545/440+aph625/440+aph673/440
and nanophytoplankton biomass proportion for each cruises. NA
indicates all NEGOM data were used. See Table 1.1 for detail of
cruise identification ……………………………………………………...115
Figure 3.16. Relationship between average of aph545/440+aph625/440+aph673/440
and picophytoplankton biomass proportion for each cruises. NA
indicates all NEGOM data were used. See Table 1.1 for detail of
cruise identification ……………………………………………………...116
Figure 3.17. Relationship between phytoplankton absorption coefficient at 440 nm
(aph(440)) and chlorophyll-a concentration for each cruises. Note the
scales on both x and y-axis are different between panels. See Table 1.1
x
for detail of cruise identification ………………………………………...119
Figure 3.18. Relationship between phytoplankton absorption coefficient at 673 nm
(aph(673)) and chlorophyll-a concentration for each cruises. Note the
scales on both x and y-axis are different between panels. See Table 1.1
for detail of cruise identification …………………………………………120
Figure 3.19. Near-surface chlorophyll-specific absorption spectra measured on the
NEGOM region during summer, spring and fall seasons between 1998
and 2000. ‘N’ followed by a number indicates cruise identification (see
Table 1.1 for cruises identification). ‘aph’ indicates phytoplankton
absorption coefficient. Note the different scales on y-axis ……………...121
Figure 3.20. Near-surface distribution of the chlorophyll-specific absorption
coefficient at 440 nm in the NEGOM. White dots show location of
water sample collections. See Table 1.1 for cruise identification ……….122
Figure 3.21. Scatter plots of a *ph (440) and biomass proportion per algal group for all
seven NEGOM cruises …………………………………………………...125
xi
SUMMARY
[I suggest a 1-page summary, and use longer text (possibly this below entirely) as a
CONCLUSIONS chapter at the end] [in table of contents you call it “Abstract” instead of
“SUMMARY” – be consistent] [I also suggest change this to a 1-page abstract, and then
add another “summary and conclusion” section at the end, and detail your findings in
brief words]
The alternating influence of nutrient inputs through riverine discharge and offshore
river plume dispersal, upwelling, and vertical mixing was studied in the northeastern Gulf
of Mexico (NEGOM) by examining weekly satellite data (ocean color, temperature,
altimetry) and in situ observations including chlorophyll-a concentration, absorption
coefficients of colored dissolved organic matter (CDOM) and particulate materials from
October 1997 to December 2000. Six major river-impacted regions, namely the
Mississippi (R1), Alabama (R2), Escambia (R3), Choctawhatchee (R4), Apalachicola
(R5), Suwannee (R6), and Tampa Bay (R7), plus one offshore region (R8) were used to
characterize the NEGOM. Nine two-week cruises were conducted in three different
seasons between 1997 and 2000 onboard the Texas A&M University R/V Gyre: spring
(April or May), summer (July and August), and fall (November).
Relatively high chlorophyll-a concentration (>1 mg m-3) was generally observed
over the inner shelf, [define what is inner shelf or inshore in terms of bathymetry, for
example, < 50m?] particularly near river mouths, and in well-defined plumes that
extended offshore and to the south into the central or eastern Gulf during the three
summer seasons (1998, 1999, and 2000). Coastal upwelling seemed to be an important
xii
local phenomenon in specific areas along the coast or shelf in the NEGOM but it was
intermittent. River discharge, however, appeared to be the major factor controlling
chlorophyll-a concentration variability in coastal and shelf areas near and off the deltas of
the Alabama, Escambia, Choctawhatchee, Apalachicola and Suwannee rivers. Strong
interannual variability (up to a factor of four) in in situ chlorophyll-a concentration was
observed during winter and spring in 1998 compared with 1999 and 2000 off Escambia,
Choctawhatchee, Apalachicola, Suwannee and Tampa Bay regions. Anomalously high[is
this correct? How anomalous and relative to what?] chlorophyll-a concentration in 1998
was related to higher fresh water discharge during the 1997-1998 El Niño-Southern
Oscillation event. The Mississippi plume extended well over 500 km from the river delta
and reached as far as the Florida Keys for periods lasting over 14 weeks between May
and September, typically flowing west of the Dry Tortugas and then into Florida Strait.
Off the Mississippi delta, the plume appeared to be forced offshore in part by eastward
winds or entrained by occasional anticyclonic eddies that approached or formed near the
slope off the delta.
In situ chlorophyll-a observations showed that the SeaWiFS algorithm OC4v4 led
to an error of +35% during fall cruises and as high as 140% during spring and summer
cruises, when river plumes were extensive, because they contained colored dissolved
organic matter (CDOM). Strong correlations between the CDOM absorption coefficient
at 443 nm (ag443) and CDOM fluorescence (330 nm excitation and 450 nm emission
filters) were consistently observed [need to give quantitative info on relationship].
Relatively high ag443 (>0.1 m-1) occurred near major river mouths, but decreased rapidly
xiii
with distance from shore except within the Mississippi River plume [need to give
quantitative info on relationship – how much did it decrease, what are values near mouth;
also, give salinity values when you mention high ag values at river mouths]. In general,
ag443 covaried linearly and inversely with salinity in nearshore areas during spring and
winter cruises, indicating conservative mixing Using all NEGOM data of winter 1998
(N4), the salinity-ag443 relationship i.e., Salinity (S)=36.59-29.86*ag443 (n=8771, r2=0.86,
0.01<ag443<0.52, 16<S<36) produced the lowest mean percentage errors of observed vs.
predicted salinities (<3% errors, correspond to <1.04 in salinity value) for all other spring
and winter seasons, therefore, this relationship may be useful to estimate salinity from
satellite ocean color data. However, estimates with this relationship led to unacceptable
errors in the summer (>10% errors in salinity estimates), due to less of a coherent
relationship between ag and salinity over the range of salinities observed. Only within the
river plumes in outer shelf and slope regions was the correlation somewhat more robust
and similar to that seen nearshore.
The absorption coefficient of phytoplankton at 440 nm [aph(440)] followed a power
law relationship (r2=0.83-0.98) with chlorophyll-a concentration. The chlorophyllspecific a *ph (440) varied by a factor of 7. Lower values of a *ph (440) (<0.06 m2 mg-1) were
observed over the inner shelf and near the outflow regions of major rivers. During
summer, lower values of a *ph (440) were also observed in the outer shelf and slope in the
western NEGOM (west of 85W) in low salinity waters of the Mississippi River plume.
Higher values of a *ph (440) (>>0.1 m2 mg-1) were otherwise observed outside the river
plumes in the outer shelf and slope, where lower chlorophyll-a concentration occurred.
xiv
Packaging effects and pigment composition were major factors in determining a *ph (440)
variability. There were no differences in the mean spectral phytoplankton absorption
coefficients between spring and fall seasons, but these were different from those observed
in the summer. The average value of a *ph (440) for microphytoplankton was also
significantly lower than that of nanophytoplankton and picophytoplantkon groups, which
suggested that an increase in cell size increased pigment packaging. Meanwhile, aph(673)
covaried linearly with chlorophyll-a concentration in all seasons (r2=0.87-0.99), and
spatial and temporal variability of the chlorophyll-specific a *ph (673) was very small.
Using microphytoplankton biomass proportion as an independent variable, we were able
to explain 48% of the variance in the phytoplankton absorption coefficient at wavelength
of 440 nm. There was no indication that a local increase of chlorophyll concentration led
to increased CDOM, except in the vicinity of Tampa Bay during summer cruises [how do
you know, how much of an increase?].
[You should conclude with a statement on nutrient inputs such as a budget that
compares upwelling and mixing with river plumes, and impacts on the NEGOM]
[Bisman: this conclusion is very important and definitely should be added]
xv
INTRODUCTION
The Northeastern Gulf of Mexico (NEGOM), defined as the region of north of
27°N between the Mississippi River Delta and the Florida coast (Figure 1.1), is one of
great national importance because of its oil, gas, fisheries, and recreation resources. Some
current and future mining activities in the NEGOM may potentially result in oil spills that
would affect its water quality, shelf, and coastal resources such as wetlands and coastal
ocean bottoms. Further, the NEGOM coast includes large areas of wetlands and beaches
supporting great economic importance for tourism and fisheries industries. Yet, this
region remains one of the least studied areas within the Gulf of Mexico.
The NEGOM region receives more fresh water and riverine sediments such as from
the Mississippi, Mobile, Apalachicola and Suwannee Rivers than any other North
American shelf which leads to strong seasonal buoyancy variation on the shelf, and to
high nutrient, chlorophyll, and high dissolved organic carbon concentration (Turner and
Rabalis, 1999; Pennock et al., 1999; Guo et al., 1999; delCastillo et al., 2000; Gilbes et
al., 1996). This region also experiences the largest upwelling within the Gulf of Mexico
(Muller-Karger, 2000; Muller-Karger, et al., 1991; Gilbes et al., 1996), and sometimes
affected by the Loop Current or by cold and warm rings shed by the large western
Boundary Current (Vukovich et al, 2000; Vukovich and Hamilton, 1989). However, the
mechanisms leading to upwelling in the NEGOM, and the spatial and temporal scales of
the process, are still poorly understood.
1
The diversity of oceanographic processes in NEGOM region leads to variability in
the distribution of chlorophyll-a, dissolved organic carbon, salinity, and bio-optical
properties. Rivers affect the chemistry and biology of estuarine zones within the
NEGOM region (Livingston et al., 1975; McPherson et al., 1990). However, synoptic
patterns of variability of chlorophyll-a, bio-optical properties and the variability in factors
that drive this are still unclear.
Earth observing satellite data such as SeaWiFS, AVHRR, and TOPEX/POSEIDON
can provide valuable information on the distribution of river water and sediments on the
continental shelf as well as on the circulation affecting the interaction of river plume with
oceanic water. Using satellite data and extensive in situ hydrographic, biological, and
optical data, I examined the variability of bio-optical properties of surface water in the
Northeastern Gulf of Mexico, with two major goals:
1) To determine the mean and variability of bio-optical properties in this region;
2) 2) To evaluate controlling mechanisms for such variations.
Specifically, my overall objectives are: [you need to change the orders of these, perhaps
(3), (4), add one for aph*, (1), (2), (5)
(1) To assess the relative importance of upwelling relative to river discharge in
controlling the variability of chlorophyll-a concentration over the NEGOM shelf
environment.
(2) To describe the effect of short-lived warm slope eddies and cold slope eddies, winds
and surface current changes on the temporal and spatial variability of chlorophyll-a
2
concentration, colored dissolved organic matter (CDOM), and absorption coefficient
of particulate materials of the NEGOM surface waters.
(3) To determine inter-annual variability of chlorophyll-a concentration.
(4) To examine the spatial and temporal variability of CDOM and the relationship
between surface salinity and CDOM, and to further
determine whether surface salinity in the NEGOM can be estimated from CDOM.
(5) To evaluate several bio-optical algorithms for estimation of chlorophyll-a
concentration. [where are these evaluations? In Chapter 1 abstract you only talked
about OC4V4?]
The dissertation is developed into three separate chapters and a conclusion
statement. Chapters are entitled as follows:
Chapter I: “The impact of rivers and oceanic processes on chlorophyll-a variability of
surface water of the Northeastern Gulf of Mexico”.
Chapter II: “Colored dissolved organic matter in the surface waters of the Northeastern
Gulf of Mexico”.
Chapter III: “Variability in the particulate absorption coefficients of the Northeastern
Gulf of Mexico surface waters”.
3
CHAPTER I:
THE IMPACT OF RIVERS AND OCEANIC PROCESSES ON
CHLOROPHYLL-A VARIABILITY OF SURFACE WATERS
OF THE NORTHEASTERN GULF OF MEXICO
4
Abstract
Spatial and temporal variability of chlorophyll-a concentration in the Northeastern Gulf
of Mexico (NEGOM) were studied using weekly satellite data and in situ observations
from October 1997 to December 2000. Relatively high chlorophyll-a concentration (>1
mg m-3) was generally observed inshore [define what is inshore with bathymetry]
particularly near the major river mouths, and in a well-defined plume extending offshore
to the south during the three summer seasons (1998, 1999, and 2000). Even though
coastal upwelling is an important factor in providing nutrients to localized phytoplankton
blooms {Briefly state how you arrived at this conclusion], the nutrients provided by river
discharge appeared to be the major factor controlling chlorophyll-a concentration
variability in coastal areas near the deltas of the Alabama, Escambia, Choctawhatchee,
Apalachicola and Suwannee rivers. Strong interannual variability (up to a factor of four)
in chlorophyll-a concentrations was observed during winter and spring in 1998 compared
with 1999 and 2000 off Escambia, Choctawhatchee, Apalachicola, Suwannee and Tampa
Bay regions. The anomalous chlorophyll-a concentration in 1998 was related to high
fresh water discharge during the 1997-1998 El Nino event. The Mississippi River and
5
upwelling both have influence over chlorophyll-a variability off the Mississippi region.
The Mississippi plume extended over 500 km from the river delta and as far as Florida
Keys for periods lasting over 14 weeks between May and September. The plume was
forced offshore in part by eastward winds and the occasional presence of anticyclonic
eddies near the slope off the Mississippi delta. In situ chlorophyll-a observations showed
that the SeaWiFS algorithm OC4v4 led to an error of +35% during fall cruises and as
high as 140% during spring and summer cruises.
[It reads like that you put all abstracts from the three chapters together to form your
summary section – this is not good. Even if you want to say the same thing, in different
places of a thesis (or paper) you need to say it differently, i.e., rephrase it]
1.1. INTRODUCTION
The Northeastern Gulf of Mexico (NEGOM) is an area that experiences significant
variability due to a variety of oceanographic processes acting on shelf and coastal waters.
Particularly, upwelling, cold- and warm-core rings, other circulation patterns, river
discharge, and winds, force significant biogeochemical variability within the NEGOM.
Several studies have focused on the effect of river input on chemistry and biology of
estuarine zones within the NEGOM (Lohrenz et al., 1990; McPherson et al., 1990;
Livingstone et al., 1975). However, synoptic patterns of variability of chlorophyll-a and
the variability in factors that drive this are still unclear.
The Gulf of Mexico is frequently considered to be an oligotrophic system (Ortner et
al., 1984; Biggs, 1992). However, satellite data (Muller-Karger et al., 1991, MullerKarger, 2000; Gilbes et al., 1996) has shown that the Gulf of Mexico does experience
6
intermediate to high phytoplankton concentration both over the shelf and in offshore
areas, and that some of these patterns are seasonal. Some changes in nearshore areas are
related to wind-driven coastal upwelling (Muller–Karger, 2000) and river plumes
(Pennock et al., 1999). Over the shelf, variation in chlorophyll-a is effected by seasonal
convective mixing, upwelling associated with eddies (Biggs and Muller-Karger, 1994),
and the offshore dispersal of riverine outflow.
Using Coastal Zone Color Scanner (CZCS) images, Muller-Karger et al. (1991)
found that to a first order variability in pigment concentration seaward of the shelf was
synchronous throughout the Gulf, with the highest values from December to February
and the lowest from May to July. Using CZCS images, Gilbes et al. (1996) observed that
an episodic plume with high pigment concentration occurs along the West Florida Shelf
each the spring and persists 1-6 weeks in a pattern that extended >250 km southward
along the shelf. During the summer, low pigment concentration was generally observed
along the outer shelf and continental slope of the West Florida Shelf (Gilbes et al., 1996).
The NEGOM coast receives significant fresh water input from rivers like the
Mississippi, Alabama, Escambia, Choctawhatchee, Apalachicola, and Suwannee Rivers.
The river discharge leads to seasonal buoyancy variation, and to high nutrient,
chlorophyll-a and high dissolved organic carbon (DOC) concentration in inner shelf areas
(Pennock et al., 1999; Des Castillo et al., 2000, Gilbes et al., 1996). The Mississippi
River provides more than half of all of the fresh water input into the Gulf of Mexico
(Deegan et al., 1986) and is the dominant source of terrestrial nutrient input (Twilley et
al., 1999). Indeed, Walsh et al. (1989) suggested that the Mississippi plume area is the
7
most productive within the Gulf of Mexico. Chlorophyll-a concentrations in the
Mississippi River plume are highly variable in space and time, ranging between 1.1 and
14.4 mg/m3 in the outer plume during four cruises in 1990-1992 (Redalje et al., 1994),
but as high as 40-80 mg/m3 in near-coastal waters during the late spring in 1993 (Nelson
and Dortch, 1996).
While buoyancy and wind forcing control the circulation of shelf waters (Weisberg
et al., 1996; Yang and Weisberg, 1999), deeper waters are also affected by northward
intrusions of the Loop Current and associated eddies (Vukovich, 1988; Huh et al., 1981;
Muller-Karger, 2000; Nowlin et al., 2000). The Loop Current rarely penetrates north of
28 °N in the eastern Gulf of Mexico (Vukovich et al., 1979); therefore, its direct
influence in the NEGOM region is considered occasional (Muller-Karger et al., 2001).
Huh et al. (1981) reported that an intrusion of the Loop Current in February 1977 reached
the DeSoto Canyon and flowed onto the west Florida Continental Shelf to within 8 km of
the shore. The duration of the event was 18 days and it affected the bio-chemical and
physical characteristics of the NEGOM waters. Advanced Very High Resolution
Radiometer (AVHRR) Seas Surface Temperature (SST) images showed that during the
period of Jan 15-28 and Mar 25-April 3, 2000, Loop Current intrusions reached 29.3 N,
close to the Mississippi Delta (Muller-Karger et al., 2001).
Using satellite data from Sea-viewing Wide-Field-of-View Sensor (SeaWiFS),
AVHRR, TOPEX/POSEIDON and field observations of hydrographic and biological
data from nine cruises within the NEGOM region, I examined the variability of
chlorophyll-a concentration in the NEGOM region and assessed factors that controlled
8
the variability. The specific objectives of this section of the dissertation are: (1) to
determine the spatial and temporal variability of chlorophyll-a and to identify the factors
affecting the variability; (2) to determine the extent, duration and forcing factors of the
offshore NEGOM plume during the three consecutive years of 1998, 1999, and 2000; and
(3) to evaluate several bio-optical algorithms for estimation of chlorophyll-a
concentration.
1.2. METHODOLOGY
1.2.1. Study Area
The study was conducted in the region extending from the Mississippi River Delta
to the West Florida Shelf off Tampa Bay. For this study the NEGOM is defined inshore
by the 10-m isobath and offshore by the 1000-m isobath. Six major river-impacted
regions, namely the Mississippi (R1), Alabama (R2), Escambia (R3), Choctawhatchee
(R4), Apalachicola (R5), Suwannee (R6), and Tampa Bay (R7), plus one offshore region
(R8) were chosen to characterize the NEGOM (Figure 1.1).
1.2.2. In-situ Data Collection, Processing and Analysis
Nine two-week cruises were conducted in three different seasons between 1997 and
2000 onboard the Texas A&M University R/V Gyre (Table 1.1): spring (April or May),
9
summer (July and August), and fall (November). Each of the cruises surveyed eleven
cross-margin transects from the 10-m to the 1000-m isobath (Figure 1.1).
About 100 stations were occupied for conductivity-temperature-depth (CTD) casts
on each cruise. During a CTD cast, water samples from twelve 10-liter niskin bottles
were collected for oxygen, salinity, nutrients, DOC, and pigment analysis. Salinity was
analyzed by Texas A&M University using Autosal Laboratory Salinometer technique as
described in Fofonoff and Millard, Jr. (1983). Nitrate and Nitrite were also analyzed by
Texas A&M University using AutoAnalyzer technique as described in Atlas et al. (1971).
Both salinity and nutrients were analyzed aboard ship.
In addition to the discrete stations, continuous surface flow-through data for
chlorophyll-a fluorescence, temperature (SST), and salinity (SSS) were recorded every 2
minutes along the transects. Flow-through data were collected using a stream pumped at
a rate of 10 liters/minute from a hull depth of 3-m into the main laboratory through a
debubbler and mixing chamber of 10-liter volume. The residence time of water in the
mixing chamber was about one minute. Using garden hoses connected by adjustable ball
valves to a “Y” splitter valve leading off the debubbler, the flow of the water was reduced
from 10 liters/minute to 1 liter/minute. The 1 liter/minute flow was directed to the
conductivity and temperature sensor (Sea-BirdTM) and to the Turner DesignsTM model-10
Fluorometer for chlorophyll-a fluorescence. The Fluorometer designated for chlorophylla fluorescence uses a blue filter for exitation and red filter for emission.
Table 1.1. Cruise identifiers and dates.
10
Cruise no.
1
2
3
4
5
6
7
8
9
Start date
16 November 1997
4 May 1998
25 July 1998
13 November 1998
15 May 1999
15 August 1999
13 November 1999
15 April 2000
28 July 2000
End date
Cruise ID
26 November 1997 N1
15 May 1998
N2
6 August 1998
N3
24 November 1998 N4
28 May 1999
N5
28 August 1999
N6
23 November 1999 N7
26 April 2000
N8
8 August 2000
N9
Cruise season
fall/winter
spring
summer
fall/winter
spring
summer
fall/winter
spring
summer
Figure 1.1. Study area: the NEGOM region, encompassing the area between 27.3 - 30.7
N and 82.6 - 89.6W. The map shows the 10, 20, 100, 200, 500, and 1000 m bathymetric
contours. A continuous line among closed triangles is a cruise transect line during which
surface chlorophyll-a fluorescence, temperature and salinity observations were collected
continuously using a flow-through system. Closed triangles are stations where CTD cast
were employed with the cruise transect lines of L1-L11. Numbered open squares show
stations at which time series of chlorophyll-a concentration and salinity were analyzed.
Rectangle area in the Gulf of Mexico inset is the area of the study.
Water samples from the fluorometer out-flow were collected at about 100 locations
to calibrate the fluorescence data into chlorophyll-a concentration. Glass fiber filters
(GF/F) were used to filter water samples and pigments were extracted using acetone.
11
Samples were extracted in 90% acetone immediately after water collection and then
analyzed at sea using a calibrated Fluorometer.
Continuous subsurface (10-14 m-deep) currents were measured using a shipboard
153 kHz Acoustic Doppler Current Profiler (ADCP). To produce circulation maps, the
ADCP data were gridded employing 40 km as the smoothing radius (Daley, 1991).
Features with scales less than 40 km are therefore likely smoothed out, but this allowed
filtering out noise features and keeping the basic large-scale patterns. Details of these
measurements and data processing can be found in Jochens et al. (2002) and Wang et al.
(2002). To see the pattern of subsurface currents, the ADCP gridded currents was
overlaid with the sea surface height (SSH) field at the middle date of each of the nine
cruises.
1.2.3. Depth of River Plumes in the NEGOM and Mixed Layer Depth
River discharge data were compiled from six major rivers entering into NEGOM
region i.e., The Mississippi, Alabama, Escambia, Choctawhatchee, Apalachicola and
Suwanne Rivers for the period of October 1997 to December 2000. The data were
obtained from the U.S. Geological Survey Water Data Report. Weekly mean time series
were built from these rivers discharge data.
The Mississippi, Alabama, Escambia, Choctawhatchee, Apalachicola, and
Suwannee discharge was recorded respectively at Tarbert Landing (30.96N, 91.66W),
Millers Ferry (32.1N, 87.40W), near Century (30.96N, 87.23W), near Bruce
(30.45N, 85.89W), Sumatra (29.95N, 85.02W), and Branford (29.96N, 82.93W).
12
To estimate the water depth influenced by fresh water in the NEGOM region, a
subjective salinity of 34.5 was used as the end-member for water influenced by the river
discharge. This is because that salinity of 34.5 was generally observed along the river
plume periphery. The depth of salinity 34.5 was determined from salinity profile of each
CTD station.
Vertical mixing is one of the mechanisms for renewal of nutrients in the euphotic
zone (Eigenheer et al., 1996; Nelson and Smith, 1991; Bishop et al., 1986; McGowan
and Hayward, 1978). The mixed layer depth (MLD) is the depth range through which
surface fluxes have been mixed in the recent past or the zone of relatively homogenous
water formed by the history of mixing (Brainerd and Gregg, 1995). Eigenheer et al.,
(1996) found that the temporal variability of the mixed layer depth caused variability in
the timing of the onset of phytoplankton blooms.
To determine the influence of changes in mixed layer depth on chlorophyll-a
concentration, MLD was calculated using both a critical temperature gradient criterion
(T/z = 0.1 C m-1) and critical density gradient criterion (/z = 0.04 kg m-2) from
CTD profiles of temperatures and density. Outside river plumes, both critical temperature
and density gradient criterions produced consistent MLD estimates. Within river plumes,
temperature and density gradient criteria produced different MLD estimates. Therefore,
within river plume region, a density gradient criterion was used to estimate MLD (Figure
1.2).
1.2.4. Wind Field and Coastal Upwelling Index
13
Hourly wind speed and direction were obtained from NDBC buoys BURL1
(28.90N, 89.40W) for the Mississippi region, DPIA1 (30.25N, 30.25W) for Alabama,
Escambia, and Choctawhatchee regions, CSBF1 (29.67N, 85.36W) for Apalachicola
region, and CDRF1 (29.14N, 83.03W) for Suwannee region. Hourly wind speed and
direction were averaged to produce daily and weekly mean wind speed and direction
Figure 1.2. Examples of temperature, density, salinity, and fluorescence vertical profiles
from three different seasons showing mixed layer depth estimate. Within river plumes,
MLD estimate results could be different (a) or similar (b) based on criterion used i.e.,
MLD1 based on critical density gradient criteria and MLD2 based on critical temperature
gradient criteria. Outside river plumes both MLD estimate results are normally
consistent (c).
employing a true vector average. In this scheme, the magnitude of the vector was
represented by the wind speed observation and the direction observations were used for
the orientation. The vectors were then broken down into their u and v components. All u
and v components are then averaged separately. The resulting average speed and
14
direction were calculated from the Pythagorean Theorem and “arctan(u/v)”, respectively.
Surface wind field data for NEGOM region of each of nine cruises were obtained from
Texas A&M University. The product was produced by gridded averaging of surface
wind field data from several NDBC buoys station in the NEGOM region for the same
cruises period of time.
Upwelling is a very important process in marine ecosystems since the upwelled
water has higher concentrations of nutrients such as nitrate, phosphate and silicate than
the original surface water (Huyer, 1983; Traganza et al., 1981; Kosro et al., 1991); such
nutrients are key to sustaining biological production. Upwelling water is also typically
cooler and more saline (Huyer, 1983; Kosro et al., 1991).
Based on Ekman’s (1905) theory and under Ekman’s assumptions of steady state
motion, uniform wind, and infinite homogenous ocean, the mass transport per unit width
of ocean surface is directed 90 to the right of the wind direction in the Northern
Hemisphere and is related to the magnitude of wind stress as follows:
y
Mx = ----f
(1)
where Mx [units?] is the offshore mass transport (Ekman transport) resulting from a wind
stress parallel to local coastline orientation, y, [units?] and f [units?] is the Coriolis
parameter. To estimate Ekman transport perpendicular to the coastline (Bakun, 1973 and
1975; Mason and Bakun, 1986; Schwing et al., 1996) for the northern coasts of the Gulf
of Mexico, the wind component parallel to the local coastline was estimated (Arfken,
1985) as follows:
15
un = uo cos  + vo sin 
(2a)
vn = -uo sin  + vo cos 
(2b)
where un and vn are u and v components of the wind after rotation so that un is parallel to
the coast and vn is perpendicular, uo and vo are u and v of original wind components, and
 is the counterclockwise rotation angle with land laying to the right side. Sea surface
wind stress was calculated using the classic square-law formula as follows:
x = aCd |un| un
(3a)
y = aCd |vn| vn
(3b)
where x is the x component of the wind stress vector, y is the y component, a is the air
density [units?], Cd [units?] is an empirical drag coefficient, and un and vn are the wind
vectors [units?] near the surface in the x and y component with magnitudes |un| and |vn|.
The magnitude of the wind vector [isnt’ this the original wind speed?] is estimated with
the following formula:
_________
|un| = |vn| = (u2n + v2n)
[this equation has serious errors –
check and correct it]
(4)
The drag coefficient (Cd) [units?] is itself a function of wind speed. In this study, I used
the approach of Hsu (1986) to estimate the Drag Coefficient as follows:
Cd = [0.4/(14.56 - 2 x ln(wind speed)]2
(5)
The coastal upwelling index (CUI) per 100 meter coastline was estimated from
Ekman Transport (Mx) using the following formula:
100 m coastline
CUIx = Mx -------------------1025 kg/m^3
16
(m3/s/100 m coastline)
(6)
[100 m in the nominator does not mean “per 100 meter”?]
Negative (offshore Ekman transport) value of the CUIx leads to an upwelling process
whereas positive (onshore Ekman transport) value of the CUIx leads to a downwelling
process.
While the derivations and calculations of the CUIx were based on established theory
and based on numerous assumptions and simplified parameterizations, they are,
nonetheless, a simplified and incomplete representation of the physics that drive coastal
circulation and specifically of Ekman transport. Schwing et al. (1996) concluded that
upwelling was as much due to the combined effect of coastal upwelling (the process that
is the basis of the coastal upwelling index) and Ekman pumping (wind curl, or the spatial
variation in the wind). The Ekman pumping process may be an equally important
contributor to surface Ekman divergence and upwelling (Bakun and Nelson, 1991),
therefore the CUIx presented here may not represent the true amount of upwelling.
1.2.5. Satellite Data Collection, Processing and Analysis
The satellite data used in this study are sea surface temperature from NOAA
AVHRR, chlorophyll-a concentration from SeaWiFS ocean color measurements, and sea
surface height from TOPEX/ERS observations. The first two were collected using a high
resolution picture transmitter (HRPT) antenna located at University of South Florida
(USF), St. Petersburg, Florida, USA.
For SST, all passes (nighttime and daytime) from each NOAA satellite were
combined over specified time periods to build the time series. The visible and nearinfrared (IR) channels were calibrated to percent albedo, and the IR channels were
17
calibrated to degrees Celsius. The navigation was manually corrected to compensate for
errors in the satellite clock and orientation. The SST was computed using the multichannel sea-surface temperature (MCSST) algorithm developed by McClain et al. (1985).
The MCSST algorithm uses channels 4 and 5 to compute the temperature, with channels
2, 3, and 4 used to determine cloud cover. Through various comparisons (unpublished
data), it has been confirmed that the approximate root mean square error of the AVHRR
SST retrievals are of the order of 0.5 °C (see also McClain et al., 1983; Walton, 1988;
Wick et al., 1992).
SeaWiFS data were processed using the atmospheric correction algorithms
described by Gordon and Wang (1994), and Chlorophyll-a concentration fields were
derived using the OC4v4 bio-optical algorithm described by O'Reilly et al. (2000). I am
aware that there are unresolved issues that lead to severe errors in estimates of
chlorophyll-a using SeaWiFS data in coastal zones, particularly in Case II waters (Morel
and Prieur, 1977) where suspended sediments and CDOM significantly affect the ocean
color signal.
A new algorithm (Carder et al., 1999) proposed for MODIS was also used to
estimate both the chlorophyll-a concentration and CDOM absorption with the remote
sensing reflectance spectra. This algorithm uses the 412 nm band data to distinguish
CDOM from chlorophyll-a pigment because of the strong absorption of CDOM at 412
nm. For validation purposes, I compared in situ chlorophyll-a concentration to
chlorophyll-a concentration derived from satellite data using the OC4v4 and Carder et al
(1999) algorithms.
18
For time series and statistical analyses, AVHRR SST (1997-2000) and SeaWiFS
(October 1997 - December 2000) images were averaged (arithmetic average) into a time
series of weekly means.
Sea surface height (SSH) anomaly fields were obtained from the University of
Colorado (courtesy of Dr. Robert Leben). This was a blended product of TOPEX/
POSEIDON and ERS-2 satellite altimeter data. SSH fields were produced by temporal
and spatial smoothing using decorrelation scales of 12 days and 100 km. Therefore,
features may appear weaker than they actually were, and smaller-scale features may not
be represented. To estimate the total dynamic topography, the residual mean in the SSH
was removed before adding a model mean to produce the synthetic height estimate. The
resulting time series of SSH fields were interpolated to obtain one SSH field per day.
More details about data processing and analysis can be found in Leben et al. (2002). For
comparison with ADCP gridded currents, the SSH field at the middle date of each of the
seven cruises was selected. For weekly time series, each middle date of a week range
(e.g., a week range of 1-7, middle date of the week is 4) for sea surface height was
overlaid over the weekly mean composite of SeaWiFS and AVHRR SST imagery.
To determine the difference between data collected in situ and data estimated by
satellite, the mean relative error (MRE) was estimated. MRE is defined as follows:
N
MRE = [∑((satellitei – in situi)/in situi)]/N,
i=1
(8)
where i is the index for all valid data points and N is total number of valid points. In
order to avoid misinterpretation, data were grouped into two categories i.e., satellitei< in
19
situi (MRE1)and satellitei> in situi (MRE2). Otherwise large positive errors could be
neutralized by any large negative errors, producing small, inaccurate mean errors.
1.3. RESULT AND DISCUSSION
1.3.1. Spatial and Temporal Distribution of Chlorophyll-a Concentration
The spatial and temporal near-surface distribution of chlorophyll-a concentration
for the different time periods based on satellite and in situ observations are shown in
Figure 1.3 and 1.4, respectively. [figures should immediately follow the text – this
applies to all figures and tables] Apparent high chlorophyll-a concentration (>1 mg m-3)
was generally found only inshore particularly near the major river mouths, or offshore in
the Mississippi River plume during summer sampling periods (N3, N6, and N9).
Figure 1.5 and 1.6 show the monthly mean values of chlorophyll-a concentration
derived from SeaWiFS and in situ observation for each region (as defined in Figure 1.1)
and for each season (cruise). [Bisman: be consistent!!! In one figure you use name to
define a region, but in another figure you use R#] Based on SeaWiFS data, mean
chlorophyll-a concentration ranged from 0.1+0.0 mg m-3 in the central NEGOM (region
R8) in July 1998 to 22.8+11.2 mg m-3 off the Mississippi (region R1) in July 1998.
While specific regions showed seasonality in chlorophyll-a concentration, there was no
consistent (synchronized) seasonal pattern for the entire NEGOM.
Figure 1.5 also shows strong interannual variability in most regions, particularly
during winter and spring. The year 1998 was different from 1999 and 2000 and showed
much higher chlorophyll-a concentration as high as a factor of 4 off Escambia (R3),
20
Choctawhatchee (R4), Apalachicola (R5), Suwannee (R6) and Tampa Bay (R7) regions,
particularly between January and May. The anomalous chlorophyll-a concentration
during winter and spring of 1998 appeared to be related to both upwelling [how do you
know it is upwelling?] and anomalously high river discharge. Indeed, Schmidt et al.
(2000) reported that Florida experienced excessive rain in winter and spring 1998 due to
the 1997-1998 El Nino event.
Considering only non-El Nino years (i.e., 1999 and 2000), a seasonal pattern in
chlorophyll-a concentration was found off the Mississippi (R1), Central NEGOM (R8),
Apalachicola (R5), Suwannee (R6) and Tampa Bay (R7) regions. Off the Mississippi
and central NEGOM regions, the maxima occurred in summer and minima in fall. Off
Apalachicola, the maximum occurred at the end of fall and early winter (December,
December, January) and minima in spring (March, May). Off the Suwannee (R6) region,
the maxima occurred during summer months (August, September) and the minima in
winter months (December, February). Off Tampa Bay (R7) region, the maxima occurred
in fall (October) and the minima in spring (March, May) (Figure 1.5). These patterns
were also consistent with in situ observation of chlorophyll-a concentration (Figure 1.6).
1.3.2. Offshore dispersal of the Mississippi Plume in the NEGOM
Based on weekly mean SeaWiFS images, a plume of apparent elevated chlorophylla concentration was seen extending east and southeastward from the Mississippi River
delta each summer in 1998, 1999, and 2000, respectively. The plume normally formed
toward the end of May and lasted until about the first week of September. In 1998, a
21
well-defined plume of 330 km to the ESE was first observed toward the end week of June
from the Mississippi Delta. Apparent chlorophyll-a concentration toward the distal end
of the plume was about 1 mg m-3. The plume reached maximum eastward extent (about
550 km to the ESE of the Mississippi Delta) in late July (Figure 1.7). The plume became
smaller afterward and was effectively absent by mid August.
In 1999, a well-defined eastward offshore NEGOM plume of ~280 km was
observed the last week of May extending ESE from the Mississippi Delta. The plume
reached its maximum extent in late July (~550 km ESE; Figure 1.7). This plume lasted
about 14 weeks, through the first week of September.
22
N9
N8
N7
N6
N5
N4
N3
N2
N1
Figure 1.3. Spatial and temporal distribution of near-surface chlorophyll-a concentration
in the NEGOM region based on SeaWiFS data during the nine separate NEGOM cruise
periods. White and purple contours are positive and negative total sea surface height,
respectively.
23
Figure 1.4. Spatial and temporal distribution of near-surface chlorophyll-a concentration
in the NEGOM region based on in situ observation data. Thin white line indicates the
cruise transect during which continuous flow-through chlorophyll fluorescence
observations were collected. See Table 1.1 for cruise identification.
24
Figure 1.5. Time series of near-surface monthly mean chlorophyll-a concentration based
on satellite SeaWiFS data [what algorithm?] for eight selected regions.
25
Figure 1.6. Near-surface mean chlorophyll-a concentration based on in situ observation
for eight selected regions. See Table 1.1 for cruise identification and Figure 1.1 for
region identification.
26
In 2000, the Mississippi plume was first observed in early July, extending some 220
km ESE off the Mississippi Delta. The plume reached its maximum ESE extent in late
early August (about 580 km from the Mississippi Delta; Figure 1.7). The plume lasted
about 9 weeks and disappeared during the first week of September.
1.3.3. Factors Affecting for Chlorophyll-a Concentration Variability
Previous studies have documented that the nearshore circulation off Mississippi and
Alabama is heavily influenced by wind (Schroeder et al., 1987; Chuang et al., 1982;
Murray, 1972). Golubev and Hsueh (submitted) also found that surface currents on the
shelf west of 85W are correlated with the zonal wind component. The nearshore
pressure gradient generated by river runoff also tends to drive a westward geostrophic
flow near the coast (Schroeder et al., 1987). Walker (1996) and Salisbury et al.
(submitted) found that the pressure gradient due to river discharge, and wind forcing are
the main factors affecting the dispersal of the Mississippi River plume. Other forcing,
such as tidal currents, seem to play a relatively weak role in the nearshore circulation of
the region (Walker, 1996; Murray, 1972).
In this study, I studied the effect of wind (as a coastal upwelling index) and sea
surface height as forcing factors affecting the variability of chlorophyll-a concentration
and the dispersal of river water in the NEGOM.
The Mississippi River discharge data from October 1997 to December 2000 showed
a seasonal cycle, with maximum discharge in spring and minimum in summer. The
apparent chlorophyll-a concentration peak off the Mississippi region occurred several
27
Figure 1.7. Initial and subsequent position of the Mississippi River plume during summer
of 1998, 1999, and 2000. White and purple contour lines indicate positive and negative
sea surface height. White arrow indicates the extent of the plume. [what are the dates]
28
weeks after the maximum in Mississippi discharge (Figure 1.8). This result was also
confirmed by cross correlation analysis. A high positive cross correlation coefficient
(r=0.55) occurred at a positive lag of 6 weeks (Figure 1.9).
A cross-correlation analysis between the upwelling index and chlorophyll-a
concentration in this region also revealed a relatively negative high cross-correlation
coefficient (r=-0.48) at a lag of 0 (zero) weeks. This demonstrated that upwelling, or at
least offshore advection of Mississippi River waters, also played an important role in the
variability of chlorophyll-a concentration off the Mississippi region.
The Alabama, Escambia, Choctawhatchee, Apalachicola, and Suwannee Rivers also
showed a seasonal discharge cycle, with maximum discharge in late winter (February) to
spring, and minimum in summer. Chlorophyll-a concentration off the Alabama,
Escambia, Choctawhatchee, Apalachicola, and Suwannee regions peaked during it’s the
corresponding river discharge maximum (Figure 1.10, 1.11, 1.12, 1.13, 1.14, and 1.15).
These results were confirmed by cross correlation analyses. High maximum positive
cross correlation coefficient off Alabama (r=0.75, lag=1), Escambia (r=0.67, lag=1),
Choctawhatchee (r=0.52, lag=1), Apalachicola (r=0.48, lag=0), and Suwannee (r=0.75,
lag=-1) were found. Meanwhile, cross correlation coefficients between upwelling index
and chlorophyll-a concentration were not significant in these regions (Figure 1.9). These
results demonstrated that changes in the river discharge preceded changes in chlorophylla concentration by a few days for the above regions. The results show that river
discharge is the major controlling factor of the variability of chlorophyll-a concentration
in these inner-shelf regions.
29
Off Alabama, Escambia, and the Choctawhatchee regions, upwelling-favorable
winds were observed intermittently between early spring and late summer. Muller-Karger
(2000) reported strong intermittent upwelling in May-June 1998 in nearshore areas,
generally within 15 km of the coast along the Florida Panhandle. Upwelling-favorable
winds also occur occasionally during winter. SeaWiFS-derived chlorophyll-a
concentrations appeared to be elevated during these periods of upwelling-favorable winds
(Figure 1.10, 1.11, and 1.12). However, variability in the freshwater discharge seemed to
have stronger influence on the variability of chlorophyll-a concentration than the
upwelling process (Figure 1.9, 1.10, 1.11, 1.12). The strength of the upwelling signal was
probably masked in part by the large size of the sampling boxes along the coast (Figure
1.1), since west of Cape San Blas upwelling was largely confined to a narrow strip along
the coast.
Off the Apalachicola, upwelling-favorable winds were frequent between winter and
summer, with the upwelling index reaching 200 m3 s-1 100m-1 in winter 1998 (Figure
1.13). Chlorophyll-a concentrations rose during upwelling periods. However, the
variability of Apalachicola River freshwater discharge (r=0.48, lag=0 week) had stronger
influence on the variability of chlorophyll-a concentration than the upwelling process
(r=0.09, lag=2 week; Figure 1.9 and 1.13). River discharge probably is the dominant
factor controlling chlorophyll-a concentration here in great measure because the sampling
box was placed immediately in front of the mouth of the Apalachicola river (Figure 1.1).
A chlorophyll plume that extends from Cape San Blas to the south parallel to the shelf
break and as far south as the Florida Keys has been observed in this region for many
30
years, each lasting 1-6 weeks during spring (Gilbes et al., 1996). Our observations
confirmed its regular occurrence. Gilbes et al. (1996) speculated that the plume might be
associated with one or a combination of the following processes: (1) discharge from
small, local rivers along the northwest Florida coast; (2) seasonal changes in steric height
differences between the shelf and deep gulf of Mexico waters; (3) circulation of water
associated with the Loop Current and upwelling in the DeSoto Canyon; and (4) discharge
from the Mississippi and Mobile Rivers. Gilbes et al. (2002) have confirmed that the
Apalachicola River has a major role in the formation of offshore blooms. A cold tongue
of water has been detected in AVHRR images in the same area as this chlorophyll plume
during spring. This is independent of the Loop Current position and is a result of
circulation driven by buoyancy and heat flux forcing (Weisberg and He, 2003; He and
Weisberg, 2002).
Off the Suwannee river, only relatively weak upwelling-favorable winds were
observed between late fall and early summer. This was also the situation in 1998, when a
significant increase in chlorophyll concentration was observed in the region. From a
purely statistical point of view, variability in the Suwannee discharge had stronger
influence on the variability of chlorophyll-a concentration (r=0.75, lag=-1 week) than the
wind-driven upwelling processes (r=-0.12, lag=-1 week; Figure 1.9 and 1.14).
In the central NEGOM (R8) region, the relatively high chlorophyll-a concentrations
observed in summer of three consecutive years (1998, 1999, 2000) appeared to be related
to wind direction and speed, to the Mississippi River discharge, and to sea surface height.
Weekly average wind data measured south and east of the Mississippi delta (BURL1,
31
28.9N, 89.4W, and NDBC 42040, 29.2N, 88.2W) recorded eastward wind from the
end of April to the end of September, with average speeds of 0.5 to 6.10 m s-1. It seemed
that a well-defined south or southeastward Mississippi plume developed concomitant
with the migration of a warm core ring (anticyclonic, high sea surface height) into the
vicinity of the Mississippi delta (Table 1.2, see also Figure 1.7). The SSH field maps
constructed for each summer cruise indicated that either small or large anticyclonic
(warm) eddies were present near the slope off the Mississippi delta. SSH ranged from
+11 cm for a mesoscale (~100 km diameter) eddy in summer 2000 cruise (N9) to +35 cm
for a large offshore eddy (centered around 88.5W and 26.5N) in summer 1999 cruise
(N6; eddy largely outside the area shown in Figure 1.17).
A relatively small warm core ring (anticyclonic) located southeast off the
Mississippi delta was also observed throughout the fall season in 1997 and 1998 (see
Figure 1.3 and 1.17). The presence of this warm core ring could advect the Mississippi
River freshwater further eastward (offshore) as it was observed during summer season.
However, the Mississippi plume was not observed in the fall season of 1997 and 1998 as
indicated by SeaWiFS data (Figure 1.3) and in situ data (Figure 1.4). Relatively high
westward wind speed during fall season may inhibit the eastward dispersal of the
Mississippi River freshwater plume. The westward wind speeds located south and east of
Mississippi Delta (BURL1 and 42040) were higher than that of all other seasons with
weekly average of up to 10.54 m s-1 (See Figure 1.18).
32
Table 1.2. Starting and ending period (in week number of 52) of eastward wind
component measured at south of Mississippi delta (BURL1) and east of Mississippi delta
(42040) in relation to sea surface height and well-defined Mississippi plume period. Sea
surface height was extracted from the location of 87.5W and 28.2N.
[NOTE: YOU SHOULD USE TABS TO ALIGN NUMBERS (ALSO INDENTATIONS
OF PARAGRAPHS ETC.]
Eastward wind component
Well-defined offshore Mississippi
BURL1
4204
plume
-----------------------------------------------------------------------------------------------------start end start end
start
maximum
end
1998 (wk # of 52) 18
sea surface height -10
30
+20
18
-10
30
+20
26
+8
30
+20
33
+16
1999 (wk # of 52)
sea surface height
20
-12
34
+4
21
-12
29
-15
22
-10
29
+10
36
-10
2000 (wk # of 52)
sea surface height
21
-7
34
+5
21
-7
35
+7
28
+4
31
+6
36
+6
33
Figure 1.8. Weekly mean time series of wind, upwelling index, river discharge, and
chlorophyll-a concentration off Mississippi region. [R#?] A negative value for upwelling
index indicates an upwelling.
34
Figure 1.9. Lagged cross-correlation analyses for weekly mean river discharge and
upwelling index vs. chlorophyll-a concentration, with lags measured in weeks. [Do these
regions have R#s?]
35
Figure 1.10. Same as Figure 1.8 but off Alabama region.
36
Figure 1.11. Same as Figure 1.8 but off Escambia region.
37
Figure 1.12. Same as Figure 1.8 but off Choctawhatchee region.
38
Figure 1.13. Same as Figure 1.8 but off Apalachicola region.
39
Figure 1.14. Same as Figure 1.8 but off Suwannee region.
40
Figure 1.15. Same as Figure 1.8 but for central NEGOM region.
41
1.3.4. In situ Observation for Offshore NEGOM Plume
During summers of three consecutive years (1998, 1999, 2000), salinity in the
offshore NEGOM sample region (R8) was lower than that measured off Alabama,
Escambia, Choctawhatchee, Apalachicola, Suwannee and Tampa Bay (Figure 1.16).
This was a result of the offshore dispersal of the Mississippi River discharge. [align your
text!]
To facilitate understanding of the motion of the offshore Mississippi freshwater
plume during summers, I examined the ship-mounted ADCP data, the SSH field and
surface wind field products. Satellite altimeter data are usually considered of limited
value over continental shelves because of poor knowledge of the geoid and because of
short time and space scales that may not be resolved by the sampling frequency of
present-day altimeters. Figure 1.17 suggested the presence of two distinct sub-surface
circulation regimes in the NEGOM, namely one to the west and one to the east of 85W
(Cape San Blas). To the west of 85W, there were significant changes in the SSH field
from season to season. To the east of 85W, a region of low (negative) sea surface height
was almost always found near the coast. Golubev and Hsueh (submitted) found that the
currents on the shelf to the west and to the east of 85W of the NEGOM had different
origins. From drifter path data, wind and SSH, they concluded that the monthly mean
surface currents on the shelf east of 85W were well correlated with sea surface height
anomalies in Florida Strait. In contrast, monthly mean surface currents on the shelf west
42
of 85W were correlated with zonal wind component, indicating that alongshore winds
drive alongshore shelf currents (see also Muller-Karger, 2000). Schroeder et al. (1987)
and Chuang et al. (1982) also reported that the nearshore circulation off eastern
Mississippi and Alabama was strongly wind dependent, in spite of the potential for
westward geostrophic flow generated by coastal fresh water input.
A well-defined offshore NEGOM plume, however, was evident after several weeks
of eastward wind and the present of a warm core ring (anticylonic) near the slope off the
Mississippi delta (Table 1.2). The anticyclonic eddy located near the slope off the
Mississippi delta entrained and transported low salinity and high nurtients of surface
water from the Mississippi River to offshore (see also Qian et al, 2003; Jochens et al.,
2002; Belabbassi, 2001). The ADCP currents at 10-14 m depth documented
northeastward and eastward (N9), or eastward and southeastward (N3, N6) flow along the
northern edge of the anticyclonic eddy, with speeds of up to 0.6 m s-1. From these ADCP
sub-surface data, it was clear that sub-surface current speed correlated well with the slope
of the eddy with higher eddy slope correspond to higher sub-surface currents.
These
vectors help explain the chlorophyll-a concentration and salinity patterns observed.
Eastward dispersal of the Mississippi plume was also recorded during summer of 1993
(Walker et al., 1994), summer of 1997 (Biggs et al, 2000), and in many other instances
(see Del Castillo et al., 2000 and 2001; Gilbes et al., 1996 and 2002; Muller-Karger,
2000; Muller-Karger et al., 1991).
In winter cruises of 1997 (N1) and 1998 (N4), SSH of +8 cm and +13 cm were seen
off the Mississippi delta (Figure 1.17). Although sub-surface currents indicated eastward
43
flow, there was no indication of eastward dispersal of the Mississippi River water. This
evidence can be seen from the relatively low chlorophyll-a concentration and high
salinity near the slope off the Mississippi delta (see Figure 1.3, 1.16 and 1.17). Based on
weekly analyses of sea surface height dynamic and SeaWiFS images, an anticyclonic
eddy and low chlorophyll-a concentration were also observed off the Mississippi delta
throughout fall season in 1997 and 1998.
Relatively high westward wind speed
component during fall season (at speeds up to ~10.5 m s-1) may have helped keep the
Mississippi water inshore and to the west of the delta (Figure 1.18).
44
Figure 1.16. Near-surface salinity distribution in the NEGOM. Thin white line indicates
the cruise transect during which continuous flow-through salinity observations were
collected. See Table 1.1 for cruise identification.
45
Figure 1.17. Sea surface height (SSH) dynamic and ADCP-generated (10-14 m-deep)
currents in the NEGOM. Note: There are no ADCP data from cruise N4 east of 87W
because of mechanical damage to ADCP equipment during the cruise N4. See Table 1.1
for cruise identification.
46
Figure 1.18. Average surface wind field in the NEGOM during the period of 9 cruise
observations (data obtained from Texas A&M University). Diamonds and their
identification indicate stations where wind speed and direction were collected. Several
offshore NDBC stations used to generate wind gridded average were not shown in the
picture because they were located south of 27N.
47
1.3.5. Nutrients Sources for Offshore NEGOM plume
The Mississippi plume seen in the eastern NEGOM and Gulf every summer of 1998
through 2000 can last over 14 weeks and extend over 550 km from the delta and beyond.
The question is whether it contains sufficient nutrients to support the high chlorophyll-a
concentrations seen. To answer this question, I examined mixed layer depth, studied the
depth of the 34.5 isohaline, and estimated the nutrient content of the 0-20 m and >20-75
m-deep intervals. The nutrient content in these layers follows the rationale of Belabbassi,
2001.
Figure 1.19 shows the mixed layer depth distribution for each season (cruise)
determined with density criterion along the river plume and temperature criterion for non
plume region). The mixed layer depth for spring and summer cruises could be as low as
3 m-deep. The results indicated that stratification of the water column increased in spring
and maximum in summer (as solar radiation input increases). [Bisman: you need to read
your own thesis thoroughly to correct grammar, typo, text alignment, figure placement
etc.] The relatively shallow mixed layer depth during spring and summer may be related
to the buoyancy factor of freshwater river discharge, relatively weak wind speed, and
increase of solar radiation input. Deeper mixed layer depths were observed during winter
of up to 80 m-deep. Heat loss and relatively high wind speed during winter may
contribute to the deepening mixed layer depth. Shoaling mixed layer depth during
48
spring and summer seasons also caused pycnocline to be uplifted and hence brought
deeper water up which rich of nutrients.
To estimate the depth of the river plume, a salinity of 34.5 criterion was used as
salinity boundary that affected by river discharge. Figure 1.20 describes depth
distribution of salinity 34.5 for each season (cruise). Deeper salinity 34.5 was observed
near the coast and broad offshore areas during spring and summer seasons. Depth
distribution pattern of salinity 34.5 was similar to the pattern of near-surface salinity
distribution. It appeared that the 34.5 isohaline exceeded the MLD during spring and
summer along the river plume indicating freshwater mix beyond the MLD. From the
34.5 isohaline distribution, it appeared that the maximum depth was about 20 m.
Therefore, subjectively, I choose criterion of 0-20 m-deep as the layer affected by plume
and >20-75 m-deep as the layer un-affected by plume and considered as bulk of euphotic
zone. These two layers were used to calculate nutrients availability. Both nutrients for
the depth of 0-20 m and >20-75 m were vertically averaged.
Figure 1.21 shows the distribution of average nutrients (NO3+NO2+NH4+Urea)
concentration for the water column of 0-20 m-deep. High nutrients concentration (as
high as 61 mol l-1, during N2 cruise) was only observed near the mouth of the
Mississippi River. Outside near the mouth of the Mississippi River, relatively low
nutrients concentration was observed. From in situ near-surface salinity and chlorophylla distributions as well as from SeaWiFS data, it was clear that a well-defined offshore
NEGOM plume was observed during summer season of the year of 1998, 1999, and
2000. The nutrients distribution of 0-20 m water column, however, did not reveal such
49
offshore plume. Nutrients from river discharge may be well consumed by phytoplankton.
In particular for spring 1998 (N2), relatively high nutrients concentration was observed
inshore in the northern part of the NEGOM from the Mississippi to Suwannee region
(Figure 1.21). This result was attributed to the anomalous northeastward upwelling
favorable winds along the northern portion of the NEGOM shelf (Figure 1.18). The
magnitude and longer lasting periods of upwelling favorable winds was observed in
spring 1998 (N2) relative to spring 1999 (N5) and spring 2000 (N8) (see also MullerKarger,2000; Weisberg and He, 2003).
Figure 1.22 shows the distribution of average nutrients (NO3+NO2+NH4+Urea)
concentration for the water column of >20-75 m-deep. Relatively high nutrient
concentrations were found near the 100 m isobath in the NEGOM region during summer
cruises (N3, 6, and 9) and during winter cruises of 1997, 1998 (N 1 and 4) and spring
cruises of 1998, 1999 (N2 and 5). These results indicated that during late spring and
summer, upwelling played a major role to provide nutrients for phytoplankton over the
shelf. This signal may, however, be obscured from the satellites.
It seemed that high nutrient patterns were always associated with upwelling.
Upwelling along the head of DeSoto Canyon during winter 1997-1998 was consistent
with the finding by Hsueh and Golubev (2002). They discovered a strong upwelling at
the head of DeSoto Canyon during the southward wind bursts in winter 1997-1998. The
current velocity around the DeSoto Canyon was in general eastward. After southward
wind bursts, the pressure gradient excess accelerated flow onshore across the shoaling
topography, leading to upwelling. Upwelling could also be generated by bottom Ekman
50
flow under the strong clockwise circulation around DeSoto Canyon after southward wind
bursts (Hsueh and Golubev, 2002). Weisberg and He (2003) also stated that cyclonic
eddy particularly at the 50 m depth in the DeSoto Canyon was often attributed to the
Loop Current. The canyon itself could generate its own eddy via potential vorticity
conservation independent of the Loop Current (Weisberg and He, 2003). A broad area of
high nutrients concentration along the midshelf of West Florida Shelf in spring 1998 (N2)
was observed (Figure 1.22). The broad area of high nutrients concentration was
consistent with the finding of Weisberg and He (2003) that massive upwelling was
observed due to combination local forcing (wind and heat flux) and anomalous Loop
Current influence.
There is no apparent relationship between the MLD (Figure 1.19) and nutrients
below 20 m. However, it seems that there is a direct correlation between the presence of
anticyclonic eddy located off the Mississippi delta and high nutrients below 20 m over
the mid-shelf (the 100 m isobath) during fall 1997 (N1), fall 1998 (N4), spring 1998
(N2), and summer 1998, 1999, 2000 (N3, N6, N9) cruises (see Figure 1.17and 1.22).
These results indicated that there is an anticyclonic eddy interacting with the NEGOM
shelf break.
51
Figure 1.19. Mixed layer depth distribution in the NEGOM. White dots indicate CTD
deployment stations. See Table 1.1 for cruise identification.
52
Figure 1.20. Depth of salinity 34.5 distribution as a boundary of fresh water influence
from oceanic water in the NEGOM. See Table 1.1 for cruise identification.
53
Figure 1.21. Nutrients (NO3+NO2+NH4+Urea) mean concentration for 0-20 m-deep of
water column in the NEGOM. See Table 1.1 for cruise identification.
54
Figure 1.22. Nutrients (NO3+NO2+NH4+Urea) mean concentration for 0-20 m-deep of
water column in the NEGOM. See Table 1.1 for cruise identification.
55
[in the summary or abstract you said river plume provides more nutrients than upwelling
– can you prove it here?]
1.3.6. Comparison between in situ and SeaWiFS chlorophyll-a concentration
Figure 1.23 shows the in vivo chlorophyll-a concentration estimated from the in situ
flow-through measurements and the SeaWiFS (OC4v4 algorithm) estimates along the
ship tracks of all nine NEGOM cruises. To improve the spatial coverage and reduce
noise, SeaWiFS images were averaged over the time period of each of the nine cruises.
Only SeaWiFS chlorophyll-a concentration data between 0.01 mg m-3 and 1.0 mg m-3
were included. [why do you restrict it to < 1.0?] The correlation between in situ and
satellite measurements, and the mean relative errors (MRE, see section 1.24 for the
formula) are shown in the Figure 1.23 and 1.24.
The satellite estimates generally agreed well with the in situ data for fall/winter
(N1, N4, N7), when the Mississippi plume influence was minimal. The MRE for these
three cruises was generally within +35% (Hooker et al., 1992). Meanwhile, for spring
and summer, SeaWiFS overestimated chlorophyll-a concentration. The positive MRE for
spring and summer cruises ranged from 57.89% to 140.20% (Figure 1.23).
Using the OC2 algorithm to estimate chlrophyll-a concentration from satellite, Hu
et al. (2003) found that MRE for the first six cruises was generally within +35%. This
56
suggests that the OC4v4 algorithm leads to larger errors in estimating chlorophyll-a
concentration in the region of high influence of freshwater discharge.
Some factors that may lead to a divergence between satellite and in situ
measurements are (Hu et al. 2003) the time difference between satellite and in situ
measurements, which is further masked by the average of SeaWiFS data over the cruise
period. It is clear that the SeaWiFS chlorophyll algorithm was affected by colored
dissolved organic matter (CDOM) absorption of light. Further, the atmospheric correction
used over turbid coastal water may have led to additional errors in estimating
chlorophyll-a concentration. Finally, the response of the in vivo fluorescence and
instrument in the river plume may be in error.
The MODIS algorithm proposed by Carder et al. (1999) to estimate chlorophyll-a
concentration as well as CDOM absorption from satellite data was also used and results
compared with in situ data (Figure 1.24). Even though the MRE is still above +35%,
there was better agreement between in situ and MODIS chlorophyll estimates vs. OC4v4
chlorophyll estimates, particularly for spring and summer cruises. Meanwhile, the MRE
using MODIS algorithm during fall seasons was generally within +35%, which agreed
with the SeaWiFS mission specification.
Comparisons between in situ and SeaWiFS for larger estimates chlorophyll-a
concentration for both OC4v4 and MODIS algorithm were also performed. In general,
positive MRE were consistently higher with higher chlorophyll-a concentrations.
Relatively large errors between in situ and satellite chlorophyll-a concentration estimates
using both OC4v4 and MODIS algorithms were most apparent in the region of freshwater
57
influence. Clearly, more research is needed to properly separate CDOM and bottom
reflectance in the algorithms.
58
Figure 1.23. Comparison between in situ and SeaWiFS estimates (OC4v4) of
chlorophyll-a concentration along ship track lines for nine NEGOM cruises. MRE1 is
negative MRE and MRE2 is positive MRE (see section 1.2.6 for the formula).
59
Figure 1.24. Same as Figure 1.23, but the SeaWiFS chlorophyll-a concentration was
estimated with the MODIS algorithm (Carder et al., 1999).
1.4. Conclusions
60
Analyses of weekly ocean color and altimeter satellite imagery and in situ data
collected between October 1997 and December 2000 showed that high chlorophyll-a
concentration (>1 mg m-3) occurred over the inner shelf of the NEGOM and offshore
during summers of 1998, 1999, and 2000. In each of these years, the NEGOM featured
one or more anticyclonic eddies which advected river water offshore into the southeastern
Gulf in late spring and summer. While off Mississippi (R1), Apalachicola (R5),
Suwannee (R6), Tampa Bay (R7) and offshore NEGOM (R8) showed seasonality in
chlorophyll-a concentration, there was no consistent (synchronized) seasonal pattern for
the entire NEGOM.
While upwelling is an important factor in providing nutrients to the phytoplankton,
the variability of river discharge appeared to be the major controlling factor in the
variability of chlorophyll-a concentration off Alabama, Escambia, Choctawhatchee,
Apalachicola and Suwannee regions. [prove it with numbers] Meanwhile, the Mississippi
river discharge and upwelling acted together to control the variability of chlorophyll-a
concentration off the Mississippi region.
A strong interannual variability of chlorophyll-a concentration in winter and spring
between 1998 compared with 1999 and 2000 was observed off Escambia (R3),
Choctawhatchee (R4), Apalachicola (R5), Suwannee (R6) and Tampa Bay (R7). Mean
chlorophyll-a concentrations during winter and spring 1998 were anomalously higher up
to about a factor of four than those of 1999 and 2000. The high concentrations related to
anomalously high river discharge in connection with 1997-1998 El Nino event.
61
A well-defined offshore low salinity and high chlorophyll-a concentration plume
was observed each summer of 1998, 1999, and 2000. The plume was observed as early
as late May and lasted as long as the first week of September. The plume extended up to
>500 km east of the Mississippi delta and to Florida Keys. The plume initiation was
facilitated by eastward wind, while the offshore plume development was facilitated by an
anticyclonic eddy. Each of these years, an anticyclonic eddy was observed in the
NEGOM near the slope off the Mississippi delta.
Relatively low surface (0-20 m-deep) nutrient concentrations were observed at all
times in the NEGOM. In the 20-75 m depth interval, relatively high nutrients
concentrations were found on the shelf during late spring and summer cruises. This
indicated strong upwelling during this time relative to other periods.
Comparison between in situ and OC4v4 SeaWiFS chlorophyll estimates indicated
that satellite estimates generally agreed to +35% with in situ data for fall/winter seasons
(N1, N4, N7). In spring and summer seasons, SeaWiFS overestimated in situ
measurements. Significant improvement compared with OC4v4 particularly for spring
and summer cruises was observed by using the Carder et al. (1999) algorithms to
estimate satellite chlorophyll-a concentration. OC2 algorithm (Hu et al., 2003) performed
better in estimating chlorophyll-a concentration than that of OC4v4.
62
CHAPTER II:
COLORED DISSOLVED ORGANIC MATTER IN THE SURFACE WATERS
OF THE NORTHEASTERN GULF OF MEXICO
63
Abstract
Colored dissolved organic matter (CDOM) fluorescence and blue light absorption
coefficient at 443 nm (ag443) were measured, along with salinity and chlorophyll
fluorescence, in surface waters of the Northeastern Gulf of Mexico (NEGOM) during
seven seasonal cruises conducted in 1998-2000. Strong correlations between the ag443
and CDOM fluorescence (330 nm excitation and 450 nm emission filters) were
consistently observed. Relatively high ag443 (>0.1 m-1) occurred near major river mouths,
but decreased rapidly with distance from shore except within river waters entrained by
oceanic currents and transported offshore during summer periods. In general, ag443
covaried linearly and inversely with salinity in nearshore areas during spring and winter
cruises, indicating conservative mixing. In outer shelf and slope regions, such correlation
was observed only during summer cruises when offshore entrainment of plume waters
occurred. There was no indication that a local increase of chlorophyll concentration
would lead to increased CDOM, except in the vicinity of Tampa Bay during summer
cruises. The “global NEGOM” winter 1998 (N4) salinity-ag443 relationship i.e., Salinity
(S)=36.59-29.86*ag443 (n=8771, r2=0.86, 0.01<ag443<0.52, 16<S<36) produced the lowest
mean percentage errors of observed vs. predicted salinities (<3% errors, correspond to
<1.04 in salinity value) for all other spring and winter seasons, therefore, this relationship
may be useful to estimate salinity from satellite ocean color data. However, estimates
with this relationship led to unacceptable errors in the summer (>10% errors in salinity
estimates), due to less of a coherent relationship between ag and salinity over the range of
salinities observed.
64
2.1. Introduction
Colored dissolved organic matter (CDOM) originates from the degradation of
organic material in terrestrial and aquatic ecosystems (Kirk, 1994). In coastal waters,
most CDOM typically derives from land drainage and river runoff. Away from
continental margins the effect of rivers declines, and CDOM is presumably related to
primary productivity as a by-product of algal cell degradation (Siegel and Michaels,
1996; Carder et al., 1989; Yentsch and Reichert, 1962; Fogg and Boalch, 1958) and
zooplankton grazing (Momzikoff et al., 1994). In upwelling areas, an increase in nutrient
availability may lead to an increase in chlorophyll concentration exceeding two orders of
magnitude. While historical observations have suggested that there may be little change
in CDOM concentration in these areas affected by upwelling (Bricaud et al., 1981) there
is evidence that, in some regions at least, upwelling source water itself may be colored by
higher background CDOM concentrations (Coble et al., 1998; see also Siegel et al.,
2002).
Knowledge of the spatial distribution and temporal changes of CDOM on the
continental shelf is important to help understand biogeochemical processes there.
Specifically, photolysis of CDOM is a pathway of carbon cycling in the ocean (Moran
and Zepp, 1997). CDOM strongly absorbs light in the biologically damaging ultraviolet
(UV) region, thus protecting phytoplankton and other biota (Arrigo and Brown, 1996;
Blough and Green, 1995). CDOM also plays an important role in light absorption in the
blue region of the spectrum, reducing the light available for phytoplankton
photosynthesis. Indeed, its absorption characteristics cause CDOM to degrade the
65
accuracy of satellite-derived chlorophyll concentrations, since these typically are
estimated with very simple empirical relationships (Morel and Prieur, 1977; Hochman et
al., 1995, 1994; Vodacek et al., 1994; Carder et al., 1989).
Since CDOM and pollutants transported in river plumes are often diluted in a
manner consistent with that of salinity, CDOM has been proposed as a proxy to estimate
surface salinity from high-resolution (~<1km) ocean color sensors (e.g., Hu et al., 2003;
D’Sa et al., 2002). Surface salinity is a conservative property (assuming balanced
precipitation/evaporation), and therefore it can be used to trace transport and mixing
processes in estuaries and coastal areas. Salinity is also frequently used as an index of the
health of coastal habitats (e.g., Montague and Ley, 1993). It would be of great utility to
establish a methodology to assess synoptic salinity distribution rapidly and repeatedly.
Unfortunately, planned space-based microwave sensors provide only coarse spatial
resolution (~80 km) and limited accuracies for instantaneous salinity estimates (~0.4-0.6
psu; Lagerloef et al., 1995). It would be ideal to complement such capabilities with
higher resolution information. The potential use of CDOM as a salinity proxy is based on
the assumptions that 1) CDOM can be estimated accurately from satellites and 2) there
exists a predictable relationship between CDOM and salinity.
Several studies have documented the relationship between CDOM and surface
salinity in coastal regions (Hu et al., 2003; Benner and Opsahl, 2001; Klinkhammer et
al., 2000; Ferrari and Dowell, 1998; Vodacek et al., 1997; Carder et al., 1999). Although
an inverse linear relationship is often found because of conservative mixing (Hu et al.,
66
2003; Ferrari and Dowell, 1998; Carder et al., 1999), non-conservative relationships have
also been reported (Blough et al., 1993; Benner and Opsahl, 2001).
The Northeastern Gulf of Mexico (NEGOM, Figure 1.1) receives significant fresh
water input from the Mississippi, Alabama, Escambia, Choctawhatchee, Apalachicola,
and the Suwannee Rivers. River discharge varies seasonally, with maximum typically
during spring and minimum during summer (Gilbes et al., 1996). The discharge leads to
strong seasonal buoyancy variation on the shelf, and to high nutrient, chlorophyll, and
dissolved organic carbon concentrations nearshore (Del Castillo et al., 2000; Guo et al.,
1995; Pennock et al., 1999; Turner and Rabalais, 1999; Gilbes et al., 1996). Del Castillo
et al. (2000) concludes that the CDOM in the West Florida Shelf is directly related to
river input, and that high phytoplankton concentrations did not automatically result in the
production of significant amounts of CDOM.
The NEGOM coast also experiences strong mixing, advection, and upwelling
processes (Muller-Karger, 2000; Muller-Karger et al., 1991; Gilbes et al., 1996). It is
sometimes impacted by northward intrusions of the Loop Current (LC) and by cyclonic
(cold) and anticyclonic (warm) eddies that may spin up along the slope and shelf after LC
intrusion (Vukovich and Hamilton, 1989; Jochens et al., 2002).
Several studies on CDOM characteristics and distribution in the NEGOM region
have been conducted (Del Castillo et al., 2001; Del Castillo et al., 2000; Carder et al.,
1989). Yet, these focused on single cruises and not on the understanding of the spatial
and temporal distribution of CDOM, or on factors controlling this distribution. Further,
67
little information has been available in the literature regarding the relationship between
CDOM and salinity, and factors affecting the relationship, in this region.
This study uses field observations from seven cruises conducted between 1998 and
2000 to examine the spatial and temporal variability of CDOM and the relationship
between surface salinity and CDOM. One goal was to determine whether surface salinity
in the NEGOM can be estimated from CDOM, specifically because this is relevant to the
possibility of using ocean color to assess the distribution of salinity in areas affected by
rivers.
2.2. Methodology
2.2.1. Study Area
The NEGOM extends from the Mississippi River Delta to the West Florida Shelf off
Tampa Bay. For our purposes, it is defined inshore by the 10-m isobath and offshore by
the 1000-m isobath. Six major river-impacted regions, namely the Mississippi (R1),
Alabama (R2), Escambia (R3), Choctawhatchee (R4), Apalachicola (R5), Suwannee
(R6), and Tampa Bay (R7), plus one offshore region (R8) were chosen to characterize the
NEGOM (Figure 1.1).
2.2.2. Data Collection, Processing and Analysis
Seven two-week cruises were conducted in three different seasons between 1998
and 2000 onboard the Texas A&M University R/V Gyre (Table 1.1): spring (April or
May), summer (July and August), and winter (November). Each of the cruises surveyed
eleven cross-margin transects from the 10-m to the 1000-m isobath (Figure 1.1). Detailed
68
description of data collection and analysis methods can be found in Hu et al., (2003). A
brief summary is given below.
Continuous near-surface salinity, chlorophyll, and CDOM fluorescence were
recorded every 2 minutes along the cruise transects with Sea-BirdTM and Turner
DesignsTM sensors arranged so that near-surface water (3 m) flowed through all sensors at
a rate of 10 liters per minute (we refer to this arrangement as a flow-through system).
Water was pumped into a de-bubbler and mixing chamber of 10-liter volume with 1minute residence time. Chlorophyll and CDOM fluorescences were measured with 440
and 330 nm excitation and 683 and 450 nm emission filters, respectively.
At selected stations, water samples from the flow-through outflow were filtered
using 0.2 μm Millipore filters into amber-colored glass bottles precombusted at 500C for
24 h. Samples were stored in a freezer for later analysis of the CDOM (Gelbstoff)
absorption coefficient (ag). The water absorption spectra were analyzed using a Hitachi
U-3000 Double Beam Spectrophotometer (10-cm pathlength). Absorption coefficients
were determined using the equation: ag()=(A()*2.303)/L where, A() is the absorbance
spectra and L is the cuvette length (m). The ag spectra were also smoothed to eliminate
instrument noise and corrected for residual scattering by subtracting residual absorption
at 700 nm from the entire spectra.
The flow-through CDOM fluorescence data were converted to ag443 (i.e., ag at 443
nm) using 30 discrete samples collected during each cruise and following the method of
Hu et al., (2003). This ensured coverage of various water types and a broad dynamic
range.
69
Continuous subsurface (10-14 m deep) currents were measured using a shipboard
153 kHz Acoustic Doppler Current Profiler (ADCP). To produce circulation maps, the
ADCP data were gridded employing 40 km as the smoothing radius (Daley, 1991).
Features with scales less than 40 km are therefore likely smoothed out, but this allowed
filtering out noise features and keeping the basic large-scale patterns. Details of these
measurements and data processing can be found in Jochens et al. (2002) and Wang et al.
(2002).
Sea surface height (SSH) anomaly fields were obtained from the University of
Colorado (courtesy of Dr. Robert Leben). This was a blended product of the
TOPEX/POSEIDON and ERS-2 satellite altimeter. The SSH anomaly fields were
produced by temporal and spatial smoothing using decorrelation scales of 12 days and
100 km. Therefore, features may appear weaker than they actually were, and smallerscale features may not be represented. To estimate the total dynamic topography, the
residual mean in the SSH anomaly was removed before adding a model mean to produce
the synthetic height estimate. The resulting time series of SSH fields were interpolated to
obtain one SSH field per day. More details about data processing and analysis can be
found in Leben et al. (2002). For comparison with ADCP gridded currents, we selected
the SSH field at the middle date of each of the seven cruises.
70
2.3. Results and Discussion
2.3.1. Spatial and Temporal Distribution of CDOM and Salinity
The near-surface distribution of the CDOM absorption coefficient at 443-nm
(ag443) for the different sampling periods is shown in Figure 2.1. Relatively high ag443
(>0.1 m-1) was generally found only inshore near the major river mouths, or offshore in
the Mississippi River plume during summer sampling periods (cruises N3, N6 and N9) .
The spatial and temporal distribution of near-surface salinity closely matched that
of ag443 (see Figure 1.16). Low salinity (<34) was generally observed over the shelf near
river mouths, but was also found in outer shelf and offshore waters during summer
(cruises N3, N6, and N9).
The results clearly show the influence of freshwater discharge on the distribution of
CDOM concentration and the importance of the Mississippi plume as a source of
terrigenous CDOM to offshore waters of the Gulf of Mexico. To facilitate understanding
of the motion of the high CDOM, low salinity Mississippi River water to the southeast
along the West Florida Shelf during summers, we examined the ship-mounted ADCP
data and the SSH field product. Satellite altimeter data are usually considered of limited
value over continental shelves because of poor knowledge of the geoid and because of
short time and space scales that may not be resolved by the sampling frequency of
present-day altimeters.
Figure 1.17 suggested the presence of two distinct surface circulation regimes in
the NEGOM, namely one to the west and one to the east of 85W (Cape San Blas). To
the west of 85W, there was a significant difference in SSH field from season to season.
71
To the east of 85W, a region of low (negative) sea surface height was almost always
found near the coast. Golubev and Hsueh (submitted) also found that the currents on the
shelf to the west and to the east of 85W of the NEGOM had different origins. From
drifter path data, they concluded that the monthly mean surface currents on the shelf east
of 85W were well correlated with sea surface height anomalies in Florida Strait,
reflecting an offshore pressure forcing that drives a barotropic flow on the west Florida
shelf on a nearly annual basis. In contrast, monthly mean surface currents on the shelf
west of 85W were correlated with zonal wind component, indicating that alongshore
winds drive alongshore shelf currents (see also Muller-Karger, 2000). Schroeder et al.
(1987) and Chuang et al. (1982) also reported that the nearshore circulation off eastern
Mississippi and Alabama was strongly wind dependent, in spite of the potential for
westward geostrophic flow generated by coastal fresh water input.
The SSH field maps constructed for each summer cruise indicated that either small
or large anticyclonic (warm) eddies were present near the slope off the Mississippi delta.
SSH ranged from +11 cm for a mesoscale (~100 km diameter) eddy in summer 2000
(cruise N9) to +35 cm for a large offshore eddy (centered around 88.5W and 26.5N) in
summer 1999 (N6; eddy largely outside the area shown in Figure 1.17). These
anticyclonic eddies entrained and transported low salinity, high CDOM surface water
from the delta area offshore (see also Qian et al, 2003; Jochens et al., 2002; Belabbassi,
2001). The ADCP currents at 10-14 m depth documented northeastward and eastward
(N9), or eastward and southeastward (N3, N6) flow along the northern edge of the
anticyclonic eddies, with speeds of up to 0.6 m s-1. These vectors help explain the ag443
72
and salinity patterns observed. Eastward dispersal of the Mississippi plume was also
recorded during summer of 1993 (Walker et al., 1994), summer of 1997 (Biggs et al,
2000), and in many other instances (see Del Castillo et al., 2000 and 2001; Gilbes et al.,
1996 and 2002; Muller-Karger, 2000; Muller-Karger et al., 1991).
Some cyclonic features appeared along the coast west of 85º W during summer
(see Figure 1.17). It is unclear what these features represent, i.e. whether they are real
cyclonic eddies or artifacts in the altimeter data.
In winter 1998 (N4), SSH of +13 cm was seen off the Mississippi Delta (Figure
1.17). This was associated with relatively high salinity and low CDOM, indicating
limited or no eastward dispersal of the Mississippi River water. Jochens et al. (2002) also
noted that the wind in fall and winter 1998 and 1999 on average blew toward the west
and southwest, at speeds up to ~10 m s-1, which may have helped keep the Mississippi
water inshore and to the west of the delta.
Figure 2.2 shows the mean values of ag443, salinity, and chlorophyll concentration
for each region (as defined in Figure 1.1) and for each season (cruise). Mean ag443 ranged
from 0.011+0.001 in the offshore NEGOM (region R8) during spring 1999 (cruise N5) to
0.201+0.065 off the Mississippi (region R1) during summer 2000 (cruise N9). While
specific locations showed seasonality in ag443 within the NEGOM, there was no
consistent (synchronized) seasonal pattern for the entire NEGOM. Seasonality was
strongest off Alabama (R2), Tampa Bay (R7), and in the central offshore NEGOM (R8).
Off Alabama (R2), the maxima were found during spring (N5, N8) and the minima
during summer and winter (N4, N6). Off Tampa Bay (R7), maxima occurred in winter
73
and minima during spring. In offshore NEGOM (R8), maxima occurred in summer and
minima during spring. If all locations are considered, a one-way analysis of variance
(ANOVA) showed that there are significant differences in the mean ag443 values among
seasons for each location (P<0.01). The mean ag443 values are also different (P<0.01)
among locations within each season (cruise). Variability of mean ag443 was higher among
locations than among seasons.
Mean salinity ranged from 28.59+1.43 off the Mississippi (R1) during summer 1999
(N6) to 36.70+0.12 off Tampa Bay (R7) during summer 2000 (N9) (Figure 2.4, middle
panel). Similar to ag443, spatial variability in salinity was higher than seasonal variability.
Seasonal variability was observed off the Mississippi (R1), Alabama (R2), and in the
central NEGOM (R8). Waters off Escambia (R3), Choctawhatchee (R4), Apalachicola
(R5), Suwannee (R6), and Tampa Bay (R7) experienced relatively low seasonal
variability. Similar to ag443, there was no clear overall seasonal pattern.
Mean chlorophyll concentration in the central offshore NEGOM (R8) ranged from
0.09+0.01 (mg/m3) during spring 2000 (N8) to 5.24+2.59 (mg/m3) in the Mississippi
region (R1) during spring 1999 (N5) (Figure 2.2, lower panel). As with ag443 and salinity,
spatial variability of surface chlorophyll was higher than seasonal variability.
74
Figure 2.1. Near-surface ag443 (m-1) distribution in the NEGOM. Thin white line indicates
the cruise transect during which continuous flow-through CDOM fluorescence
observations were collected. See Table 1.1 for cruise identification.
75
0.34
N3
N4
N5
N6
N7
N8
N9
0.32
0.30
0.28
0.26
ag443 (1/m)
0.24
0.22
0.20
0.18
0.16
0.14
0.12
0.10
0.08
0.06
0.04
0.02
0.00
R1
R2
R3
R4
R5
R6
R7
R8
R7
R8
Region
38
N3
37
N4
N5
N6
N7
N8
N9
36
Salinity
35
34
33
32
31
30
29
28
R1
R2
R3
R4
R5
R6
Region
Chlorophyll Conc. (mg/m^3)
8.0
N3
7.0
N4
6.0
N5
5.0
N6
N7
4.0
3.0
N8
2.0
N9
1.0
0.0
R1
R2
R3
R4
R5
R6
R7
R8
Region
Figure 2.2. Mean and standard deviation of near-surface ag443 (m-1) (top panel), salinity
(middle panel) and chlorophyll concentration (bottom panel) values for each subregion
and for different time periods sampled. See Table 1.1 for cruise identification.
76
2.3.2. CDOM versus Salinity
Figure 2.3 shows salinity-ag443 scatter plots within each of the selected sampling
regions organized by date (cruises). Statistics are provided in Table 2.1. In general, ag443
covaried linearly and inversely with salinity. Using ANOVA, there were also significant
differences among slopes of salinity-ag443 regressions, both among regions within each
season, and among seasons within each region (P<0.01). For nearshore regions, mostly
in spring and winter seasons, high coefficient of determinations (r2>0.70 to r2=0.98) were
found, indicating conservative mixing behavior. Some low (r2<0.70) to non-significant
(r2=0.00) coefficient of determinations were also found but mostly in summers, for
example in the Mississippi river region (R1) during summer 1999 (N6) when effectively
no high salinity waters were found within this sample region.
The non-conservative CDOM mixing curves found during most of the summertime
cruises in the nearshore (low salinity) regions suggest that destructive (removal)
processes (e.g. CDOM photo-degradation, consumption) may play a role. The removal
processes are apparent as concave salinity-ag443 curves. Strong thermal and salinity
stratification during summertime may also enhance rates of CDOM photolytic decay in
near-surface waters (Vodacek et al., 1997). A non-conservative behavior with net
removal of dissolved organic carbon in the Mississippi River plume was also observed
during summer by Benner et al. (1992).
The central offshore NEGOM (R8) showed low to no correlation between salinity
and ag443 during winter and spring seasons. High negative correlation and steep negative
slopes were observed in this region during summer (Table 2.1, Figure 2.3). The high
77
correlation was due to the east and southward extension and transport of the Mississippi
River plume during these summer periods (Hu et al., 2003; Del Castillo et al., 2001;
Nowlin et al., 2000; Muller-Karger, 2000, and others).
Figure 2.3. Scatter plots of salinity vs ag443 for each region in different seasons (cruises).
See Table 1 for cruise identification. Yellow line is the global NEGOM regression line
for Winter 1998 (N4), yielding the lowest error of observed vs. predicted salinities for
spring and winter seasons within 8 selected regions and global NEGOM region.
78
Table 2.1. Regression coefficient of determination (r2), intercept, and slope, with
respective standard error (se), between salinity and ag443 (S=a + b*ag443), and mean
percentage error (MPE) for salinity prediction based on ag443 data within 95% prediction
interval of each region in different cruises (seasons).
The Mississippi River region (R1)
cid n intercept(se) slope(se)
N3 178 34.46(0.14) -31.16(0.76)
N4 202 36.49(0.09) -34.60(0.39)
N5 176 36.99(0.09) -33.81(1.06)
N6 240 30.98(0.21) -17.42(1.44)
N7 496 36.49(0.03) -25.87(0.41)
N8 272 36.36(0.03) -36.19(0.57)
N9 283 35.66(0.22) -23.30(0.95)
Escambia River region (R3)
cid n intercept(se) slope(se)
N3 140 36.36(0.14) -38.84(2.07)
N4 170 36.65(0.05) -27.44(0.53)
N5 152 36.37(0.16) -36.01(1.23)
N6 139 35.44(0.25) -20.34(3.87)
N7 135 36.84(0.07) -21.40(0.76)
N8 135 36.66(0.13) -34.03(1.32)
N9 139 35.07(0.12) -2.61(2.24)
Apalachicola River region (R5)
cid n intercept(se) slope(se)
N3 158 36.37(0.16) -30.14(3.83)
N4 138 35.92(0.04) -11.74(0.29)
N5 140 36.51(0.02) -35.75(0.67)
N6 132 34.41(0.46) 7.31(9.47)
N7 126 36.59(0.04) -20.16(0.53)
N8 218 37.16(0.03) -32.71(0.62)
N9 290 36.56(0.04) -13.63(0.69)
Tampa Bay region (R7)
cid n intercept(se) slope(se)
N3 175 33.704(0.06) 7.00(2.03)
N4 183 35.58(0.03) -13.91(0.35)
N5 173 36.27(0.01) -10.81(0.34)
N6 160 36.21(0.02) -3.08(0.45)
N7 87 36.81(0.05) -12.13(0.86)
N8 69 36.33(0.01)
2.70(0.17)
N9 343 36.69(0.03) 0.04(0.56)
r^2
0.90
0.98
0.84
0.38
0.88
0.94
0.69
MPE(%)
4.75
5.71
8.17
7.67
2.33
1.41
7.22
r^2
0.72
0.94
0.85
0.17
0.86
0.83
0.01
MPE(%)
3.55
1.68
3.43
5.81
1.39
2.01
1.04
r^2
0.28
0.92
0.96
0.01
0.92
0.92
0.58
MPE(%)
2.44
1.01
0.42
3.17
0.78
0.66
1.13
r^2
0.06
0.90
0.84
0.23
0.70
0.79
0.00
MPE(%)
1.27
0.84
0.44
0.53
1.10
0.09
0.63
Alabama River region (R2)
cid n intercept(se) slope(se)
r^2
N3 143 31.27(0.18) 2.92(2.09) 0.01
N4 152 36.32(0.04) -26.07(0.68) 0.90
N5 147 32.52(0.17) -18.95(0.84) 0.77
N6 145 31.25(0.24) 18.11(3.26) 0.18
N7 221 36.52(0.04) -33.07(0.46) 0.96
N8 179 35.44(0.07) -24.08(0.39) 0.96
N9 129 35.88(0.12) -16.46(1.56) 0.46
Choctawhatchee River region (R4)
cid n intercept(se) slope(se)
r^2
N3 168 38.03(0.14) -87.69(1.47) 0.96
N4 237 36.74(0.04) -25.26(0.42) 0.94
N5 161 37.83(0.22) -54.11(2.25) 0.79
N6 178 35.07(0.12) -17.70(1.91) 0.32
N7 155 37.02(0.10) -23.39(1.62) 0.58
N8 157 36.98(0.03) -26.76(0.99) 0.83
N9 148 35.94(0.09) -31.45(2.79) 0.46
Suwanne River region (R6)
cid n intercept(se) slope(se)
r^2
N3 190 35.84(0.01) -10.87(0.23) 0.92
N4 175 35.42(0.02) -9.34(0.17) 0.94
N5 188 36.22(0.02) -21.67(0.34) 0.96
N6 161 36.43(0.06) -17.23(0.71) 0.79
N7 123 35.94(0.04) 1.42(0.85) 0.02
N8 219 36.89(0.03) -15.99(0.05) 0.94
N9 296 37.38(0.04) -20.41(0.42) 0.88
Central of the NEGOM region (R8)
cid n intercept(se) slope(se)
r^2
N3 358 33.32(0.06) -38.86(0.60) 0.92
N4 371 35.95(0.09) 8.11(5.72) 0.01
N5 307 36.75(0.03) -41.28(2.46) 0.47
N6 299 36.74(0.06) -76.55(1.63) 0.88
N7 221 37.18(0.05) -63.01(3.01) 0.67
N8 293 36.44(0.01) 0.56(0.84) 0.00
N9 352 35.84(0.03) -28.62(0.27) 0.96
Note: cid=cruise identification (see Table 1), n=number of observations
79
MPE(%)
2.10
0.86
4.30
2.81
1.51
2.36
1.42
MPE(%)
2.28
1.46
4.82
2.62
1.37
0.55
1.00
MPE(%)
0.51
0.70
0.70
0.83
0.73
0.55
0.78
MPE(%)
3.62
1.00
0.23
3.48
0.33
0.12
1.36
2.3.3. CDOM versus Chlorophyll
River plumes and upwelling are the two significant sources of new nutrients to the
NEGOM. In upwelling areas, chlorophyll concentration and CDOM often covary or
chlorophyll is the dominant bio-optical constituent, indicating a classical Case I water
type (Morel and Prieur, 1977). In contrast, in river plumes (Case II water) there are
frequently conditions where CDOM and chlorophyll are not correlated (Hu et al., in
press).
High coefficient of determination for CDOM vs. chlorophyll concentration
regressions (r2 >0.70) were coincident with high r2 (>0.70) of salinity-ag443 relationships.
This suggested that riverine nutrient inputs controlled both chlorophyll and CDOM
levels, and that chlorophyll itself played only a minor role in the addition of CDOM to
plume waters. Del Castillo et al. (2000) reported similar results for the West Florida
Shelf. Off Tampa Bay, however, we found positive correlations between chlorophyll
concentration and ag443 (r2>0.6) during summer cruises (N3, N6, and N9) associated with
non-conservative mixing curves of salinity-ag443 (r2<0.23, see Table 2.1 and Figure 2.3).
A convex curvature in the salinity-ag443 relationships during summer cruises indicated
introduction of autochthonous CDOM to surface waters (either via upwelling and/or
phytoplankton degradation or grazing (Zweifel et al., 1995) or the release of CDOM
from sediments (Skoog, et al., 1996). Weisberg et al. (2000) described upwelling in this
region by examining an episode during summer 1994. The increase of chlorophyll
concentration due to upwelling during summer may have therefore contributed to an
increase in the CDOM pool.
80
2.3.4. Estimating Salinity from CDOM
Can surface salinity be estimated from CDOM observations? CDOM has been
proposed as a proxy to estimate surface salinity from satellite ocean color sensors
previously (Hu et al., 2003; D'Sa et al., 2002). However, this will depend on whether
CDOM can be estimated accurately from satellite (Carder et al., 1999; Hu et al., 2003)
and whether or not there is a robust relationship between CDOM and salinity.
To answer this question we analyzed the errors for salinities estimated from ag443
data, using empirical relationships between these variables (Table 2.1). We first estimated
salinities at the 95% confidence interval (also called the predicted interval). We then
calculated the mean percentage error (MPE) i.e., the average of the absolute value of the
difference between the estimated salinities at the 95% confidence interval and the
estimated salinities at the regression line, divided by the estimated salinity at the
regression line multiplied by 100% (graphical example presented in Figure 2.4).
While in estuarine and plume regions there was generally an inverse ag443-salinity
relationship, there were instances when ag443 increased with salinity at high salinities
(Figure 2.4); this was particularly apparent in the mid-shelf upwelling area off Tampa
Bay (R7, N8). The MPE of ag443-derived salinity varied regionally and seasonally,
ranging between 0.09% (R7, N8, correspond to 0.03 in salinity value) and 8.17% (R1,
N5, correspond to 2.29 in salinity value). These results indicated that using CDOM data
to estimate salinity of river plumes results in MPE of less than 9%. The lowest MPE was
found off the Suwannee River (0.51%-0.83%, correspond to 0.18-0.29 in salinity value),
and the highest MPE off the Mississippi River (1.41%-8.17%, correspond to 0.43-2.29 in
81
salinity value) (Table 2.1). The high MPE found in the Mississippi River region may
result from variation in the CDOM content from the various tributaries and marshlands.
Lower MPE variability was found during winter and higher during summer seasons.
Widespread patchiness, greater influence from adjacent rivers, and photobleaching may
all contribute to variation in the salinity-ag relationship.
To assess whether a single universal relationship between CDOM and salinity is
applicable within each of eight selected regions and within the whole NEGOM area, we
derived a salinity-ag443 regression (applying all NEGOM combined salinity-ag443 data) for
each season of each year, and attempted to predict salinity in other seasons and in each of
eight selected regions. First, we performed a global (whole NEGOM area) error analysis
on observed vs. predicted salinity. For example, the salinity-ag443 regression for winter
1998 (N4) was used to predict salinities based on ag443 data of spring 1999 (N5). Our
results showed that mean percentage error of observed vs. predicted salinity for the global
analysis ranged from 0.67% to 10.33% (correspond to 0.24-3.28 in salinity value; Table
2.2; Figure 2.5). Relatively high global mean percentage errors (up to 10.33%) were
found when winter or spring regression formulas were applied to predict salinities for
summer cruises and vice-versa. Meanwhile, relatively low global mean percentage errors
were obtained when applying global relationships based only on winter and spring data to
only those seasons in other years (Table 2.2). The winter 1998 (N4) relationship i.e.,
Salinity (S)=36.59-29.86*ag443 (n=8771, r2=0.86, 0.01<ag443<0.52, 16<S<36) produced
the lowest mean percentage errors for winter and spring seasons (0.82%-2.92%;
correspond to 0.29-1.01 in salinity value; Table 2.2). Employing global winter 1998 (N4)
82
relationship to predict salinities in winter and spring seasons both globally and regionally
produced the lowest mean percentage errors, therefore, this relationship may be useful to
estimate salinity from satellite ocean color data.
Table 2.2. Mean percentage error of observed vs. predicted salinity employing seasonal
global NEGOM salinity-ag443 relationships to predict salinities in other seasons.
Cruise Number
Mean percentage error (%)
Id
of obs.
R_N3 R_N4 R_N5
R_N6 R_N7
R_N8 R_N9
N3
8899
3.46
8.75
6.01
4.00
10.33
9.59
8.12
N4
8771
6.72
1.08
2.09
4.68
1.26
1.37
1.23
N5
9057
5.06
2.92
1.83
3.13
4.02
3.55
2.67
N6
8733
4.92
7.89
5.74
4.86
9.44
8.56
7.44
N7
7483
6.32
0.82
1.93
4.21
0.67
0.94
1.07
N8
7865
6.51
1.00
2.15
4.40
1.13
0.81
1.43
N9
11817
6.85
2.27
2.36
4.87
3.01
2.68
2.01
Note: R_N3, R_N4, R_N5, R_N6, R_N7, R_N8 and R_N9 refers to regression formula
derived from global NEGOM salinity-ag443 relationship during N3, N4, N5, N6, N7, N8
and N9, respectively used to predict salinities in others seasons (cruises).
83
Figure 2.4. Example of the error analysis of salinity predictions based on ag443 data. Mean
percentage error (MPE) equals to average of absolute values of 95% predicted interval
values minus regression line values divided by regression line value multiplied by 100%.
MPE for (A)=0.04% and (B)=8.12% (see also Table2). Note axis scales are different for
better display purpose.
Figure 2.5. Some example of observed and predicted global NEGOM salinity
employing global NEGOM salinity-ag443 relationship of other season. MPE equals to
average of absolute values of estimate salinity minus observed salinity divided by
observed salinity multiplied by 100%. These two cases represent the best and the worst
scenarios.
84
2.4. Conclusions
Coastal and shelf waters of the NEGOM experience seasonal influence of fresh
water discharge. Strong relationships between ag443 and CDOM fluorescence were
consistently observed during each of 7 cruises conducted to the NEGOM region in
different seasons over the course of three years. Relatively high ag443 (>0.1 m-1) were
observed in surface waters on the inner shelf (bottom depth < 100 m) near the mouths of
rivers during all seasons. During summers, riverine CDOM from the Mississippi River
reached the outer shelf and oceanic waters of the Gulf. Otherwise, riverine CDOM was
limited to coastal (inshore) regions and freshwater outflow was the major controlling
factor of CDOM concentration and distribution. The Mississippi can be a major source of
terrigenous CDOM to the outer shelf of the NEGOM, and eddies contribute to the
sporadic but frequent offshore dispersal of river plume waters.
In general, ag443 covaried linearly and inversely with salinity, particularly below
salinities of ~36. Therefore, salinity-CDOM relationships are sufficiently robust for river
plume identification. For all near-shore regions, high coefficient of determinations
(r2~0.70 to r2~0.98) between salinity and ag443 were observed during spring and winter
seasons, indicating a conservative mixing behavior. In contrast, during summer seasons
low (r2<0.70) to non-significant (r2=0.0) coefficient of determination were observed
between salinity and ag443, indicating non-conservative mixing behavior. In contrast to
near-shore regions, outer shelf region showed high negative correlations during summer
seasons.
85
We did not observe any indication that the increase of chlorophyll concentration
due to riverine nutrient input would result in an increase in the CDOM pool. Off Tampa
Bay, an increase in CDOM was observed concurrent with higher chlorophyll and
salinities, indicating possible CDOM contribution via either upwelling or from the
concomitant phytoplankton bloom. The high inverse correlation between ag443 and surface
salinity during winter and spring seasons in near-shore regions suggested that it might be
possible to derive salinity from satellite ocean color sensors, provided that ag443 is
estimated consistently from ocean color and relationship between ag443 and salinity is
obtained empirically from field measurements. The global NEGOM salinity-ag443
relationship of winter 1998 (N4), i.e., Salinity=36.59-29.86*ag443 (n=8771, r2=0.86;
0.01<ag443<0.52, 16<S<36) produced the lowest mean percentage errors of observed vs.
predicted salinities for all other winter and spring seasons (<3% errors; correspond to
<1.03 in salinity value), therefore, this relationship may be useful to estimate salinity
from satellite ocean color data. However, there was no robust global relationship that
helped explain summer values.
86
CHAPTER III:
VARIABILITY IN THE PARTICULATE ABSORPTION COEFFICIENTS OF
THE NORTHEASTERN GULF OF MEXICO SURFACE WATERS
87
Abstract
Variability in the light absorption coefficient of particulate materials in near-surface
waters of the Northeastern Gulf of Mexico (NEGOM) was examined during 7 seasonal
cruises in 1998-2000. Generally, the absorption coefficient of phytoplankton at 440 nm
[aph(440)] allowed a power law relationship (r2=0.83-0.98) with chlorophyll-a
concentration. The chlorophyll-specific a *ph (440) varied by a factor of 7. In general,
lower values of a *ph (440) (<0.06 m2 mg-1) were observed in the inner shelf and near the
outflow regions of major rivers. During summer, lower values of a *ph (440) were also
observed in the outer shelf and slope in the western NEGOM (west of 85W) in low
salinity waters of the Mississippi River plume. Higher values of a *ph (440) (>>0.1 m2 mg1
) [>>?] were otherwise observed outside the river plumes in the outer shelf and slope,
where lower chlorophyll-a concentration occurred. Packaging effects and pigment
composition were major factors in determining a *ph (440) variability. There were no
differences in the mean spectral phytoplankton absorption coefficients between spring
and fall seasons, but these were different from those observed in the summer. The
average value of a *ph (440) for microphytoplankton was also significantly lower than that
of nanophytoplankton and picophytoplantkon groups, which suggested that an increase in
cell size increased pigment packaging. Meanwhile, aph(673) covaried linearly with
chlorophyll-a concentration in all seasons (r2=0.87-0.99), and spatial and temporal
variability of the chlorophyll-specific a *ph (673) was very small. [what happened to your
original conclusion: “As opposed to previous studies by Smith and Baker (1978), Morel
88
(1988), and Gordon (1989) that the proportion of detritus to total particulate absorption
increased as waters became more oligotrophic, this study indicated that there was no clear
relationship between detritus absorption coefficient and chlorophyll-a concentration”?]
3.1. INTRODUCTION
Light absorption by phytoplankton and particulate material (detritus) play a
significant role in determining the variability of optical properties of oceanic and coastal
waters (Allali, et al., 1997). This optical variability also affects primary production,
remote sensing of pigment biomass and mixed layer heating (Suzuki et al., 1998;
Cleveland, 1995; Sosik and Mitchell, 1995; Nelson et al., 1993; Bricaud and Stramski,
1990; Yentsch and Phinney, 1989). For these reasons and an improvement in remote
sensing capabilities, information on the magnitude, range and sources of variability in the
absorption coefficient of phytoplankton and other particulate matter is very important.
Given the traditional difficulty of estimating phytoplankton pigment concentration in
coastal waters, information from these areas may be of particular interest for the
refinement of Case-II algorithms.
The phytoplankton absorption coefficient per unit of chlorophyll concentration
(chlorophyll-specific absorption coefficient, a *ph ()) is a key factor when modeling light
propagation within the ocean and ocean color (e.q., Gordon et al., 1988; Morel, 1988),
carbon fixation by phytoplankton (Kiefer and Mitchell, 1983), the contribution of
phytoplankton to the total absorption coefficient of seawater, as well as for modeling
89
marine primary production (Ishizaka, 1998; Morel and Andre, 1991; Sakshaug et al.,
1997).
In many studies, the value of a *ph () has been considered to be relatively constant
(Bannister, 1974; Kiefer and Mitchel, 1983; Berthon and Morel, 1992). However, the
variability in a *ph () has been extensively documented for both laboratory cultures (e.g.,
Fujiki and Taguchi, 2002; Ahn et al., 1992; Bidigare et al., 1990; Stramski and Morel,
1990; Berner et al., 1989; Mitchell and Kiefer, 1988a; Bricaud et al., 1988, 1983) and
natural populations (e.g., Lohrenz et al., 2003; Suzuki et al., 1998; Allali et al., 1997;
Sosik and Mitchell, 1995; Nelson et al., 1993; Bricaud et al., 1995; Cleveland, 1995;
Babin et al., 1993; Hoepffner and Sathyendranath, 1992, 1991; Carder et al., 1991;
Bricaud and Stramski, 1990; Yentsch and Phinney, 1989; Mitchell and Kiefer, 1988b).
[What are the variabilities then? From what to what? And where?]
Variability in the magnitude and spectral shape of a *ph () can be attributed to two
factors: (1) packaging effect i.e., pigments packed into chloroplast are less efficient in
absorbing light per unit pigment mass than in an optically thin solution (Kirk, 1994), and
(2) pigment composition. An increase in pigment packaging can occur either as cell size
increases or the internal concentration of pigments increases (Lohrenz et al., 2003; Kirk,
1994; Morel and Bricaud, 1981). Differences in phytoplankton species, as well as
variation within a species grown under different environment conditions (e.g., growth
irradiance), also causes variability in a *ph () due to pigmentation and packaging effects
90
(Fujiki and Taguchi, 2002; Mitchell and Kiefer, 1988a; Bricaud et al., 1988; Morel and
Bricaud, 1986; Dubinsky et al., 1986; Berner et al., 1989, Stramski and Morel 1990).
Several investigators have reported an increase in a *ph () with a decrease in
chlorophyll concentration (Bricaud et al., 1998, 1995; Cleveland, 1995; Bricaud and
Stramski, 1990). They interpreted this result mainly due to packaging effect or a general
trend of decreasing cell size with decreasing chlorophyll concentration. However, the
above relationship can also be attributed solely to the increasing contribution of accessory
pigments in waters with low chlorophyll concentration, with no consideration of
packaging effect (Ciotti et al., 1999; Sakshaug et al., 1997; Bricaud et al., 1995; Wozniak
and Ostrowska, 1990). Carder et al. (1991) also suggested that chlorophyll dependency of
a *ph () could differ between regions such as subtropical and temperate regions due to the
typical differences in cell size and light regimes. In contrast, Hoepffner and
Sathyendranath (1992) reported that there was no dependency of a *ph (440) on chlorophyll
concentration in the Gulf of Maine and Georges Bank regions within the range of 0.052.5 mg m-3 chlorophyll.
The Northeastern Gulf of Mexico (NEGOM) encompasses a wide variety of
ecosystems which are influenced by a combination of nutrient-rich input from rivers and
estuaries, coastal upwelling, and vertical mixing (Muller-Karger, 2000; Pennock et al.,
1999; Gilbes et al., 1986). These processes exert a strong influence on the phytoplankton
abundance and distribution in the NEGOM. To date, a *ph () variability had not been
studied in the NEGOM region.
91
This study uses field observations from seven of two-week cruises in the NEGOM
region (1998 through 2000), to examine the relationship between variability in a *ph (),
riverine influence (salinity), nutrient supply, and pigment concentration. The specific
objectives of this dissertation chapter are to: (1) determine the seasonal and spatial
variation of the chlorophyll-specific absorption coefficient of phytoplankton in the
NEGOM, (2) determine the percentage contribution of phytoplankton and detritus to the
total particulate absorption coefficient, (3) determine whether the phytoplankton species
or group can be estimated from phytoplankton absorption coefficient (aph()) using
remote sensing data, [? How do you derive aph from remote sensing data?] and (4)
understand the relationship between community structure and pigment composition of
surface waters.
3.2. METHODOLOGY
3.2.1. Sample Collection
Near-surface (+ 3 m-deep) water samples were collected from the Northeastern
Gulf of Mexico (NEGOM; Figure 1.1, Chapter I) during seven of two-week cruises
aboard the Texas A&M University R/V Gyre. Cruises were conducted in spring (April or
May; cruises N5, N8), summer (July/August; cruises N3, N6, N9), and fall (November;
cruises N4, N7) of 1998, 1999, and 2000 (Table 1.1). Each cruise surveyed eleven crossmargin transects from the 10-m to the 100-m isobath. Water samples for determining the
total particulate, phytoplankton, and detritus absorption coefficients were collected from
the outflow of a flow-through system that pumped at a rate of 10 liters/minute from a hull
depth of about 3-m, and which passed water via a 10-liter debubbler and mixing chamber.
92
Seawater samples (0.5-5.0 L) were collected and filtered immediately aboard ship
through Whatman GF/F filters (25 mm diameter) under low vacuum pressure (<0.5 atm).
The volume of water filtered varied between ~0.05 and 5.0 liters depending on the
concentration of pigmented particles in the sample. Filtering was discontinued once the
filter displayed sufficient color to the naked eye. Each filter pad was half-folded and
placed into a 2.0 ml sterile Nalgene cryogenic vial, and stored in liquid N2 for analysis.
After arrival from the field, samples were stored in a deep-freezer.
Water samples for pigment analyses were collected from one of twelve 10-L
Niskin bottles operated with a Sea-Bird SBE 911 profiling system used to collect
conductivity-temperature-depth (CTD) profiles. Pigment sampling and analyses,
including the calibration of a fluorometer on the flow-through system, were conducted by
Texas A&M University. Detailed description of data collection and analysis methods can
be found in Qian et al. (2003).
3.2.2. Absorption Measurements
The quantitative filter technique (Yentsch, 1962; Kiefer and SooHoo, 1982) was
used to determine absorption spectra due to total particulate (phytoplankton and detritus,
aph()) and detritus (ad()). Prior to analysis, sample and reference filter pads were
allowed to thaw slowly at room temperature for about 5-10 minutes prior to being placed
in a dark petri dish and moistened with a drop of Milli-Q water. The moist sample and
reference filter pads were placed on individual glass plates (diameter=2.4cm) in a
custom-made transmissometer box. Prior to each scan, the filters were slid one at a time
over a tungsten-halogen light source that shone through a blue long-pass/cut-off filter and
93
a quartz glass diffuser. Using a custom made, 512-channel spectroradiometer (~350850nm), the transmittances of the sample filter (Tsample()) and the reference filter
(Treference()) were measured three times. These measurements were averaged and used to
obtain the optical densities of total particulate matter (ODp()) as shown below.
The sample filter was then soaked with ~40-50ml of hot 100% methanol for 10-15
minutes in the dark to extract phytoplankton pigments (Kishino et al., 1985; Roesler et
al., 1989). Transmittances of the extracted filter (Tdetritus()) and the reference filter were
once again measured three times, and the optical density of detritus (ODd()) calculated.
Optical densities were calculated as follows:
T
λ  

reference
OD λ   log 10 
p
 Tsample λ  


(1a)
T
λ  
OD λ   log10  reference
d
 T
λ  
 detritus
(1b)
The absorption coefficients of particulate matter (ap()) and detritus (ad()), were
calculated as follows:
a p λ  
a
λ  
d
ln 10  OD p λ   β
l
ln 10  OD λ   β
d
l
(2a)
(2b)
where “l” is the geometric pathlength equal to the volume of seawater filtered divided by
the effective filtration area of the filter (r2, r=0.0215/2 m) and  is the pathlength
94
amplification or “ factor” (Butler, 1962). The  factor is an empirical formulation
defined as the ratio of optical to geometric pathlength that corrects for multiple scattering
inside the filter. In this study, an average of two published  factor formulations (Bricaud
and Stramski, 1990; Nelson et al., 1993) was used as follows:
0.5
β  1.0  0.6  ODp λ 
(3)
Spectra with ODp(675) less than 0.04 were omitted from this study to minimize artifacts
due to uncertainty in the  factor (Mitchell and Kiefer, 1988; Bricaud and Stramski,
1990; Cleveland and Weidemann, 1993; Nelson et al., 1993; Moore et al., 1998;
Lohrenz, 2000). All other spectra were set to zero at 750 nm to correct for residual
scattering caused by non-uniformity in wetness between the sample and reference filters
or for stray light.
Using Eqs. 2a and 2b, the absorption spectra due to phytoplankton pigments,
aph(), was calculated as follows:
a ph ( )  a p ( )  a d ( )
(4)
Fluorometric chlorophyll and phaeopigment concentrations were determined on the
filtrate of phytoplankton pigments extracted from the sample filter with hot 100%
methanol using a Turner 10-AU-005 fluorometer according to the methods of HolmHansen and Riemann (1978). Finally aph() was converted to chlorophyll-a specific
absorption coefficient ( a *ph ()) dividing by chlorophyll-a concentration.
The local maximum of the total particulate absorption coefficient spectra near 440,
545, 625 and 673 nm was normally observed in the range of 440+5, 545+5 and 673+5
95
nm, respectively. Therefore, the magnitude of phytoplankton, detritus, total particulate
and chlorophyll-specific absorption coefficients at these wavelengths (440, 545 and 673
nm) was obtained by averaging spectra from the above range of wavelengths.
3.3. RESULTS AND DISCUSSION
3.3.1. Total Particulate, Detritus and Phytoplankton Absorption Coefficients
The near-surface absorption coefficients for total particulate (ap()), detritus (ad()),
and phytoplankton (aph()) in the NEGOM from 1998 to 2000 are presented in Figure
3.1, 3.2 and 3.3. Higher total particulate absorption coefficients at 440 nm were observed
near the coast and particularly near the Mississippi river mouth (as high as 0.85 m-1). [for
what salinity?] Lower values were observed offshore (as low as 0.01 m-1) (Figure 3.1).
Detritus and phytoplankton absorption coefficients at 440 nm also showed higher values
in the coastal region relative to offshore (range of 0.01 to 0.51 m-1 for detritus and 0.01 to
0.61 m-1 for phytoplankton; Figure 3.2 and 3.3). A similar pattern was observed for all
absorption coefficients at 673 nm.
In coastal waters, particularly near river mouths, chlorophyll concentrations were
relatively high but absorption by detritus in the blue (ad(440)) exceeded that of
phytoplankton absorption (aph(440)). In other words, detritus contributed on average
between 52% and 66% to the total particulate absorption in coastal areas near the river
mouths. In contrast, in offshore waters where chlorophyll concentration was relatively
low, absorption by phytoplankton generally exceeded that by detritus (phytoplankton
96
contributed on average between 68% and 77% to the total particulate absorption). This
result is contrary to many previous studies that have reported that the proportion of
detritus absorption increases offshore (especially in oligotrophic areas) relative to
nearshore (Gordon, 1989; Morel, 1988; Smith and Baker, 1978). [why? What’s the
reason?]
The mean absorption coefficient at 440 nm for total particulate, detritus, and
phytoplankton varied seasonally, with maxima during summer cruises (i.e.,
ap(440)=0.0875 m-1, ad(440)=0.0277 m-1 and aph(440)=0.0605 m-1, respectively) and
minima during spring cruises (i.e., ap(440)=0.0206 m-1, ad(440)=0.0076 m-1 and
aph(440)=0.0130 m-1, respectively).
There was no clear relationship between the detritus absorption coefficient at 440
nm and chlorophyll-a concentration in individual samples (Figure 3.4). At low
chlorophyll-a conentration (<1 mg m-3), ad/ap seems to vary significantly. However, the
average absorption coefficients for each cruise showed a tendency to covary with
chlorophyll concentration averaged for the cruise (Figure 3.5). These results indicate that
chlorophyll-a played a major role in determining total particulate absorption coefficient
in the NEGOM region. [we know this even without measurements] The exceptions were
the river mouths of the Mississippi, Alabama, Apalachicola Rivers and in some offshore
of the Mississippi river plume, which detritus contributed to more than 50% of total
particulate absorption coefficients.
97
98
Figure 3.1. Near-surface total particulate absorption spectra in the NEGOM between
1998 and 2000. ‘N’ indentifies cruise number (Table 3.1). Note different scales on y-axis.
99
Figure 3.2. Near-surface detritus absorption spectra in the NEGOM between 1998 and
2000. ‘N’ indentifies cruise number (Table 3.1). Note different scales on y-axis.
100
Figure 3.3. Near-surface phytoplankton absorption spectra in the NEGOM between 1998
and 2000. ‘N’ indentifies cruise number (Table 3.1). Note different scales on y-axis.
101
Figure 3.4. Proportion of detrital to total particulate absorption coefficient at 440 nm, as a
function of chlorophyll-a concentration for each cruise. See Table 1.1 for cruise
identification.
0.1
1.4
ad440
0.09
1.2
Absorption coefficient
0.08
aph440
1
0.07
ap440
0.06
0.8
chl
0.05
0.6
0.04
0.03
0.4
0.02
0.2
0.01
0
0
N3
N4
N5
N6
N7
N8
N9
Cruise Identification
Figure 3.5. Time series of mean ap(440), ad(440), aph(440) and chlorophyll-a
concentration of each cruise, indicating possible concurrent seasonal variability in the
NEGOM. Right y-axis is for chlorophyll-a concentration. See Table 1.1 for detail cruise
102
identification. [Bisman: these are my original comments: “[need to overlay standard
deviation. Can these “mean” values represent the mean state of NEGOM? In other words,
if I take a sample randomly in the NEGOM during summer, and during the winter, can I
say (without measurement) the former is higher than the latter? If the stddev is larger than
the mean, than these mean values don’t mean much.” Why did you not take care of these
comments?]
103
3.3.2. Variation in Spectral Shape of the Phytoplankton Absorption Coefficient
Phytoplankton absorption (Figure 3.3) shows peaks at approximately 440 and 673
nm, typical of absorption by chlorophyll-a (Cleveland, 1995). The spectral shape of the
absorption coefficient associated with phytoplankton pigments was examined by
normalizing each of the spectra to its value at 440 nm for each cruise (sampling season).
Average spectra for spring, summer, and fall are presented in Figure 3.6. There was
considerable variation around the mean for all seasons, which is largely attributed to
variation in the region along the cruise track. The secondary peak observed around 545
nm was generally observed in the inner shelf (water depth less than 50 m), except during
summer cruises when this also occurred in the river plume extending over outer shelf
(water depth 50-400 m) and slope (water depth >400 m) regions.
The 545 nm peak may be due to the present of fucoxanthin and/or phycoerythrobilin pigments. Anderson and Barrett (1986) reported that the fucoxanthinchlorophylla/c-protein complex enhanced absorption near 542 nm. Phycoerythrobilin,
found in cyanobacteria, also has an absorption band around 545 nm (Moore et al., 1995;
Morel et al., 1993, Hoepffner and Sathyendranath, 1991). Indeed, Qian et al. (2003)
reported that fucoxanthin concentrations (as biomarker for diatoms) were generally high
in the inner shelf of the NEGOM and offshore areas affected by fresh water discharge.
On the other hand, prokaryotes (cyanobacteria) containing phycoerythrobilin were also
observed in patches over the entire study area (Qian et al., 2003).
There was no significant difference between the spectra of spring (cruises N5, N8)
and fall (N4, N7). However, a one-way analysis of variance (ANOVA) showed that there
104
were significant differences in the mean values of normalized phytoplankton absorption
coefficient at 545 nm and 673 nm between those of summer (cruises N3, N6, N9) and
those of spring and fall (P>0.05). There were no significant differences between spring
and fall. For example, the mean values of normalized phytoplankton absorption
coefficient for 545 nm and 673 nm were higher during summer (0.209 and 0.272) than
that of spring (0.122 and 0.223) and fall (0.124 and 0.236) seasons (Figure 3.6, plot H).
Higher mean normalized phytoplankton absorption coefficients at 545 and 673 nm
during summer may be attributed to the offshore dispersal of the Mississippi River plume
offshore in the NEGOM, which leads to higher mean chlorophyll-a concentration [No!
No! No! Bisman – how many times do I need to tell you about the difference of absolute
aph and relative (you call it normalized) aph? Read my original comments] and more
large-celled diatoms. [why do diatoms have higher relative aph at 545 and 673?]
The seasonal variations in the spectral shape of the phytoplankton absorption,
however, were relatively small. [I don’t understand this – you just said there was big
difference between summer and spring/winter, now you said it is not significant. What’s
the truth? Read my original comments] The variations in the magnitudes of chlorophyllspecific absorption coefficients in the blue band were more pronounced.
Seasonal and spatial distribution of aph545/aph440, aph625/aph440, aph673/aph440 and
the average are presented in Figure 3.7, 3.8, 3.9 and 3.10. Relatively high values were
observed in the inner shelf especially in the river major mouths and relatively low values
in the offshore, except during summer season where high values were also observed in
the offshore associated with the Mississippi river plume. The spatial distribution of
105
aph545/aph440, aph625/aph440 and aph673/aph440 also seems to correlate with spatial
distributions of salinity and chlorophyll-a. The correlation seems better when
aph545/aph440, aph625/aph440 and aph673/aph440 were averaged (Figure 3.10).
To study the effect of taxonomic groups on aph545/aph440, aph625/aph440 and
aph673/aph440 in the NEGOM, an index was computed based on pigment composition as
per Vidussi et al., (2001). They used pigments to derive cell size class markers of
phototroph groups, such as picophytoplankton (<2 m), nanophytoplankton (2-20 m),
and microphytoplankton (>20 m). Pigments marker for picophytoplankton (<2 m)
were used such as: Zeaxanthin (biomarker for cyanobacteria and prochlorophytes;
Gieskes et al, 1988; Guillard et al., 1985; Chisholm et al., 1988), Divinyl-chlorophyll a
(biomarker for prochlorophytes; Goericke and Repeta, 1992; Gieskes and Kraay, 1983;
Chisholm et al., 1988) and Chlorophyll b+Divinil-chlorophyll b (biomarkers for green
flagellates and prochlorophytes; Partensky et al., 1993; Moore et al., 1995; Jeffrey, 1976;
Simon et al., 1994). Pigments marker for nanophytoplankton were used such as:
19’hexanoyloxyfucoxanthin and 19’butanoyloxy-fucoxanthin (biomarker for
chromophytes nanoflagellates; Bjornland and Liaaen-Jensen, 1989; Hooks et al., 1988;
Arpin et al., 1976; Wright and Jeffrey, 1987; Jeffrey and Vesk, 1977; Andersen et al.,
1993; Bjornland et al., 1989), and Alloxanthin (a biomarker for cryptophytes; Jeffrey
and Vesk, 1977; Gieskes and Kraay, 1983). Pigments marker for microphytoplankton
were used such as: Fucoxanthin (a biomarker for diatoms; Bjornland and Liaaen-Jensen,
1989; Hooks et al., 1988; Wright and Jeffrey, 1987; Kimor et al., 1987), and Peridinin (a
106
biomarker for dinoflagellates; Kimor et al., 1987; Johansen et al., 1974; Jeffrey et al.,
1975). The formula to derive cell size class marker is as follows:
DP (Diagnostic Pigments)=Zea+chl_b+Allo+19’-HF+19’-BF+Fuco+Peri
(5)
BPpico = (Zea+chl_b)/DP
(6)
BPnano = (Allo+19’-HF+19’-BF)/DP
(7)
BPmicro= (Fuco+Peri)/DP
(8)
Where BPpico=Biomass Proportion of picophytoplankton, BPnano= Biomass Proportion of
nanophytoplankton, BPmicro= Biomass Proportion of microphytoplankton, Zea= [zeaxanthin], chl_b=[chlorophyll_b], Allo=[alloxanthin], 19’-HF=[19’hexanoyloxyfucoxanthin],
19’-BF=[19’butanoyloxyfucoxanthin], Fuco=[fucoxanthin], Peri=[peridinin].
Biomass proportion distribution for microphytoplankton, nanophytoplankton and
picophytoplankton are presented in Figure 3.11, 3.12, and 3.13. Relatively high biomass
proportions of microphytoplankton (diatoms and dinoflagellates) were observed in the
inner shelf particularly in the major river mouths and the relatively low values in
offshore. During summer cruises, relatively high biomass proportions of microphytoplankton were also observed in the offshore region (Figure 3.11) along the Mississippi
River plume . It appears that distribution pattern of microphytoplankton biomass
proportion correlates with distribution of aph545/440, aph625/440 and aph673/440 (see
Figure 3.7, 3.8, and 3.9).
Relatively high values of biomass proportion for nanophytoplankton (chromophytes
nanoflagellates and cryptophytes) were observed particularly in offshore and the the
relatively low in inshore. During summer, it appeared that relatively low
107
nanophytoplankton biomass proportion extended offshore from the Mississippi delta
(Figure 3.12).
The abundance of pycophytoplankton group (cyanobacteria, prochlorophytes and
green flagellates) was mixed along the NEGOM but relatively high biomass proportions
were only observed in late spring (N5) and summer (N3, N6, N9) (Figure 3.13).
To examine whether algal group can be estimated from satellite assuming
phytoplankton absorption coefficient can be estimated accurately from satellite, I
examined the relationship between aph545/440, aph625/440, aph673/440 and biomass
proportion of algal size markers i.e., microphytoplankton, nanophytoplankton and
picophytoplankton. Positive relationships were found between aph545/440, aph625/440,
aph673/440 and microphytoplankton biomass proportion with coefficient of determination
(r2) was lower in summer than in spring and fall. Correlations between normalized
phytoplankton absorption coefficients and algal size groups improved especially during
summer when aph545/440, aph625/440 and aph673/440 were averaged and correlated with
algal size groups. Using the above average values, positive correlations with r2=0.2 (N6)
to r2=0.76 (N5) were found between normalized phytoplankton absorption coefficients
and microphytoplankton biomass proportion (Figure 3.14). Negative low correlations
were found between normalized phytoplankton absorption coefficients and
nanophytoplankton biomass proportion (Figure 3.15). Meanwhile, non significant
correlations were found between normalized phytoplankton absorption coefficients and
picophytoplankton biomass proportion (Figure 3.16).
108
Relatively low correlation values (in average r2=0.48) between normalized
phytoplankton absorption coefficients and microphytoplankton biomass proportion
indicates that it is impossible to estimate phytoplankton group from satellite.
Figure 3.6. Phytoplankton absorption spectra normalized to 440 nm from seven NEGOM
cruises. Individual normalized spectra are shown in black and the average normalized
spectra in blue. Panel H shows the average normalized spectra for spring, summer and
fall seasons. ‘N’ followed by a number identifies each cruise (Table 1.1).
109
Figure 3.7. Near-surface distribution of the aph545/440 in the NEGOM. White dots show
location of water sample collections. See Table 1.1 for cruise identification.
110
Figure 3.8. Near-surface distribution of the aph625/440 in the NEGOM. White dots show
location of water sample collections. See Table 1.1 for cruise identification.
111
Figure 3.9. Near-surface distribution of the aph673/440 in the NEGOM. White dots show
location of water sample collections. See Table 1.1 for cruise identification.
112
Figure 3.10. Near-surface distribution of the of mean aph545/440+aph625/440+
aph673/440 in the NEGOM. White dots show location of water sample collections. See
Table 1.1 for cruise identification.
113
Figure 3.11. Spatial distribution of biomass proportion of cell size marker for
microphytoplankton. See Table 1.1 for cruise identification.
114
Figure 3.12. Spatial distribution of biomass proportion of cell size marker for
nanophytoplankton. See Table 1.1 for cruise identification.
115
Figure 3.13. Spatial distribution of biomass proportion of cell size marker for
pycophytoplankton. See Table 1.1 for cruise identification.
116
Figure 3.14. Relationship between the average of aph545/440+aph625/440+aph673/440
and microphytoplankton biomass proportion for each cruises. NA indicates all NEGOM
data were used. See Table 1.1 for detail of cruise identification.
117
Figure 3.15. Relationship between average of aph545/440+aph625/440+aph673/440 and
nanophytoplankton biomass proportion for each cruises. NA indicates all NEGOM data
were used. See Table 1.1 for detail of cruise identification.
118
Figure 3.16. Relationship between average of aph545/440+aph625/440+aph673/440 and
picophytoplankton biomass proportion for each cruises. NA indicates all NEGOM data
were used. See Table 1.1 for detail of cruise identification.
119
3.3.3. Phytoplankton Absorption Coefficient vs. Chlorophyll-a Concentration
We examined the relationship between phytoplankton absorption coefficient
(aph()) and chlorophyll-a concentration at 440 and 673 nm (Figure 3.17 and 3.18). The
intercepts at zero chlorophyll-a concentration were very low, which suggested that aph()
in low-chlorophyll waters co-varied strongly with chlorophyll and did not occur in excess
of chlorophyll.
Most cruises did not show a linear relationship between aph(440) and chlorophyll-a
concentration, but instead exhibited a power law relationship with coefficient of
determination (r2) range from 0.83 to 0.98. Only two cruises showed a linear relationship
(N3 and N9, Figure 3.17).
Strong linear relationship between aph(673) and chlorophyll-a concentration were
generally observed during all seasons (r2=0.87-0.98, Figure 3.18). The slopes of the
regressions also spanned within the small range (0.013-0.017). A strong linear
relationships between aph(673) and chlorophyll-a concentration for all cruises indicated
that variability in chlorophyll-specific absorption coefficient at 673 nm was small. By
combining all data from seven cruises aph(673) and chlorophyll-a concentration are
related through the following equation:
aph(673) = 0.0153*chla (r2=0.96, n=436).
120
3.3.4. Variation in Chlorophyll-Specific Absorption Coefficient
The chlorophyll-specific absorption coefficient at 440 nm was found to be highly
variable in the NEGOM region in all cruises (Figure 3.9). a *ph (440) varied by about a
factor 7 ranging from 0.02 to 0.15 m2 mg-1 over the study period (Figure 3.19).
Variability of a *ph (673) was less pronounced and only varied by a factor 2 to 3.
Near-surface spatial and temporal variability of a *ph (440) for seven cruises is
presented in Figure 3.20. There was a general trend of increasing a *ph (440) with distance
from shore where chlorophyll-a concentration also decreased. During summer cruises,
relatively low a *ph (440) was also observed in the outer shelf and slope of the western
NEGOM (Figure 3.10) and higher a *ph (440) values (~0.15 m2 mg-1) were observed only
in the middle shelf off Florida. Relatively low a *ph (440) were observed along the inner
shelf and specifically near river mouths including the Mississippi, Alabama, Apalachicola
and Suwannee river outflow regions.
Since a *ph (440) was higher in areas of lower chlorophyll-a concentration and vice
versa, this indicated a packaging effect. Because the relationship between a *ph (440) and
chlorophyll-a concentration was not linear, this also indicated pigment composition also
played a role in determining a *ph (440) variability.
121
Figure 3.17. Relationship between phytoplankton absorption coefficient at 440 nm
(aph(440)) and chlorophyll-a concentration for each cruises. Note the scales on both x and
y-axis are different between panels. See Table 1.1 for detail of cruise identification.
122
Figure 3.18. Relationship between phytoplankton absorption coefficient at 673 nm
(aph(673)) and chlorophyll-a concentration for each cruises. Note the scales on both x and
y-axis are different between panels. See Table 1.1 for detail of cruise identification.
123
Figure 3.19. Near-surface chlorophyll-specific absorption spectra measured on the
NEGOM region during summer, spring and fall seasons between 1998 and 2000. ‘N’
followed by a number indicates cruise identification (see Table 1.1 for cruises
identification). ‘aph’ indicates phytoplankton absorption coefficient. Note the different
scales on y-axis.
124
Figure 3.20. Near-surface distribution of the chlorophyll-specific absorption coefficient at
440 nm in the NEGOM. White dots show location of water sample collections. See Table
1.1 for cruise identification.
125
126
3.3.5. Sources of Variability in Chlorophyll-Specific Absorption Coefficient
Variability in a *ph () might be expected due to the pigment composition and
packaging effects (Lohrenz et al., 2003; Fujiki and Taguchi, 2002; Kirk, 1994; Stramski
and Morel 1990; Berner et al., 1989; Mitchell and Kiefer, 1988a; Bricaud et al., 1988;
Morel and Bricaud, 1986; Dubinsky et al., 1986; Morel and Bricaud, 1981). In the
NEGOM, spatial distribution of a *ph (440) followed chlorophyll-a concentration and
salinity patterns. Low salinity, with a higher nutrient content, may have played a role in
lowering a *ph (440). However, there was not statistical relationship between surface
nutrient concentrations (NO2+NO3+NH4+Urea) and a *ph (440), likely because nutrients
are consumed as fast as they are supplied.
To study the effect of taxonomic groups on a *ph (440) in the NEGOM, an index was
computed based on pigment composition as per Vidussi et al., (2001) as discussed in
section 3.3.2. For each major Biomass Proportion (BP>0.5), the a *ph (440) values were
grouped into the three corresponding cell size markers. Scatter plots of a *ph (440) and
Biomass Proportion of cell size markers is presented in Figure 3.21. Although scatter is
high, it seems that biomass of nanophytoplankton and picophytoplankton have a positive
correlation with a *ph (440). Microphytoplankton showed either no relation or perhaps a
slight negative correlation suggesting higher abundance of microphytoplankton lowered
the value of a *ph (440) (Figure 3.21). The average value of a *ph (440) for microphytoplankton was also significantly lower than that of nanophytoplankton and
127
picophytoplantkon groups (Table 3.1). These results strongly suggest that an increase in
cell size resulted in increase in pigment packaging.
The spatial variability of a *ph (440) (Figure 3.20) may be explained by the findings
of Qian et al. (2003), who observed large-celled diatoms in the inner shelf of the
NEGOM, near major river outflow regions and in river plumes, and low chlorophyll-a
concentration and small-celled nanophytoplankton and picophytoplankton offshore and
outside the river plumes. Our computations of biomass proportion in different
phytoplankton groups (Figure 3.11, 3.12, and 3.13) corroborated these findings.
The small size of the prokaryotes and prymnesiophytes should lead to a higher
absorption efficiency compared to that of the large-celled diatoms (Morel and Bricaud,
1981; Kirk, 1975; Duysens, 1956). Therefore, region associated with predominantly
microphytoplankton or diatoms would contribute to the lower values of a *ph (440) and in
contrast region associated with predominantly nanophytoplankton and picophytoplankton
or prokaryotes and prymnesiophytes would contribute to the higher values of a *ph (440).
Table 3.1. Mean values of a *ph (440) for all seven NEGOM cruises data corresponding
with the three major size cell markers of phytoplankton.
ID
Cell size markers
N
Mean
SD
Different From ID*
A
nanophytoplankton
136
0.0875 0.0020 C
B
picophytoplankton
64
0.0815 0.0029 C
C
microphytoplankton
29
0.0484 0.0044 A,B
*Average values were compared for the differences among cell size markers using the
Least Significant Difference (LSD) Method
128
Figure 3.21. Scatter plots of a *ph (440) and biomass proportion per algal group for all
seven NEGOM cruises.
129
3.4. Conclusions
In the coastal NEGOM near river outflows where chlorophyll concentration was
relatively high, light absorption in the blue band (440 nm) by detritus normally exceeded
that of phytoplankton absorption (ad was 54% to 66% of the total particulate absorption
or ap). In contrast, in offshore waters, absorption by phytoplankton contributed more than
that of detritus (aph was 68% to 77% of ap). There was no clear relationship between
detritus absorption coefficient and chlorophyll-a concentration.
There were no differences in the mean spectral phytoplankton absorption
coefficients between spring and fall seasons, but these were different from those observed
in the summer. Higher mean phytoplankton absorption coefficients during summer may
be attributed to the Mississippi River freshwater input to offshore NEGOM waters.
In general, aph(440) covaried non-linearly with chlorophyll-a concentration
following a power law (r2=0.83-0.98). Meanwhile, aph(673) covaried linearly with
chlorophyll-a concentration in all seasons (r2=0.87-0.99), and the variability of
chlorophyll-specific absorption coefficient at 673 nm was very small.
Even though positive correlation between phytoplankton absorption coefficient and
microphytoplankton biomass proportion was found, relatively low the correlation
indicates that the estimation of algal group (species) from satellite is still impossible.
Variability of a *ph (440) was more pronounced within a cruise than among the
seasons. It varied by about a factor of 7. In general, lower values of a *ph (440) were
observed in the inner shelf and major rivers outflow regions. Higher a *ph (440) were
observed in outer shelf and slope, associated with high salinity, and lower chlorophyll-a,
130
except during summer when lower a *ph (440) were also observed in outer shelf and slope
to the west of about 85W in the low salinity Mississippi plume. Mean values of
a *ph (440) in waters dominated by microphytoplankton (diatoms) were significantly lower
than in waters with nanophytoplankton and picophytoplankton communities. Therefore,
a *ph (440) was controlled by the chlorophyll-a concentration and phytoplankton
community structure, indicating packaging effects and pigment composition play a major
role in determining a *ph (440) variability in NEGOM region.
131
REFERENCES
Ahn, Y. H., A. Bricaud and A. Morel. (1992). Light backscattering efficiency and
related properties of some phytoplankters. Deep Sea Res., Part A, 39:1835-1855.
Allali, K., A. Bricaud and H. Claustre. (1997). Spatial variations in the
chlorophyll-specific absorption coefficients of phytoplankton and photosynthetically
active pigments in the equatorial Pacific. J. Geophys. Res., 102(C6):12,413-12,423.
Anderson, J. M., and J. Barrett. (1986). Light-harvesting pigment-protein
complexes of algae. In: Staehelin, L. A., and C. J. Arntzen (Eds), Photosynthesis III.
Photosynthe-thic membranes and light harvesting system, vol. 19. Springer, Berlin, pp.
269-285.
Arfken, G. (1985). Mathematical methods for physicists (3rd edition). Academic
Press, Inc., New York. 985pp.
Arrigo, K. R., and C. W. Brown. (1996). Impact of chromorphoric dissolved
organic matter on UV inhibition of primary productivity in the sea. Mar. Ecol. Prog.
Ser., 140:207-216.
Argyrou, M. E., T. S. Bianchi, and C. D. Lambert. (1997). Transport and fate of
particulate and dissolved organic carbon in the Lake Pontchartrain estuary, Louisiana.
U.S.A. Biogeochemistry, 38:207-226.
Atlas, E. L., L. I. Gordon, S.W. Hager and P. K. Park. (1971). A practical manual
for use of the technic on AutoAnalyzer in seawater nutrients analysis (revised).
Technical report 215, Dept. of Oceanography, Oregon State University, Corvallis,
Oregon. 49pp.
Austin, R.W. (1974). Inherent spectral radiance signatures of the ocean surface.
Ocean Color Analysis (Technical Report), SIO Ref. 7410:1-20.
Babin, M., A. Morel, H. Claustre, A. Bricaud, Z. Kolber, and P. G. Falkowski.
(1996). Nitrogen and irradiance-dependent variations of the maximum quantum yield of
carbon fixation in eutrophic, mesotrophic and oligotrophic marune system. Deep-Sea
Res. I, 43:1241-1272.
Babin, M., J. C. Therriault, L. Legendre and A. Condal. (1993). Variations in the
specific absorption coefficient in natural phytoplankton assemblage: Impact on estimates
of primary production. Limnol. Oceanogr., 38:154-177.
Bakun, A. (1973). Coastal upwelling indices, west coast of North America, 194671. U.S. Dept. Commer., NOAA Tech. Rep., NMFS SSRF-671, 103pp.
132
Bakun, A. (1975). Daily nad weekly upwelling indices, west coast of North
America, 1967-73. U.S. Dept. Commer., NOAA Tech. Rep., NMFS SSRF-693, 114pp.
Bakun, A. and C.S. Nelson. (1977). Climatology of upwelling related processes
off Baja California. Calif. Coop. Oceanic Fish. Invest. Rep., 19:107-127.
Belabbassi, L. (2001). Importance of physical processes on near-surface nutrient
distributions in summer in the northeastern Gulf of Mexico, M.S. thesis, 75 pp., Texas
A&M University, Texas.
Benner, R., C. Chin_leo, W. Gardner, B. Eadie, and J. Cotner. (1992). The fates
and effects of riverine and shelf-derived DOM on Mississippi River plume/Gulf shelf
processes. In: Nutrient Enhanced Coastal Ocean Productivity. NCOP Workshop
Proceeding, October 1991, NOAA Coastal Ocean Program, pp. 84-89, TAMU-SG-92109, 1992.
Benner, R., and S. Opsahl. (2001). Molecular indicators of the sources and
transformations of dissolved organic matter in the Mississippi river plume. Organic
Geochem., 32:597-611.
Berner, T., K. Wayman and P. G. Falkowski. (1989). Photoadaptation and the
“package” effect in Dunaliella tertiolecta (Chlorophyceae). J. Phycol., 25:70-78.
Berthon, J. F., and A. Morel. (1992). Validation of spectral light-photosynthesis
model and use of the model in conjunction with remotely sensed pigment observation.
Limnol. Oceanogr., 37:781-796.
Bianchi, T. S., J. R. Pennock, and R. R. Twilley. (1999). Biochemistry of Gulf of
Mexico Estuaries. John Wiley. New York. 428 pp.
Bidigare, R. R., M. E. Ondrusek, J. H. Morrow and D. A. Kiefer. (1990). In vivo
absorption properties of algal pigments. Ocean Optics X, Proc. SPIE Int. Soc. Opt. Eng.,
1302:290-302.
Biggs, D. C., R. R. Leben, and J. G. Ortega-Ortiz. (2000). Ship and satellite studies
of mesoscale circulation and sperm whale habitats in the Northeast Gulf of Mexico
during GulfCet II. Gulf of Mex. Sci., 18:15-22.
Biggs, D. C. (1992). Nutrients, plankton, and productivity in a warm-core ring in
the western Gulf of Mexico. J. Geophys. Res., 97:2143-2154.
Biggs, D. C., and F.E. Muller-Karger. (1994). Ship and satellite observations of
chlorophyll-a stocks in interacting cyclone-anticyclone pairs in the western Gulf of
Mexico. J. Geophys. Res., 99:7371-7384.
133
Bishop, J. K. B., M. H. Conte, P. H. Wiebe, M. R. Roman, and C. Langdon.
(1986). Particulate matter production and consumption in deep mixed layers:
observations in a warm-core ring. Deep Sea Res., 33(11/12):1813-1841.
Blough, N. V., and S. A. Green. (1995). Spectroscopic characterization and remote
sensing of non-living organic matter. In: The role of non-living organic matter in the
earth’s carbon cycle. R. G. Zepp and C. Sonntag (Eds.), pp. 23-45, John Wiley & Sons.
Blough, N. V., O. C. Zafiriou, and J. Bonilla. (1993). Optical absorption spectra of
waters from the Orinoco river outflow: Terrestrial input of colored organic matter to the
Caribbean. J. Geophys. Res., 98:2271-2278.
Brainerd, K.E. and M.C. Gregg. (1995). Surface mixed and mixing layer depths.
Deep Sea Res. I, 42(9):1521-1543.
Bricaud, A., M. Babin, A. Morel and H. Claustre. (1995). Variability in the
chlorophyll-specific absorption coefficients of natural phytoplankton: Analysis and
parameterization. J. Geophys. Res., 100:13,321-13,332.
Bricaud, A. and D. Stramski. (1990). Spectral absorption coefficients of living
phytoplankton and nonalgal biogenous matter: A comparison between the Peru upwelling
area and the Sargasso Sea. Limnol. Oceanogr., 35(3):562-582.
Bricaud, A., A. L. Bedhomme and A. Morel. (1988). Optical properties of diverse
phytoplanktonic species: Experimental results and theoretical interpretation. J. Plankton
Res., 10:851-873.
Bricaud, A., A. Morel and L. Prieur. (1983). Optical efficiency factors of some
phytoplankters. Limnol. Oceanogr., 28:816-832.
Bricaud, A., A. Morel, and L. Prieur. (1981). Absorption by dissolved organic
matter in the sea (yellow substance) in the UV and visible domains. Limnol. Oceanogr.,
26:43-53.
Burton, J. D., and P. S. Liss. (1976). Estuarine chemistry. Academic press.
Butler, W. L. (1962). Absorption of light by turbid materials. J. Opt. Soc. Am.,
52:292-299.
Carder, K. L., F. R. Chen, Z. P. Lee, S. K. Hawes, and D. Kamykowski. (1999).
Semianalytic moderate-resolution imaging spectrometer algorithms for chlorophyll-a-a
and absorption with bio-optical domains based on nitrate depletion temperatures. J.
Geophys. Res., 104:5403-5421.
134
Carder, K. L., S. K. Hawes, K. A. Baker, R. C. Smith, R. G. Steward and B. G.
Mitchell. (1991). Reflectance model for quantifying chlorophyll a in the presence of
productivity degradation products. J. Geophys. Res., 96:20,599-20,611.
Carder, K. L., R. G. Steward, G. R. Harvey, and P. B. Ortner. (1989). Marine humic
and fulvic acids: Their effects on remote sensing of ocean chlorophyll. Limnol.
Oceanogr., 34:68-81.
Chuang, W. S., W. W. Schroeder, and W. J. Weisman, Jr. (1982). Summer current
observations off the Alabama coast. Contrib. Mar. Sci., 25:121-131.
Ciotti, A. M., M. R. Lewis and J. J. Cullen. (2002). Assessment of relationship
between dominant cell size in natural phytoplankton communities and the spectral shape
of the absorption coefficient. Limnol. Ocenaogr., 47:404-417.
Cleveland, J. S. (1995). Regional models for phytoplankton absorption as a
function of chlorophyll a concentration. J. Geophys. Res., 100(C7):13,333-13,344.
Coble, P. G., C. E. Del Castillo, and B. Avril. (1998). Distribution and optical
properties of CDOM in the Arabian Sea during the 1995 Southwest Monsoon. Deep Sea
Research Part II, 45:2195–2223.
Daley, R. (1991). Atmospheric Data Analysis. Cambridge Univ. Press, New York,
477 pp.
Del Castillo, C. E., P. G. Coble, R. N. Conmy, F. E. Muller-Karger, L.
Vanderbloemen, and G. A. Vargo. (2001). Multispectral in situ measurements of organic
matter and chlorophyll fluorescence in seawater: Documenting the intrusion of the
Mississippi River plume in the West Florida Shelf. Limnol. Oceanogr., 46(7):1836-1843.
Del Castillo, C. E., F. Gilbes, P. G. Coble, and F. E. Muller-Karger. (2000). On the
dispersal of riverine colored dissolved organic matter over the West Florida Shelf,
Limnol. Oceanogr., 45(6):1425-1432.
Deegan, L. A., J. W. Day, Jr., J. G. Gosselink, A. Yanez-Arancibia, G. SoberonChavez, and P. Snachez-Gil. (1986). Relationships among physical characteristics,
vegetation distributions and fisheries yield in Gulf of Mexico Estuaries, p. 83-100. In:
Estuarine variability. D. A. Wolfe(ed.). Academic Press.
D’Sa, E. J., C. Hu, F. E. Muller-Karger, and K. L. Carder. (2002). Estimation of
colored dissolved organic matter and salinity fields in case 2 waters using SeaWiFS:
Examples from Florida Bay and Florida Shelf. Proc. Indian Acad. Sci. (Earth Planet Sci.),
111(3):197-207, Sept., 2002.
135
Dubinsky, Z., P. G. Falkowski and K. Wyman. (1986). Light harvesting and
utilization by phytoplankton. Plant and Cell Physiology, 27:1335-1349.
Duysens, L. N. M. (1956). The flattening of the absorption spectrum of
suspensions, as compared to that of solutions. Biochim. Biophys. Acta, 19:1-12.
Eigenheer, A., W. Kuhn, and G. Radach. (1996). On the sensitivity of ecosystem
box model simulations on mixed-layer depth estimates. Deep-Sea Research I,
43(7):1011-1027.
Ekman, V. W. (1905). On the influence of the earth’s rotation on ocean currents.
Ark. Mat. Astron. Fys. 2(11):1-52.
Ferrari, G. M., and M. D. Dowell. (1998). CDOM absorption characteristics with
relation to fluorescence and salinity in coastal areas of the Southern Baltic Sea.
Estuarine, Coastal and Shelf Science, 47:91-105.
Fofonoff, N. P. and R. C. Millard, Jr. (1983). Algorithms for computation of
fundamental properties of seawater. Unesco. Technical papers in marine science 44.
Paris, France. 53 pp.
Fogg, F. E., and G. T. (1958). Boalch, Extracellular products in pure cultures of
brown alga. Nature, 181:789-790.
Fujiki, T., and S. Taguchi. (2002). Variability in chlorophyll a specific absorption
coefficient in marine phytoplankton as a function of cell size and irradiance. J. Plankton
Res., 24(9):859-874.
Gilbes, F., F. E. Muller-Karger, and C. E. Del Castillo. (2002). New evidence for
the West Florida Shelf plume. Continental Shelf Res., 22:2479-2496.
Gilbes, F., C. Thomas, J. J. Walsh, and F. E. Muller-Karger. (1996). An episodic
chlorophyll-a plume on the west Florida shelf. Continental Shelf Res. 16(9):1201-1224.
Golubev, Y., and Y. Hsueh, Low-frequency variability of surface currents on the
northeastern Gulf of Mexico (NEGOM) shelf, submitted.
Guo, L., P. H. Santschi, and K. W. Warnken. (1995). Dynamics of dissolved
organic carbon (DOC) in oceanic environments, Limnol. Oceanogr., 40(8):1392-1403.
Gordon, H. R. and Wang, M. (1994). Retrieval of water-leaving radiance and
aerosol optical thickness over the oceans with SeaWiFS: A Preliminary Algorithm.
Applied Optics, 33(3):443-451.
136
He, R. and R. H. Weisberg. (2002). West Florida shelf circulation and temperature
budget for the 1999 spring transition. Continental Shelf Res., 22:719-748.
Hellerman, S., M. Rosenstein. (1983). Normal monthly wind over the world ocean
with error estimates. J. Phys. Oceanogr. 13:1093-1104.
Hitchcock, G. L., W. J. Wiseman Jr., W. C. Boicourt, A. J. Mariano, N. Walker, T.
A. Nelsen and E. Ryan. (1997). Property fields in an effluent plume of the Mississippi
river. J. of Marine Sci., 12:109-126.
Hsu, S. A. (1986). A mechanism for the increase of wind stress (drag) coefficient
with wind speed over water surfaces: a parametric model. J. Phys. Oceanogr., 16:144150.
Hochman, H. T., F. E. Muller-Karger, and J. J. Walsh. (1994). Interpretation of the
Coastal Zone Color Scanner signature of the Orinoco River plume, J. Geophys. Res.,
99(C4):7443-7455.
Hochman, H., J. J. Walsh, K. L. Carder, A. Sournia, and F. E. Muller-Karger.
(1995). Analysis of ocean color components within stratified and well-mixed waters of
the western English channel. J. Geophys. Res., 100:10777-10787.
Hoepffner, N., and S. Sathyendranath. (1992). Bio-optical characteristics of
coastal waters: Absorption spectra of phytoplankton and pigment distribution in the
western North Atlantic. Limnol. Oceanogr., 37:1660-1679.
Hoepffner, N., and S. Sathyendranath. (1991). Effect of pigment composition on
absorption properties of phytoplankton. Mar. Ecol. Prog. Ser., 73:11-23.
Holm-Hansen, O., and B. Riemann. (1978). Chlorophyll a determination:
improvements in methodology. Oikos, 30:438-447.
Hu, C., E. T. Montgomery, R. W. Schmitt, and F. E. Muller-Karger. 2003. (in
press). The dispersal of the Amazon and Orinoco River water in the tropical Atlantic and
Caribbean Sea: Observation from space and S-PALACE floats, Deep-Sea Research II.
Hu, C., F. E. Muller-Karger, D. C. Biggs, K. L. Carder, B. Nababan, D. Nadeau,
and J. Vanderbloemen. (2003). Comparison of ship and satellite bio-optical
measurements o the continental margin of the NE Gulf of Mexico. Int. J. Rem. Sensing,
24(13):2597-2612.
Huh, O. K., W. J. Wiseman, and L. J. Rouse. (1981). Intrusion of Loop Current
waters onto the West Florida shelf. Journal of Geophysical Research, 86(C5):4,1864,192.
137
Huyer, A. (1983). Coastal upwelling in the California Current. Prog. Oceanogr.,
12:259-284.
JGOFS. (1994). Protocols for the Joint Global Ocean Flux Study (JGOFS) Core
Measurements. UNESCO. Manual and Guides, 29:91-100, 104-118.
Jochens, A. E., S. F. DiMarco, W. D. Nowlin, Jr., R. O. Reid, and M. C. Kennicutt
II. (2002). Northeastern Gulf of Mexico Chemical Oceanography and Hydrography
Study: Synthesis Report, U. S. Department of the Interior, Minerals Management
Service, Gulf of Mexico OCS Region, New Orleans, LA., OCS Study MMS 2002-055.
586 pp.
Jochens, A. E. and W. D. Nowlin, Jr., eds. (1999). Northeastern Gulf of Mexico
Chemical Oceanography and Hydrography Study: Year2-Annual Report. OCS Study
MMS 99-0054, U.S. Dept. of the Interior, Minerals Management Service, Gulf of Mexico
OCS Region, New Orleans, LA. 123 pp.
Kiefer, D. A., and B. G. Mitchell. (1983). A Simple, steady state description of
phytoplankton growth based on absorption cross section and quantum efficiency.
Limnol. Ocenaogr., 28:770-776.
Kiefer, D. A., and J. B. SooHoo. (1982). Spectral absorption by marine particles of
coastal waters of Baja California. Limnol. Oceanogr., 27:492-499.
Kirk, J. T. O. (1994). Light and hotosynthesis in aquatic ecosystem. 2nd ed.
Cambridge University Press. Cambridge, 509 pp.
Kishino, M., M. Takahashi, N. Okami and S. Ichimura. (1985). Estimation of the
spectral absorption coefficients of phytoplankton in the sea. Bull. Mar. Sci., 37:634-642.
Klinkhammer, G. P., J. McManus, D. Colbert, and M. D. Rudnicki. (2000).
Behavior of terrestrial dissolved organic matter at the continent-ocean boundary from
high-resolution distribution, Geochimica et Cosmochimica Acta, 64(16):2765-2774.
Kosro, P. M., A. Huyer, S. R. Ramp, R. L. Smith, F. P. Chavez, T. J. Cowles, M.
P. Abbott, P. T. Strub, R. T. Barber, P. Jessen, and L. F. Small. (1991). The structure of
the transition zone between coastal waters and the open ocean off northern California,
winter and summer 1987. J. Geophys. Res., 96:14,707-14,730.
Lazzara, L., A. Bricaud, and H. Claustre. (1996). Spectral absorption and
fluorescence excitation properties of phytoplanktonic populations at a mesotrophic and an
oligotrophic site in the tropical North Atlantic (EUMELI program). Deep-Sea Res. I,
43:1215-1240.
138
Lagerloef, G., C. Swift and D. LeVine. (1995). Sea surface salinity: The next
remote sensing challenge. Oceanography, 8:44-50.
Leben, R. O., G. H. Born, and B. R. Engebreth. (2002). Operational altimeter data
processing for mesoscale monitoring. Marine Geodesy, 25:3-18.
Lillibridge, J., R. Leben, and F. Vossepoel. (1997). Real-time altimetry from ERS2, Proc. Third ERS Symp., Vol. 3, Florence, Italy, European Space Agency SP.,
414:1449-1453.
Livingston, R. J., R. L. Iverson, R. H. Estabrook, V. E. Keys, and J. Taylor.
(1975). Major features of the Apalachicola Bay system: physiography, biota, and
resource management. Florida Science, 37(4):217-244.
Lohrenz, S. E., A. D. Weidemann and M. Tuel. (2003). Phytoplankton spectral
absorption as influenced by community size structure and pigment composition. J.
Plankton Res., 25(1):35-61.
Lohrenz, S. E. (2000). A novel theoretical approach to correct for path-length
amplification and variable sample loading in measurements of particulate spectral
absorption by the quantitavie filter technique. J. Plankton Res., 22:639-657.
Lohrenz, S. E., M. J. Dagg, and T. E. Whitledge. (1990). Enhance primary
production at the plume/oceanic interface of the Mississippi River. Continental Shelf
Research, 88(C7):639-664.
Mason, J. E. and A. Bakun. (1986). Upwelling index update, U.S. west coast, 33N48N latitude. U.S. Dep. Commer., NOAA Tech. Memo., NOAA-TM-NMFS-SWFC-67,
81 pp.
McClain, E. P., W. G. Pichel, and C. C. Walton. (1985). Comparative performance
of (AVHRR) based multichannel sea surface temperatures, J. Geophys. Res., 90:11,58711,601.
McClain, E. P., W. G. Pichel, C.C. Walton, Z. Ahmad, and J. Sutton. (1983). Multichannel improvements to satellite-derived global sea-surface temperatures. Adv. Space
Res., 2(6), 43-47.
McGowan, J. A., and T. L. Hayward. (1978). Mixing and oceanic productivity.
Deep Sea Res., 25:771-793.
McPherson, B. F., R. T. Montgomery, and E. E. Emmons. (1990). Phytoplankton
productivity and biomass in the Charlotte Harbor estuarine system, Florida. Water
Resources Bulletin, 26(5):787-800.
139
Mitchell, B. G., and D. A. Kiefer. (1988a). Chlorophyll a-specific absorption and
fluorescence excitation spectra for light-limited phytoplankton. Deep Sea Res., Part A.,
35:639-663.
Mitchell, B. G., and D. A. Kiefer. (1988b). Variability in pigment specific
particulate fluorescence and absorption spectra in the northeastern Pacific Ocean. Deep
Sea Res., Part A., 35:665-689.
Morel, A., Y. Ahn, F. Partensky, D. Vaulot and H. Claustre. (1993). Prochlorococcus and Synechococcus: A comparative study of their optical properties in relation to
their size and pigmnetation. J. Mar. Res., 51:617-649.
Moore, L. R., R. Goericke and S. W. Chisholm. (1995). Comparative physiology
of Synechococcus and Prochlorococcus: influence of light and temperature on growth,
pigments, fluorescence and absorptive properties. Mar. Ecol. Prog. Ser., 116:259-275.
Milliman, J. D., and R. H. Meade. (1983). World-wide delivery of river sediment
of the oceans. Journal of Geology, 91:1-21.
Mitchell, B. G., and D. A. Kiefer. (1988). Chlorophyll-a a specific absorption and
fluorescence excitation spectra for light-limited phytoplankton. Deep Sea Res.,
35(5):639-663.
Mobley, C. D. (1994). Light and Water. Academic Press, Inc. San Diego. 592 pp.
Momzikoff, A., S. Dallot, and G. Gondry. (1994). Distribution of seawater
fluorescence and dissolved flavins in the Almeria-Orant front (Alboran Sea, Western
Mediterranean Sea). J. Mar. Sys., 5:361-376.
Montague, C. L., and J. A. Ley. (1993). A possible effects of salinity fluctuation
on abundance of benthic vegetation and associated fauna in northeastern Florida Bay.
Estuaries. 16:703-717.
Moran, M. A., and R. G. Zepp. (1997). Role of photoreactions in the formation of
biologically labile compounds from dissolved organic matter. Limnol. Oceanogr., 42:
1307-1316.
Morel A., and L. Prieur. (1997). Analysis of variations in ocean color. Limnol.
Oceanogr., 22:709-722.
Muller-Karger, F. E. (2000). The spring 1998 Northeastern Gulf of Mexico
(NEGOM) cold water event: Remote sensing evidence for upwelling and for eastward
advection of Mississippi water (or: How an errant Loop Current Anticyclone took the
NEGOM for a spin). Gulf of Mex. Sci., 18(1):55-67.
140
Müller-Karger, F. E., J. J. Walsh, R. H. Evans, and M. B. Meyers. (1991). On the
Seasonal Phytoplankton Concentration and Sea Surface Temperature Cycles of the Gulf
of Mexico as Determined by Satellites. J. Geophys. Res., 96(C7):12645-12665.
Muller-Karger, F. E., F. Vukovich, R. Leben, B. Nababan, C. Hu and D. Myhre.
(2001). Surface circulation and the transport of the Loop Current into the Northeastern
Gulf of Mexico: Final Report. OCS Study MMS 2001-102, U.S. Dept. of the Interior,
Minerals Management Service, Gulf of Mexico OCS Region, New Orleans, La. 39 pp.
Murray, S. P. (1972). Observations on wind, tidal and density-driven currents in
the vicinity of the Mississippi River delta. In: Swift, Duane, and Pilkey (eds.). Shelf
sediment transport. Dowden, Hutchinson and Ross, Stroudsburg, PA, pp. 127-142.
National Oceanic and Atmospheric Administration (NOAA). (1985). The Gulf of
Mexico coastal and ocean zone strategic assessment: Data Atlas. Strategic Assessment
Branch, Ocean Assessment Division.
Nelson, D. M., and Q. Dortch. (1996). Silicic acid depletion and silicon limitation
in the plume of the Mississippi River: Evidence from kinematic studies in spring and
summer. Marine Ecology Progress Series, 136:136-178.
Neslon, J. R., and S. Guarda. (1995). Particulate and dissolved spectral absorption
on the continental shelf of the southeastern United States. J. Geophys. Res.,
100(C5):8715-8732.
Nelson, N. B., B. B. Prezelin, and R. R. Bidigare. (1993). Phytoplankton light
absorption and the package effect in California coastal waters. Mar. Ecol. Prog. Ser.,
94:217-227.
Nelson, D. M. and W. O. Smith, Jr. (1991). Sverdrup revisited: Critical depths,
maximum chlorophyll-a levels, and the control of Southern Ocean productivity by the
irradiance-mixing regime. Limnol. Oceanogr., 36(8):1650-1661.
Nowlin,Jr., W. D., A. E. Jochens, M. K. Howard, S. F. DiMarco, and W. W.
Schroeder. (2000). Hydrographic properties and inferred circulation over the
Northeastern Shelves of the Gulf of Mexico during Spring to Midsummer of 1998. Gulf
of Mexico Science, 18(1):40-45.
141
O’Reilly, J. E., S. Maritorena, D. A. Siegel, M. C. O’Brien, D. Toole, B. G.
Mitchell, M. Kahru, F. P. Chavez, P. Strutton, G. F. Cota, S. B. Hooker, C. R. McClain,
K. L. Carder, F. Muller-Karger, L. H. Harding, A. Magnuson, D. Phinney, G. F. Moore,
J. Aiken, K. R. Arrigo, R. Letelier, and M. Culver. (2000). Ocean color chlorophyll-a a
algorithms for SeaWiFS, OC2, and OC4: Version 4. In: Hooker, S.B. and E. R. Firestone
(eds.): SeaWiFS postlaunch technical report series, Volume 11, SeaWiFS postlaunch
calibration and validation analyses, Part 3. Goddard Space Flight Center, Greenbelt,
Maryland. NASA/TM-2000-206892, Vol.11. pp. 9-23.
O’Reilly, J. E., S. Manitorena, B. G. Mitchell, D. A. Siegel, K. L. Carder, S. A.
Garver, and C. R. McClain. (1998). Ocean color chlorophyll-a algorithm for SeaWiFS.
J. Geophys. Res., 103(C11):24,937-24,953.
Ortner, P. B., T. N. Lee, P. J. Milne, R. G. Zika, M. E. Clarke, G. P. Podesta, P. K.
Swart, P. A. Tester, L. P. Atkinson, and W. R. Johnson. (1995). Mississippi River flood
waters that reached the Gulf Stream. J. of Geophys. Res., 100(C7):13,595-13,601.
Pennock, J. R., J. N. Boyer, J. A. Herrera-Silveira, R. L. Iverson, T. E. Whitledge,
B. Mortazavi, and F. A. Comin. (1999). Nutrient Behavior and phytoplankton
production in Gulf of Mexico estuaries. In: T.S. Bianchi, J.R. Pennock, and R.R. Twilley
(eds.). Biogeochemistry of Gulf of Mexico Estuaries. John Wiley and Sons, Inc. New
York. pp. 109-162.
Qian, Y., A. E. Jochens, M. C. Kennicutt, and D. C. Biggs. (2003). Spatial and
temporal variability of phytoplankton biomass and community structure over the
continental margin of the NEGOM based on pigment analysis. Continental Shelf Res., 23:
1-17.
Redalje, D. G., S. E. Lohrenz, and G. L. Fahnenstiel. (1994). The relationship
between primary production and the vertical export of particulate organic matter in a
riverine impacted coastal ecosystem. Estuaries, 17:829-838.
Rhodes, R. C., A. J. Wallcraft, and J. D. Thompson. (1985). Navy-corrected
geostrophic wind set for the Gulf of Mexico. NORDA Technical Note 310, Stennis
Space Center, MS., 103pp.
Roesler, C. S., M. J. Perry and K. L. Carder. (1989). Modeling in situ
phytoplankton absorption from total absorption spectra in productive inland marine
waters. Limnol. Ocenaogr., 34:1510-1523.
Salisbury, J. E., J. W. Campbell, E. Linder, L. D., Meeker, F. E. Muller-Karger, C.
J., Vorosmarty. (submitted). On the seasonal correlation of surface particle fields with
wind stress and Mississippi discharge on the northern Gulf of Mexico.
142
Sakshaug, E., A. Bricaud, Y. Dandonneau, P. G. Falkowski, D. A. Kiefer, L.
Legendre, A. Morel, J. Parslow and M. Takahashi. (1997). Parameter of photosynthesis:
Definition, theory and interpretation of results. J. Plankton Res., 19:1637-1670.
Schmidt, N., E. K. Lipp, J. B. Rose, and M. E. Luther. (2001). ENSO influences
on seasonal rainfall and river discharge in Florida. Journal of Climate, 14:615-628.
Schroeder, W. W., S. P. Dinnell, W. J. Wiseman, Jr., and W. J. Merrel. (1987).
Circulation patterns inferred from the movement of detached buoys in the eastern Gulf of
Mexico. Continental Shelf Res., 7(8):883-894.
Schwing, F. B., M. O’Farrell, J. Steger, and K. Baltz. (1996). Coastal upwelling
indices, west coast on North America 1946-1995. U.S. Dept. Commer., NOAA Tech.
Rep., NOAA-TM-NMFS-SWFSC-231.
Siegel, D. A., S. Maritorena, N. B. Nelson, D. A. Hansell, and M. Lorenzi-Kayser.
(2002). Global distribution and dynamics of colored dissolved and detrital organic
materials. J. of Geophys. Res., 107(C12):3228-3242.
Siegel, D. A., and A. F. Michaels. (1996). Quantification of non-algal attenuation
in the Sargasso Sea: Implication for biogeochemistry and remote sensing. Deep-Sea Res.
II., 43:321-345.
Skoog, A., P. O. Hall, S. Hulth, N. Paxeus, M. R. Van der Loeff, and S. Westerlund.
(1996). Early diagenetic production and sediment-water exchange of fluorescent
dissolved organic matter in the coastal environment. Geochimica et Cosmochimia Acta,
60(19): 3619-3629.
Sosik, H. M., and B. G. Mitchell. (1991). Absorption, fluorescence, and quantum
yield for growth in nitrogen-limited Dunaliella tertiolecta. Limnol. Oceanogr., 36:910921.
Sosik, H. M., and B. G. Mitchell. (1995). Light absorption by phytoplankton,
photosynthetic pigments and detritus in the California Current System. Deep-Sea
Research I, 42(10):1717-1748.
Stramski, D., A. Sciandra, and H. Claustre. (2002). Effects of temperature,
nitrogen, and light limitationon the optical properties of the marine diatom Thalassiosira
pseudonana. Limnol. Oceanogr., 47:392-403.
Stramski, D., and A. Morel. (1990). Optical properties of photosynthetic
picoplankton in different physiological states as affected by growth irradiance. Deep Sea
Res., Part A., 37:245-266.
143
Stuart, V., S. Sathyendranath, T. Platt, H. Maass, and B. D. Irwin. (1998).
Pigments and species composition of natural phytoplankton populations: effects on the
absorption spectra. J. Plankton Res., 20:187-217.
Suzuki, K., M. Kishino, K. Sasaoka, S. Saitoh, and T. Saino. (1998). Chlorophyllspecific absorption coefficients and pigments of phytoplankton off Sanriku, Northwestern
North Pacific. J. Oceanogr., 54:517-526.
Traganza, E.D., J.C. Conrad, and L.C. Breaker. (1981). Satellite observations of a
cyclonic upwelling system and giant plume in the California Current. In: F.A. Richards
(ed.), Coastal Upwelling. Amer. Geophys. Union. Washington, D.C. p.228-241.
Turner, R. E., and N. N. Rabalais, Changes in Mississippi River quality this
century. (1991). Bioscience, 3:140-147.
Twardowski, M. S., J. M. Sullivan, P. L. Donaghay, and J. R. V. Zaneveld. (1999).
Microscale Quantification of the Absorption by Dissolved and Particulate Material in
Coastal Waters with AC-9. J. of Atm. and Oceanic Tech., 16:691-707
Twilley, R. R., J. Cowan, T. Miller-Way, P. A. Montagna, and B. Mortazavi.
(1999). Benthic nutrient fluxes in selected estuaries in the Gulf of Mexico. In: T. S.
Bianchi, J. R. Pennock, and R. R. Twilley (eds.). Biogeochemistry of Gulf of Mexico
Estuaries. John Wiley and Sons, Inc. New York. pp. 163-209.
Vidussi, F., H. Claustre, B. B. Manca, A. Luchetta, J-C. Marty. (2001).
Phytoplankton pigment distribution in relation to the upper thermocline circulation in the
eastern Mediterranean Sea during winter. J. Geophys. Res., 106:19,939-19,956.
Vodacek, A., N. V. Blough, M. D. DeGrandpre, E. T. Peltzer, and R. K. Nelson.
(1997). Seasonal variation of CDOM and DOC in the Middle Atlantic Bight: Terrestrial
inputs and photooxidation. Limnol. Oceanogr., 42(4):674-686.
Vukovich, F. M. and P. Hamilton. (1989). New Atlas of Front Locations in the
Gulf of Mexico, Proceedings of the MMS Information Transfer Meeting, New Orleans,
Louisiana, May 1989.
Vukovich, F. M. (1988). Loop Current Boundary Variations. J. Geophys. Res.,
93(C12):15,585-15,591.
Vukovich, F. M. and P. Hamilton. (1989). New Atlas of Front Locations in the Gulf
of Mexico. Proceedings of the MMS Information Transfer Meeting, New Orleans,
Louisiana, May 1989.
144
Vukovich, F. M., B. W. Crissman, M. Bushnel, and W. J. King. (1979). Some
aspects of the oceanography of the Gulf of Mexico using satellite and in situ data. J.
Geophys. Res., 84(C12): 7749-7768.
Walker, N. D. (1996). Satellite assessment of Mississippi River plume variability:
causes and predictability. Remote Sensing Environment, 58:21-35.
Walker, N. D., G. S. Fargion, L. J. Rouse, and D. C. Biggs. (1994). The great flood
of summer 1993: Mississippi River discharge studies. Eos Trans. AGU, 75:409-411.
Wang, O., R. O. Reid, S. F. DiMarco, and W. D. Nowlin, Jr. (2002). Diagnostic
calculations of circulation over the northeastern Gulf of Mexico. TAMU Oceanography
Technical Report No. 02-1-D. Texas A&M University, College Station, Texas, 2002.
Weisberg, R. H., and R. He. (2003). Local and deep-ocean forcing contributions to
anomalous water properties on the West Florida Shelf. J. Geophys. Res., 108(C6): 3184,
doi:10,1029/2002JC001407.
Weisberg, R. H., B. D. Black, and Z. Li. (2000). An upwelling case study of
Florida's west coast, J. Geophys. Res., 105(C5):11,459-11,469.
Weisberg, R. H., B. D. Black, and H. Yang. (1996). Seasonal modulation of the
west Florida continental shelf circulation. Geophys. Res. Let., 23(17):2247-2250.
Walsh, J. J., D. A. Dieterle, M. B. Meyers, and F. E. Muller-Karger. (1989).
Nitrogen exchange at the continental margin: a numerical study of the Gulf of Mexico.
Prog. Oceanogr. 23:245-301.
Walton, C. C. (1988). Nonlinear multichannel algorithms for estimating sea surface
temperature with AVHRR satellite data. JAM, 27:115-27.
Wick, G. A., W. J. Emery, and P. Schluessel. (1992). A comprehensive comparison
between satellite-measured skin and multichannel sea surface temperature. J. Geophys.
Res., 97(C4):5569-5595.
Wozniak, B., and M. Ostrowska. (1990). Optical absorption properties of
phytoplankton in various seas. Oceanologia, 29:117-146.
Yang, H., and R. H. Weisberg. (1999). Response of the West Florida Shelf to
climatological wind stress forcing. J. Geophys. Res. 104(C3):5301-5320.
Yentsch, C. S., and D. A. Phinney. (1989). A bridge between ocean optics and
microbial ecology. Limnol. Oceanogr., 34(8):1694-1705.
145
Yentsch, C. S. (1962). Measurement of visible light absorption by particulate
matter in the ocean. Limnol. Oceanogr., 7:207-217.
Yentsch, C. S., and C. A. Reichert. (1962). The interrelationship between watersoluble yellow substances and chloroplastic pigments in marine algae. Bot. Mar., 3:6574.
Zweifel, U. L., J. Wikner, A. Hagstrom, E. Lundberg, and B. Norman, Dynamics
of dissolved organic carbon in a coastal ecosystem. (1995). Limnol. Oceanogr., 40(2):
299-305.
146