JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 114, C03023, doi:10.1029/2008JC004934, 2009
A study of a Hurricane Katrina–induced phytoplankton bloom using
satellite observations and model simulations
Xiaoming Liu,1,2 Menghua Wang,1 and Wei Shi1,3
Received 27 May 2008; revised 31 December 2008; accepted 13 January 2009; published 27 March 2009.
[1] Satellite observations of sea surface temperature (SST) and chlorophyll-a
concentration from the Moderate Resolution Imaging Spectroradiometer (MODIS) on
Aqua revealed a phytoplankton bloom event in the vicinity of the Loop Current (LC) in
the Gulf of Mexico after Hurricane Katrina’s passage in August 2005. The sea surface
height (SSH) anomaly data derived from satellite altimetry measurements indicated that
the phytoplankton bloom coincided with a LC frontal eddy and thus suggested that the
cold-core eddy may have played a critical role in the phytoplankton bloom event. Using
the Hybrid Coordinate Ocean Model (HYCOM), we have carried out numerical
simulations to understand the mechanism that triggers the phytoplankton bloom after
Katrina passage. The model study shows that the preexisting cold-core eddy strengthens
the upper ocean dynamics and nutrient responses with significant increase of nutrient
concentration after the passage of Hurricane Katrina, because of the uplift of nutrient fields
within the cold-core eddy. The Hurricane Katrina–triggered phytoplankton bloom is
attributed to the preexisting cold eddy as well as the slowly passing hurricane. The nutrient
distribution after Hurricane Katrina indicates that the increase of nitrate concentration
plays a dominant role in the development of the phytoplankton bloom.
Citation: Liu, X., M. Wang, and W. Shi (2009), A study of a Hurricane Katrina – induced phytoplankton bloom using satellite
observations and model simulations, J. Geophys. Res., 114, C03023, doi:10.1029/2008JC004934.
1. Introduction
[2] Early observations of upper ocean response to hurricanes have been reported and studied by Leipper [1967],
Brooks [1983], Sanford et al. [1987], Shay and Elsberry
[1987], Shay et al. [1989, 1998], Dickey et al. [1998], and
Jacob et al. [2000]. Hurricanes are known to lower sea
surface temperature (SST) significantly during its passage,
and this upper ocean process depends on many factors
such as mixed layer depth, thermocline depth, air-sea heat
fluxes, and the storm intensity and transition speed. Threedimensional ocean numerical models were also used to study
the process [Price, 1981; Price et al., 1994; Prasad and
Hogan, 2007]. Entrainment has been emphasized as the
dominant term in lowering the SST beneath a moving
hurricane. On the basis of observations, Jacob et al. [2000]
suggested that entrainment at the mixed layer base generally
accounts for 75– 90% of the cooling, while Price [1981]
model results indicated that 85% of heat flux into the
mixed layer was through entrainment. Price [1981] also
concluded that upwelling causes a significant enhancement
of the SST response to a slowly moving hurricane (4 m/s).
1
Center for Satellite Applications and Research, National Environmental Satellite, Data, and Information Service, NOAA, Camp Springs,
Maryland, USA.
2
SP Systems Incorporated, Greenbelt, Maryland, USA.
3
CIRA, Colorado State University, Fort Collins, Colorado, USA.
Copyright 2009 by the American Geophysical Union.
0148-0227/09/2008JC004934
In addition to the SST cooling, hurricane-induced vertical
mixing and upwelling can also bring up the nutrient-rich
water to the upper euphotic zone and thus stimulate biological production. With the advancement of satellite ocean
color remote sensing from Sea-viewing Wide Field-of-view
Sensor (SeaWiFS) and Moderate Resolution Imaging Spectroradiometer (MODIS), hurricane-induced phytoplankton
blooms were also observed and reported in the past few
years [Babin et al., 2004; Lin et al., 2003; Miller et al.,
2006; Walker et al., 2005]. Most recently, Shi and Wang
[2007] reported a Hurricane Katrina –induced SST cooling
and a phytoplankton bloom in the Gulf of Mexico in August
2005. In their study, MODIS-Aqua data were processed
using a recently developed short-wave infrared (SWIR)
atmospheric correction algorithm [Wang, 2007; Wang and
Shi, 2005]. It has been demonstrated that, comparing with
the NASA standard algorithm [Gordon and Wang, 1994],
using the SWIR algorithm ocean color products can be
improved over the productive ocean waters [Wang et al.,
2007; Wang et al., 2009]. Therefore, the phytoplankton
bloom can be distinguished from sediment resuspension
and transportation, which can also be induced by hurricane
in coastal waters. A notable phytoplankton bloom centered
at a location of (24°N, 84°W) was observed 4 days after
Katrina passage [Shi and Wang, 2007].
[3] The Gulf of Mexico is greatly influenced by its
unique ocean circulation feature, i.e., the Loop Current
(LC). Warm water from the Caribbean Sea enters the Gulf
of Mexico through the Yucatan Straits and then forms the
LC. The LC penetrates northward into the Gulf of Mexico
C03023
1 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
and then turns anticyclonically to flow southward along the
steep west Florida shelf before turning eastward into the
southern Straits of Florida (SSF). The Straits of Florida
forms a conduit for the Florida Current (FC), which connects the LC with the Gulf Stream in the North Atlantic.
The instability processes that are associated with LC shed
warm core eddies (WCE), which propagate westward and
eventually dissipate along the shelf break. Cyclonic, coldcore eddies (CCE) are also formed along the shoreward
boundary of the LC [Paluszkiewicz et al., 1983; Vukovich,
1988; Fratantoni et al., 1998], and they are called Loop
Current Frontal Eddies (LCFE). LCFE is one type of CCE
that occurs at LC boundary. LCFEs travel rapidly southward
downstream with the LC at speeds of 20– 50 cm/s. Since the
preexisting oceanic features such as LC, WCE and CCE can
modulate the upper ocean vertical and horizontal structures
and heat balance, they are also very important in the oceanic
response process to hurricanes. When a hurricane enters the
warm (cold) core eddy, SST cooling may be greatly suppressed (intensified) due primarily to deep (shallow) mixed
layer and thermocline depth.
[4] Shi and Wang [2007] used passive multiple-sensor
remote sensing data to analyze the physical, optical, and
biological responses. The phytoplankton bloom is attributed
to the vertical mixing and upwelling caused by the slow
moving hurricane at the bloom location near (24°N, 84°W).
However, the effects of the passing hurricane could not
completely explain why the phytoplankton bloom developed at this location, especially because it was on the left
side of the storm track. To the right of the hurricane track,
chlorophyll-a concentration shows modest increases as the
storm moved slowly forward and the sustained winds
strengthened before the Gulf coast landfall.
[5] In this paper, we present some evidence from sea
surface height (SSH) anomaly data derived from satellite
altimetry measurements, which show the Hurricane
Katrina – induced phytoplankton bloom near (24°N, 84°W)
was coincident with a cyclonic eddy near Dry Tortugas.
These observations indicate that the cyclonic eddy may
have played a critical role in the phytoplankton bloom
event. Near the Dry Tortugas, the Tortugas gyre often
occurs when the southward flowing LC overshoots entry
into the Straits of Florida and approaches the coast of Cuba,
before abruptly turning to the east [Lee et al., 1995]. This
offshore position of the Florida Current combined with the
strong cyclonic curvature of the flow field can cause a
counterclockwise recirculation off the Tortugas, i.e., the
‘‘Tortugas gyre’’ [Lee et al., 1995]. Fratantoni et al.
[1998] used a combination of numerical model experiments
with time series of satellite thermal images to show that the
evolution of Tortugas gyres depend strongly on the variability of the LC and LCFEs. Romanou et al. [2004]
reported that in their numerical simulations cyclone eddies
occur very often near Dry Tortugas. They can be either
generated locally or evolved from LCFEs ending up there
after moving around LC boundary. In a cyclonic eddy,
upwelling can bring cold nutrient-rich water up from the
deep layer to the euphotic zone, and thus stimulate primary
production. It has been reported that Hurricane-forced
upwelling and chlorophyll-a concentration increase could
be enhanced within cold-core cyclones in the Gulf of
Mexico [Walker et al., 2005].
C03023
[6] The purpose of this study is to identify the mechanism
that triggers the phytoplankton bloom near (24°N, 84°W)
after the passage of hurricane. The process of upper ocean
response to Hurricane Katrina is simulated using the Hybrid
Coordinate Ocean Model (HYCOM) [Bleck, 2002]. Particularly, two experiments with and without the presence of a
preexisting cyclone eddy near Dry Tortugas are simulated
and the results are compared. In addition to the standard
physical parameters such as temperature and salinity in
HYCOM, three nutrient parameters, nitrate, phosphate,
and silicate, are also included in the simulation to study
the process of the Katrina-induced nutrient enhancement in
the upper ocean.
2. Satellite Observations
2.1. Hurricane Katrina
[7] Hurricane Katrina formed as a tropical depression
over the southeastern Bahamas on 23 August 2005, and
was then upgraded to tropical storm status on 24 August.
The tropical storm continued to move toward Florida, and
became a hurricane near 2100 UTC on 25 August 2005,
only 2 h before it made landfall on the southeastern coast of
Florida. The storm weakened over land, but it regained
hurricane status about 1 h after entering the Gulf of Mexico.
The storm rapidly intensified after entering the Gulf because
of the storm’s movement over the LC warm waters. On
27 August, the storm reached Category 3 intensity, becoming the third major hurricane of the season. Katrina again
rapidly intensified, attaining Category 5 status on 28 August.
The pressure measurement made Katrina the fourth most
intense Atlantic hurricane on record, and it was also the
strongest hurricane ever recorded in the Gulf of Mexico at
the time. Katrina made its second landfall on 29 August as
a Category 3 hurricane with sustained winds of 125 mph
(205 km/h) near Buras-Triumph, Louisiana.
2.2. Satellite SST Measurements
[8] Figures 1a – 1c show the MODIS SST before and after
Hurricane Katrina’s arrival. On 22 August, Gulf of Mexico
was characterized by warm SST predominantly above 30°C
(Figure 1a). Shi and Wang [2007] reported a maximum SST
drop of 6 – 7°C near (24°N, 84°W) on 29 August from
Advanced Microwave Scanning Radiometer EOS (AMSR-E)
SST. For that specific day, MODIS SST data were not
available because of cloud cover. On 31 August (Figure 1b),
4 days after Katrina passage, a cold SST patch was observed
near (24°N, 84°W) from MODIS SST, and another cold
SST patch was also observed near Katrina landfall in the
Louisiana coast. Overall, the SST dropped at a magnitude of
1– 2°C in the Gulf of Mexico after Katrina passage [Shi
and Wang, 2007]. Our particular interest is in the cold SST
patch around (24°N, 84°W), where a phytoplankton bloom
is also found from satellite-derived chlorophyll-a concentration. The cold SST patch moved eastward after Katrina
passage and reached (23.8°N, 84°W) on 7 September
(Figure 1c).
2.3. Satellite SSH Anomaly Data
[9] The MODIS SST in the eastern Gulf of Mexico did
not indicate the presence of mesoscale features like warmor cold-core eddies before Hurricane Katrina. However,
2 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
C03023
Figure 1. Satellite-derived (a – c) MODIS SST, (d – f) SSH Anomaly, and (g – i) MODIS chlorophyll-a
concentration on 22 August, 31 August, and 7 September 2005, respectively. SSH Anomaly data is from
Center for Astrodynamics Research (CCAR). Figure 1b also shows the Hurricane Katrina eye locations
(pink squares) with interval of every 6 h. Latitude and longitude for the region are indicated in Figure 1i.
these features were revealed in the SSH observations
provided by satellite altimeters. In a cyclonic eddy like
LCFE, the water column is denser than normal, which
results in a lower than average SSH or a negative SSH
anomaly. On the contrary, in an anticyclone eddy like LC
and WCE, there is a positive SSH anomaly. Figures 1d –1f
show the SSH anomaly maps of 22 August, 31 August, and
7 September, derived from satellite altimetry, which were
obtained from Colorado Center for Astrodynamics Research
(CCAR), University of Colorado. CCAR SSH anomaly data
sets are created using multiple satellite altimeter measurements from Jason-1, TOPEX/POSEIDON, Envisat, Geosat
Follow-On, and ERS-2. The SSH anomaly data clearly
indicate a positive anomaly in LC. On the eastside of LC,
there was a southward propagating LCFE (negative anomaly) moving downstream with LC. On 22 August, the
frontal eddy was at the location around (24.5°N, 84.5°W),
and it moved to around (24°N, 84°W) on 31 August, where
the phytoplankton bloom was observed. After Katrina
passage, the frontal eddy moved eastward downstream of
Florida Current to (23.8°N, 84°W) on 7 September.
2.4. Satellite Chlorophyll-a Data
[10] Figures 1g – 1i show the MODIS chlorophyll-a
concentration measured on 22 August, 31 August, and
7 September, respectively. A patch of high chlorophyll-a
concentration with value over 1.5 mg/m3 was observed at
(24°N, 84.2°W) on 31 August [Shi and Wang, 2007]. The
high chlorophyll-a concentration patch clearly showed a
cyclonic circulation pattern, of which the center coincided
with the center of the frontal eddy as indicated in the SSH
anomaly map. Furthermore, the patch of phytoplankton
bloom moved in the same direction and speed with the
frontal eddy: on 7 September, the eddy moved southeastward to (23.8°N, 84°W), and the phytoplankton bloom
patch also moved to the same location. The colocation of the
phytoplankton bloom and the cold-core eddy evidently
indicated that the phytoplankton bloom was related to the
preexisting cold-core eddy. In fact, it can be seen that
elevation of the chlorophyll-a concentration was not continuous along the Katrina track, but occurred primarily at the
location of the eddy, which suggested that the eddy may
have played a critical role in the bloom event.
3 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
[11] On the other hand, before Katrina’s arrival, the eddy
itself did not generate a phytoplankton bloom at the surface
as can be seen from the chlorophyll-a concentration map of
22 August. One possible mechanism of the phytoplankton
bloom is that the cold nutrient-rich water, which is normally
in the thermocline depth of 100 m, was brought up by the
eddy to the subsurface layer, but did not reach the surface
layer because of the strong stratification caused by the
summer heating. Indeed, it was the strong vertical mixing
and upwelling induced by Hurricane Katrina that further
brought the nutrient up to the surface, and thus stimulated
the phytoplankton bloom. Therefore, to simulate the oceanic
response process to Hurricane Katrina, the model must be
eddy-resolving, so that mesoscale cold-core eddies can be
included in the model.
3. Numerical Model and Experiment
3.1. Numerical Model
[12] The Hybrid-Coordinate Ocean Model (HYCOM)
[Bleck, 2002] is used in this study. HYCOM is a primitive
equation ocean general circulation model that was evolved
from the Miami Isopycnic-Coordinate Ocean Model
(MICOM) [Bleck et al., 1992]. HYCOM was developed
to address known shortcomings of the MICOM vertical
coordinate scheme. Vertical coordinates in HYCOM remain
isopycnic in the open and stratified ocean. However, they
are smoothly transition to z coordinate in the weakly
stratified upper ocean mixed layer, then to terrain-following
sigma coordinates in shallow water regions, and finally back
to level coordinates in very shallow water. The latter
transition prevents layers from becoming too thin where
the water is very shallow. HYCOM has been widely used to
study the North Atlantic circulation and Gulf Stream variability [Chassignet et al., 2003, and references therein].
[13] HYCOM is also equipped with biogeochemical
models such as Nutrient-Phytoplankton-Zooplankton
(NPZ) and Nutrient-Phytoplankton-Zooplankton-Detritus
(NPZD) models. However, since there are some uncertainties in setting up model’s initial condition of zooplankton
and biological parameters (such as the phytoplankton
growth/mortality rates, nutrient uptake rate, zooplankton
grazing rate, etc.) for this specific region in the Katrinainduced phytoplankton bloom event, we are not using these
biogeochemical models in our simulations. Instead, it makes
more sense to focus on the effect of physical processes on
nutrient distribution in response to the hurricane by treating
the three nutrient fields, i.e., nitrate, phosphate and silicate,
as passive tracers. The three types of nutrient concentrations
are initialized from World Ocean Atlas data set, and
subsequently influenced by physical processes (advection
and turbulent diffusion) only. There is no nutrient flow from
its dissolved phase to phytoplankton or zooplankton in our
simulations. The purpose of this passive tracer-based nutrient response study is to identify the dominant nutrient type
in the Katrina-induced phytoplankton event.
3.2. Model Simulations
[14] A simulation nested in the North Atlantic HYCOM
ocean prediction system [Chassignet et al., 2007] was
attempted before this study. However, the cyclonic eddy
did not occur exactly at (24°N, 84°W), and the effect of
C03023
upwelling on the vertical profile was not significant at that
location during the Katrina passage. This is because the
hurricane response is sensitive to the initial condition of the
LC and its frontal eddy. The position of the SST cooling in
the nested simulation did not match with the location of the
observed phytoplankton bloom and maximum SST drop.
Thus, we were motivated to do a two-case study to
investigate the importance of the LC frontal eddy in the
event of the Katrina-induced phytoplankton bloom.
[15] To evaluate the importance of the cold-core eddy on
physical and nutrient responses of the upper ocean to the
hurricane forcing, and identify the mechanism of the hurricane-driven bloom, we use the same north Atlantic
HYCOM model configuration used by Xie et al. [2007]
for Gulf Stream meanders and eddies in the South Atlantic
Bight to spin up the HYCOM model. Two cases with and
without cold-core front eddy are then identified. The ocean
conditions for these two cases are set up as the initial
condition, and then wind-forcing associated with Hurricane
Katrina is applied to these two cases. The model domain
covers the north and equatorial Atlantic (98°W – 14°E,
17°S – 52°N). The horizontal resolution is 0.18° and the
vertical resolution is 22 layers that move vertically as a
function of total depth. The model is spun up for 1 year with
an initial condition constructed from the Levitus Data Set
[Levitus, 1982] and forced by Comprehensive Ocean Atmosphere Data Set (COADS) monthly climatology atmospheric forcing including wind, air-temperature, relative
humidity, and short-wave radiation and long-wave radiations. The results are then used as the initial conditions of
the Hurricane response for simulations. Cyclone eddies are
found very often near Dry Tortugas area in the climatology
run, and they can be either generated locally or evolved
from LCFEs ending up there after moving around the LC
front. In their numerical study using MICOM statistically,
Romanou et al. [2004] reported that in the Gulf of Mexico
cyclonic eddies are most frequently found in the Dry
Tortugas area. In this study, we choose two cases of LC
states near Dry Tortugas on two different model days
(Figure 2): Experiment I, with the presence of a cyclone
eddy near (24°N, 84°W) on model Day 600 (Year 2,
30 August, Figure 2a); and Experiment II, without the
presence of a cyclone eddy near Dry Tortugas on model
Day 549 (Year 2, 9 July, Figure 2b). Since the differences
between July and August climatology monthly mean in the
vertical structure are small when compared with the differences between the two cases with and without the presence
of the eddy, it is assumed that the differences in the
simulations between the two cases are mainly caused by
the eddy. Here we focus on the response of upper ocean
nutrient distribution to Hurricane Katrina. The advection and
diffusion based multitracer code equipped in HYCOM are
used for tracer distribution. Since biological processes are
not considered in this present study, the nutrient tracers
are treated as passive tracers. The annual mean fields of
nitrate, phosphate, and silicate are extracted from World
Ocean Atlas 2005 and set as the initial conditions of nutrient
at the beginning of the both experiments.
[16] The two cases of hurricane simulations are carried
out for the time period of 25– 31 August 2005, and they are
driven by NCEP atmospheric forcing fields at the ocean
surface: wind, air temperature, and relative humidity. Wind
4 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
C03023
Figure 2. Initial conditions of SSH for (a) Experiment I: with the presence of a cyclonic eddy near
(24°N, 84°W); and (b) Experiment II: without the presence of the eddy (units: m).
stress is computed using the conventional bulk aerodynamic
formula:
t ¼ rCd U10 U10 ;
ð1Þ
where r is the density of air and U10 is the wind speed at
10 m height. The drag coefficient, Cd is computed using
Garratt’s [1977] composite form:
Cd ¼ ð0:73 þ 0:069U10 Þ 103 :
ð2Þ
[17] Since the total contribution of the surface heat flux
during the event accounts for only 10% of the mixed layer
heat budget, and with the cloud effect and warm air
temperature associated with the storm, the sum of the
short-wave and long-wave radiation is only a small portion
of the total air-sea heat exchange, which is dominated by the
latent heat flux due to the strong hurricane wind speed. Thus,
the short-wave and long-wave radiation heat flux at the
ocean surface are not included in the simulations. HYCOM
is equipped with several options of vertical mixing model
including Kraus-Turner model (KT), Priece-Weller-Pinkel
dynamical instability model (PWP), the K-Profile Parameterization model (KPP), the Mellor-Yamada level 2.5 model
(MY), and the NASA Goddard Institute for Space Studies
level 2 model (GISS). In this study, KPP model is used for
vertical mixing.
4. Results and Discussions
[18] In this section, results of the two-case studies with
numerical simulations for the hurricane response are compared and discussed. The physical response is presented in
section 4.1, nutrient response in section 4.2, with discussion
in section 4.3.
4.1. Physical Response
[19] Figure 3 shows the SST map near the LC before and
after Hurricane Katrina passage in Experiment I (Figures 3a
and 3b) and Experiment II (Figures 3c and 3d). In both
cases, the area is dominated by warm SSTs of 30°C on
27 August before Hurricane Katrina (Figures 3a and 3c).
However, after Hurricane Katrina, the SST cooling shows
quite large differences between the two cases: in Experiment I (Figure 3b), the maximum SST drop (4°C) is at
the location of the cold-core eddy, and the SST cooling
inside of the LC frontal eddy is much more significant
than other areas; in Experiment II (Figure 3d), SST cooling
shows no significant difference in the observed phytoplankton bloom location than other areas on the Hurricane
track. Figure 4 shows the time series of the SST drop at
the location of the observed phytoplankton bloom in
Experiments I and II. The SST drop derived from
AMSR-E satellite measurement is also plotted in Figure 4
for comparison. SST cooling from 27 August to 28 August is
4°C in Experiment I and 2.5°C in Experiment II, while the
SST dropped 6°C in the satellite measurement. At the
phytoplankton bloom location, the difference of SST cooling
between the two experiments is 1.5°C. Because the model
settings of Experiments I and II are identical except the
preexisting LC frontal eddy at the phytoplankton bloom
location in Experiment I, the presence of this eddy appears to
be responsible for this difference in the SST cooling at the
location of the phytoplankton bloom. On the right side of the
hurricane track, both Experiments I and II show a hurricaneinduced SST drop of 1 – 2°C. Although the maximum SST
cooling in Experiment I is 2°C less than the satellite
measurement, it is in better agreement with the satellite
measurement than in Experiment II. The difference between
the Experiment I and satellite measurement is mainly due to
the model setting of the initial condition. Since data assimilation is not used in the model experiments, the model’s
initial vertical thermal and dynamic structures are different
from the conditions of the real ocean before Hurricane
Katrina passage, leading to the model-measurement difference. In fact, both SST and SSH are variables of the sea
surface, and the method to accurately project the satellitemeasured sea surface property down to the water column
below the surface is an area of active research in data
assimilation. With the focus to investigate the mechanism
of observed phytoplankton bloom at the location (24°N,
84°W), the twin experiment design in this study helps to
5 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
C03023
Figure 3. Model-simulated SST in (a– b) Experiment I and (c– d) Experiment II (units: °C). Figures 3a
and 3c are on 27 August, and Figures 3b and 3d are on 29 August. Katrina track is overlaid on Figure 3a
as a dashed line with square symbols.
overcome this model limitation, and reveals that a preexisting cold core eddy makes more significant SST cooling at the
location of the phytoplankton bloom where maximum SST
drop also observed from the satellite.
[20] The time series of vertical temperature profile and
mixed layer depth near the location of the observed phytoplankton bloom (24°N, 84°W) in the two experiments are
shown in Figure 5. Before Hurricane Katrina, the surface
waters down to 30 m depth are dominated by warm
temperature > 29°C in both cases. However, the vertical
structure below the surface shows large difference between
the two cases due to different LC states. Without the
presence of the cyclone eddy in Experiment II (Figure 5b),
the location of the observed phytoplankton bloom is on the
path of the LC, and thus is dominated by warm LC waters. In
this case, the isotherms are deep, e.g., the 26°C isotherm is at
85 m depth. With the presence of the cyclone eddy in
Experiment I (Figure 5a), the isotherms are elevated, and the
26°C isotherm is now at 60 m, which is 25 m shallower
than in Experiment II. In Experiment I, strong hurricaneinduced upwelling lifts up the isotherms by 30 m on
27 August, and the mixed layer depth is also decreased.
Because of strong vertical diffusion, the mixed layer depth
deepens by 40 m on 28 August, and then it starts to
decrease gradually. The maximum mixed layer depth is
about 50 m, and coldest temperature below the mixed layer
depth is 25°C, which is originally located at 70 m depth
before Katrina. Thus, the deepest layer that contributes to the
surface cooling is at 70 m depth. The significant mixed
layer temperature cooling occurs between 27 and 28 August,
which is similar to satellite SST observations. In Experiment
II, because the location is under the influence of LC
advection, changes of mixed layer depth and vertical profiles
are different than from those in Experiment I. The upwelling
Figure 4. Time series of SST drop at (24°N, 84°W) in
Experiments I and II and the satellite-derived AMSR-E data
from 26 to 31 August 2005.
6 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
C03023
Figure 5. Time series of vertical temperature profile at (24°N, 84°W) in (a) Experiment I and
(b) Experiment II (units: °C). Mixed layer depth is indicated as a dashed line.
and vertical diffusion processes are not as strong as in
Experiment I, and the isotherms are deeper. The 27°C
isotherm is well below the bottom of the mixed layer during
and after the Katrina passage, and it never reaches the
surface.
[21] To further explain the mechanism of the mixed layer
cooling, we analyzed the roles of the major heat budget
terms that contribute to the cooling on the basis of the
thermal evolution equation, QT = QU+V QW + QDV +
QDH + QS, where QT is the rate of change of heat storage,
QU+V is horizontal advection, QW is vertical advection
(upwelling), QDV is vertical diffusion, QDH is horizontal
diffusion, and QS is surface heat flux. Figure 6 shows
vertical cross section of the rate of change in water
temperature due to (Figure 6a) vertical advection QW,
(Figure 6b) vertical diffusion QDV, and (Figure 6c) horizontal advection QU+V averaged from 27 August 0600 UTC to
28 August 0000 UTC along 84°W in Experiment I. Below
the mixed layer depth (50 m), the upwelling process
dominates the temperature change. The maximum rate of
change in temperature due to upwelling reaches 5°C/d just
below the mixed layer base. Within the mixed layer, the
vertical diffusive mixing process dominates the temperature
change with a cooling of 2– 4°C/d. Below the mixed layer,
the vertical mixing process causes a temperature warming of
2– 4°C/d. The horizontal advection also has a contribution
to the temperature cooling due to divergence and horizontal
temperature gradient. The maximum rate of change in
temperature due to horizontal advection reaches 0.2 –
0.6°C/d at the surface, which account for 5 –15% of the
Figure 6. Time rate of change in water temperature (°C/d) due to (a) vertical advection, (b) vertical
diffusion, and (c) horizontal advection averaged from 27 August 0600 UTC to 28 August 0000 UTC
along 84°W cross section in Experiment I.
7 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
total rate of temperature change. These results are in
agreement with the mixed layer heat budget analysis by
Prasad and Hogan [2007].
[22] In summary, SST cooling is more significant with the
presence of the cyclone eddy than in case of absence of the
eddy. The vertical thermal structure shows that the cyclone
eddy lifts up the isotherms prior to the arrival of the Katrina,
thus leads to lower water temperature than in case of
absence of the eddy at the same depth. It should be pointed
out that the mechanism of mixed layer cooling is the same
as Price [1981]: the entrainment (vertical diffusion) process
dominates mixed layer cooling, while upwelling induced by
the hurricane wind stress-curl is significant below the mixed
layer base. Our model results indicate that the preexisting
cold eddy makes a significant difference in the location and
magnitude of the hurricane-induced mixed layer cooling
between the two experiments.
4.2. Nutrient Responses
4.2.1. Overall Description
[23] The initial surface nitrate, phosphate, and silicate
concentrations are about 0.8 mM, 0.1 mM and 1.6 mM,
respectively, in both cases before Hurricane Katrina. The
basic mechanism of nutrient response is the same as
temperature: deep nutrient-rich waters are brought up to
the surface by hurricane-induced strong upwelling and
vertical diffusion processes during the Katrina passage.
Since they are treated as passive tracers in the simulations,
their response to the hurricane colocated well with temperature, and thus shows quite large differences between the
two cases. Figure 7 shows the post-Katrina/prior-Katrina
ratio of surface nitrate, phosphate, and silicate concentrations. In Experiment I, all of the three types of nutrient
concentrations within the cyclone eddy increase more
significantly than other areas on the hurricane track. The
maximum nutrient concentration increase occurs at the location of the observed phytoplankton bloom (24°N, 84°W).
The nitrate concentration increases most significantly by
100% at the location of observed phytoplankton bloom,
while phosphate and silicate concentration increases 40%
and 20%, respectively. In Experiment II, nitrate, phosphate,
and silicate concentrations at (24°N, 84°W) do not show
more significant increase than other areas on the hurricane
track. In general, the enhancement of all of the three nutrient
concentrations is less than 20%.
4.2.2. Nitrate Responses
[24] Figure 8 shows the time series of nitrate, phosphate,
and silicate vertical profiles at (24°N, 84°W). Before
Katrina passage, the subsurface nitrate concentrations in
Experiment I are higher than in Experiment II at the same
depth (Figures 8a and 8b). For example, at 70 m depth, the
nitrate concentration is 2.0 mM in Experiment I, while it is
only 1.2 mM in Experiment II. This is primarily due to
cyclone eddy induced upwelling, which elevates the nitrate
levels in Experiment I. The differences in the initial vertical
structure between the two cases consequently lead to
different results in the nitrate response to Katrina. In
Experiment I, waters at 70 m depth with nitrate concentration of 2.0 mM are lifted up by 30 m and further
entrained into the mixed layer, which significantly increases
the nitrate concentrations at the surface. The surface nitrate
concentration increases dramatically between 27 and 28
C03023
August, and the increase is in phase with SST cooling as
shown in Figure 5a. In Experiment II, however, the 2.0 mM
contour line is well below the bottom of the mixed layer,
and does not contribute to the surface nitrate enhancement.
The maximum nitrate concentration at the mixed layer
base is about 1.2 mM, and thus the change of the nitrate
concentration at the surface is less significant compared with
Experiment I.
4.2.3. Phosphate Responses
[25] The phosphate vertical profile (Figures 8c and 8d)
shows the similar pattern as in nitrate in the two experiments. In Experiment I, the surface phosphate concentration
increases from 0.1 mM to 0.14 mM after the Katrina
passage near the location of the observed phytoplankton
bloom. Before Katrina, the 0.14 mM contour line is originally at 70 m depth, and it is uplifted and further entrained
into the mixed layer during the Katrina passage. The 0.16 mM
contour line, which is originally at 80 m depth, never
reaches the surface. Therefore, the impact of the hurricaneinduced upwelling and vertical diffusion can reach 70 m
depth, which is consistent with the results of nitrate concentration. In Experiment II, the contour lines are much deeper
than in Experiment I. In this case, waters with 0.14 mM
phosphate concentration is originally at 90 m depth, and it is
well below the bottom of the mixed layer during the Katrina
passage. Before Katrina, the phosphate concentration at 70 m
depth is 0.12 mM. Although it is brought up to the surface
by the upwelling and vertical diffusion, it does not enhance
the surface phosphate concentration due to little vertical
gradient in the top 70 m.
4.2.4. Silicate Responses
[26] The surface silicate concentrations show much less
significant changes after Katrina passage compared with
nitrate and phosphate (Figures 8e and 8f). Before Katrina,
although the silicate concentration is higher in Experiment I
than at the same depth in Experiment II, it can be seen that
in both cases, the vertical gradient of silicate concentration
is minimal in the top 70 m depth. Consequently, the
upwelling and vertical diffusion process does not significantly increase the surface silicate levels in either of the
experiments.
4.3. Discussion
[27] It is well documented that the phytoplankton concentrations are increased over large areas because of intensified surface nutrient supply caused by strong wind-forcing
[Babin et al., 2004; Lin et al., 2003; Miller et al., 2006;
Walker et al., 2005]. However, the location of the phytoplankton bloom at (24°N, 84°W) cannot be explained only
by upwelling and vertical diffusion associated with Hurricane Katrina. In fact, it has been well-documented that the
maximum SST drop usually occurs to the right of the
hurricane track because of the asymmetry in turning direction of the wind stress vector that drives a very strong
asymmetry in the mixed layer velocity [Price, 1981]. On
the contrary, the observed Katrina-induced maximum SST
drop and phytoplankton bloom at (24°N, 84°W) occurred on
the left of the hurricane track. According to National
Hurricane Center report (http://www.nhc.noaa.gov/pdf/
TCR-AL122005_Katrina.pdf), the center of Hurricane
Katrina is at (24.5°N, 84°W) while moving toward the west
on 27 August at 0600 UTC. Thus, the location of the
8 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
Figure 7. Post-Katrina/prior-Katrina ratio of (a – b) nitrate, (c– d) phosphate, and (e – f) silicate.
Figures 7a, 7c, and 7e are from Experiment I, and Figures 7b, 7d, and 7f are from Experiment II. The
27 and 29 August data are used as the prior-Katrina and post-Katrina data, respectively.
9 of 12
C03023
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
Figure 8. Time series of vertical profile of (a –b) nitrate, (c –d) phosphate, and (e – f) silicate at (24°N,
84°W). Figures 8a, 8c, and 8e are from Experiment I, and Figures 8b, 8d, and 8f are from Experiment II
(units: mM). Mixed layer depth is indicated as a dashed line.
10 of 12
C03023
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
observed maximum SST drop and the center of the phytoplankton bloom is 20– 30 km left on the hurricane track. The
colocation of the cyclone eddy and the phytoplankton bloom
explains this contradiction to the classic rightward bias
response theories: the preexisting cold-core eddy at (24°N,
84°W) played an important role in the bloom event.
[28] The numerical simulations in this study have further
supported this hypothesis. The simulation results show that
the SST cooling is more significant in case of the presence
of the cyclone eddy at the location of the observed phytoplankton bloom. Although the response of the surface
nitrate, phosphate, and silicate concentrations is due to the
same mechanism of upwelling and vertical diffusion that
induces SST cooling, the enhancement of the three types of
nutrient concentrations at the surface are quite different. The
surface nitrate and phosphate concentrations increase by
100% and 40%, respectively, while the surface silicate
concentration increase only 20%. Given the same strength
of upwelling and vertical diffusion, the enhancement of the
surface nutrient levels depends on the initial vertical gradient of the subsurface nutrient concentrations within the
maximum depth that the upwelling and vertical diffusion
processes can reach. Generally, the vertical gradient of the
nitrate and phosphate concentrations within the top 70 m
depth are larger than that of silicate, and thus lead to much
more significant surface enhancement. In addition, the
preexisting cyclonic eddy greatly enhances the vertical
gradient of nitrate and phosphate in the top 70 m depth,
and consequently leads to more significant surface nitrate
and phosphate concentrations in Experiment I. However,
because of deep nutricline in the silicate vertical profile, the
eddy does little enhancement on its vertical gradient of the
top 70 m depth. Thus, there is little difference in the surface
silicate enhancement between the two simulated cases.
[29] The observed phytoplankton bloom is near the Tortugas eddy. Lee et al. [1992] reported that the eddy-induced
upwelling could uplift of the nutricline to within 50 m of
the surface near the eddy center. Nitrate and phosphate
concentrations were about 10 mM and 1.0 mM, respectively,
at 100 m in the eddy center. Without the eddy-induced
upwelling, nitrate levels would be only from 1 to 3 mM, and
phosphate levels is about 0.1 mM at the 100 m. In our
simulations, the cyclone eddy increases the nitrate concentration from 2 mM to 4 mM, and the phosphate concentration
from 0.15 mM to 0.3 mM at 100 m depth. The hurricaneinduced upwelling further increases the nitrate and phosphate concentrations to 8 mM and 0.4 mM, respectively.
[30] The focus of this study is different from the other
similar study of ocean responses of major Gulf of Mexico
hurricanes in 2005 [Gierach and Subrahmanyam, 2008],
which analyzed satellite-derived observations of SST, Chl-a,
SSH anomaly, and surface winds during the passage of
Hurricanes Katrina, Rita, and Wilma. Their satellite analysis
of the upper ocean responses to the three hurricanes
provided a preliminary understanding of the biophysical
effects that transpired in the Gulf of Mexico. In addition,
various factors that contribute to the SST cooling and Chl-a
enhancement were analyzed. However, these observations
only provided insight near the ocean surface. In this study,
we focus on using the numerical model to study the overall
mechanics of the physical and nutrient responses to Hurricane Katrina and the subsurface dynamics are also analyzed.
C03023
Our particular interest is in the investigation of the mechanism of Katrina-induced phytoplankton bloom event near
(24°N, 84°W).
[31] Color dissolved organic matter (CDOM) is also a big
component that contributes to the ocean surface radiance
spectrum. With its maxima depth at subsurface layer, there
are some reports showing the elevated CDOM concentration driven by wind-forcing in North Atlantic region [Hoge
and Lyon, 2002]. Since our knowledge of CDOM such as
vertical profile, global distribution and its relation with
biological and geochemical processes in the ocean are still
very limited, it is difficult to simulate and quantify its role in
ocean responses after the hurricane passage. The radiance
spectrum from Katrina-induced bloom shows strong absorption at the wavelength 443 nm, suggesting that it was a
phytoplankton (not CDOM) bloom [Shi and Wang, 2007].
As also shown by Shi and Wang [2007], a time lag between
the chlorophyll-a concentration and the SST drop in the
satellite remote sensing further strengthens this conclusion.
5. Summary and Conclusions
[32] Satellite observations revealed that Hurricane Katrina
led to a significant SST cooling and a phytoplankton bloom
near (24°N, 84°W) in the Gulf of Mexico. The SSH
anomaly derived from satellite altimetry also showed that
a southward propagating Loop Current frontal eddy colocated with the phytoplankton bloom, which suggested that
the eddy was an important factor in the phytoplankton
bloom event. In this study, HYCOM is used to study the
process of upper ocean responses to Hurricane Katrina. Two
different cases with and without the presence of a cyclone
eddy at the location of the observed phytoplankton bloom
are simulated. The model run demonstrates that the Katrinainduced phytoplankton bloom is attributed to the preexisting eddy, as well as to slow moving hurricane. The cyclonic
eddy plays a critical role in the development of the Katrinainduced bloom, and it significantly strengthens the upper
ocean dynamics and nutrient responses. In the simulations,
Katrina-induced upwelling and vertical diffusion process
can reach 70 m depth. The cyclonic eddy in the phytoplankton bloom area uplift the isotherms and increase
vertical temperature and nitrate and phosphate gradient in
the upper 70 m depth, and thus leads to more significant
SST cooling and increase in nitrate and phosphate concentrations than in the noneddy case. Our model results also
suggest that the nitrate concentration plays a dominant role
in the development of the phytoplankton bloom with over
100% increase in the surface concentration, while the
phosphate increases 40%. The silicate has a minimal effect
on the bloom with the smallest increase for its surface
concentration.
[33] Acknowledgments. This research was supported by NASA
(Ocean Biology and Biogeochemistry Program) and NOAA funding and
grants. MODIS Aqua SST and chlorophyll-a concentration data were
acquired from the NASA/GSFC ocean color data archive. SSH data were
provided and remapped from Colorado Center for Astrodynamics Research
(CCAR). AMSR-E data are produced by Remote Sensing Systems and
sponsored by the NASA Earth Science REASON DISCOVER project and
the AMSR-E science team. We thank three anonymous reviewers for their
useful comments. The views, opinions, and findings contained in this paper
are those of the authors and should not be construed as an official NOAA or
U.S. government position, policy, or decision.
11 of 12
C03023
LIU ET AL.: A KATRINA-INDUCED PHYTOPLANKTON BLOOM
References
Babin, S. M., J. A. Carton, T. D. Dickey, and J. D. Wiggert (2004), Satellite
evidence of hurricane-induced phytoplankton blooms in an oceanic desert, J. Geophys. Res., 109, C03043, doi:10.1029/2003JC001938.
Bleck, R. (2002), An oceanic general circulation model framed in hybrid
isopycnic-Cartesian coordinate, Ocean Modell., 4, 55 – 88, doi:10.1016/
S1463-5003(01)00012-9.
Bleck, R., C. Rooth, D. Hu, and L. T. Smith (1992), Salinity-driven thermocline transients in a wind and thermohaline-forced isopycnic coordinate model of the North Atlantic, J. Phys. Oceanogr., 22, 1486 – 1505,
doi:10.1175/1520-0485(1992)022<1486:SDTTIA>2.0.CO;2.
Brooks, D. A. (1983), The wake of Hurricane Allen in the western Gulf
of Mexico, J. Phys. Oceanogr., 13, 117 – 129, doi:10.1175/15200485(1983)013<0117:TWOHAI>2.0.CO;2.
Chassignet, E., L. Smith, G. Halliwell, and R. Bleck (2003), North Atlantic
simulations with the Hybrid Coordinate Ocean Model (HYCOM): Impact
of coordinate choice, reference pressure, and thermobaricity, J. Phys.
Oceanogr., 33, 2504 – 2526, doi:10.1175/1520-0485(2003)033<2504:
NASWTH>2.0.CO;2.
Chassignet, E. P., H. E. Hurlburt, O. M. Smedstad, G. R. Halliwell, P. J.
Hogan, A. J. Wallcraft, R. Baraille, and R. Bleck (2007), The HYCOM
(Hybrid Coordinate Ocean Model) data assimilative system, J. Mar. Syst.,
65, 60 – 83, doi:10.1016/j.jmarsys.2005.09.016.
Dickey, T., et al. (1998), Upper-ocean temperature response to Hurricane
Felix as measured by the Bermuda test-bed mooring, Mon. Weather Rev.,
126, 1195 – 1200, doi:10.1175/1520-0493(1998)126<1195:UOTRTH>
2.0.CO;2.
Fratantoni, P. S., T. N. Lee, G. P. Podesta, and F. Muller-Karger (1998), The
influence of Loop Current perturbations on the formation and evolution
of Tortugas eddies in the southern Straits of Florida, J. Geophys. Res.,
103(C11), 24,759 – 24,799, doi:10.1029/98JC02147.
Garratt, J. R. (1977), Review of drag coefficients over oceans and continents, Mon. Weather Rev., 105, 915 – 929, doi:10.1175/1520-0493(1977)
105<0915:RODCOO>2.0.CO;2.
Gierach, M. M., and B. Subrahmanyam (2008), Biophysical responses of
the upper ocean to major Gulf of Mexico hurricanes in 2005, J. Geophys.
Res., 113, C04029, doi:10.1029/2007JC004419.
Gordon, H. R., and M. Wang (1994), Retrieval of water-leaving radiance and
aerosol optical thickness over the oceans with SeaWiFS: A preliminary
algorithm, Appl. Opt., 33, 443 – 452.
Hoge, F. E., and P. E. Lyon (2002), Satellite observation of Chromophoric
Dissovled Organic Matter (CDOM) variability in the wake of hurricanes
and typhoons, Geophys. Res. Lett., 29(19), 1908, doi:10.1029/
2002GL015114.
Jacob, S. D., L. K. Shay, and A. J. Mariano (2000), The 3D oceanic mixed
layer response to Hurricane Gilbert, J. Phys. Oceanogr., 30, 1407 – 1429,
doi:10.1175/1520-0485(2000)030<1407:TOMLRT>2.0.CO;2.
Lee, T. N., C. Rooth, E. Williams, M. McGowan, A. F. Szmant, and M. E.
Clarke (1992), Influence of Florida Current, gyres and wind-driven circulation on transport of larvae and recruitment in the Florida Keys coral
reefs, Cont. Shelf Res., 12, 971 – 1002, doi:10.1016/0278-4343(92)
90055-O.
Lee, T. N., K. Leaman, E. Williams, T. Berger, and L. Atkinson (1995),
Florida Current meanders and gyre formation in the southern Straits of
Florida, J. Geophys. Res., 100(C5), 8607 – 8620, doi:10.1029/
94JC02795.
Leipper, D. F. (1967), Observed ocean conditions and Hurricane Hilda,
1964, J. Atmos. Sci., 24, 182 – 196, doi:10.1175/1520-0469(1967)
024<0182:OOCAHH>2.0.CO;2.
Levitus, S. (1982), Climatological Atlas of the World Ocean, NOAA Prof.
Pap. 13, 173 pp., U. S. Govt. Print. Off, Washington, D.C.
Lin, I., W. T. Liu, C-C. Wu, G. T. F. Wong, C. Hu, Z. Chen, W-D. Liang,
Y. Yang, and K-K. Liu (2003), New evidence for enhanced ocean
primary production triggered by tropical cyclone, Geophys. Res. Lett.,
30(13), 1718, doi:10.1029/2003GL017141.
C03023
Miller, W. D., L. W. H. Jr., and J. E. Adolf (2006), Hurricane Isabel
generated an unusual fall bloom in Chesapeake Bay, Geophys. Res. Lett.,
33, L06612, doi:10.1029/2005GL025658.
Paluszkiewicz, T., L. P. Atkinson, E. S. Posmentier, and C. R. McClain
(1983), Observations of a loop current frontal eddy intrusion onto the
West Florida Shelf, J. Geophys. Res., 88, 9639 – 9652, doi:10.1029/
JC088iC14p09639.
Prasad, T. G., and P. J. Hogan (2007), Upper-ocean response to Hurricane
Ivan in a 1/25° nested Gulf of Mexico HYCOM, J. Geophys. Res., 112,
C04013, doi:10.1029/2006JC003695.
Price, J. F. (1981), Upper ocean response to a hurricane, J. Phys. Oceanogr.,
11, 153 – 175, doi:10.1175/1520-0485(1981)011<0153:UORTAH>2.0.
CO;2.
Price, J. F., T. B. Sanford, and G. Z. Forristall (1994), Forced stage response
to a moving hurricane, J. Phys. Oceanogr., 24, 233 – 260, doi:10.1175/
1520-0485(1994)024<0233:FSRTAM>2.0.CO;2.
Romanou, A., E. P. Chassignet, and W. Sturges (2004), Gulf of Mexico
circulation within a high-resolution numerical simulation of the North
Atlantic Ocean, J. Geophys. Res., 109, C01003, doi:10.1029/
2003JC001770.
Sanford, T. B., P. G. Black, J. R. Haustein, J. W. Feeney, G. Z. Forristall,
and J. F. Price (1987), Ocean response to a hurricane. part I: Observations, J. Phys. Oceanogr., 17, 2065 – 2083, doi:10.1175/1520-0485(1987)
017<2065:ORTAHP>2.0.CO;2.
Shay, L. K., and R. L. Elsberry (1987), Near-inertial ocean current response
to Hurricane Frederic, J. Phys. Oceanogr., 17, 1249 – 1269, doi:10.1175/
1520-0485(1987)017<1249:NIOCRT>2.0.CO;2.
Shay, L. K., R. L. Elsberry, and P. G. Black (1989), Vertical structure of the
ocean current response to a hurricane, J. Phys. Oceanogr., 19, 649 – 669,
doi:10.1175/1520-0485(1989)019<0649:VSOTOC>2.0.CO;2.
Shay, L. K., A. J. Mariano, S. D. Jacob, and E. H. Ryan (1998), Mean and
near-inertial ocean current response to Hurricane Gilbert, J. Phys. Oceanogr., 28, 858 – 889, doi:10.1175/1520-0485(1998)028<0858:MANIOC>
2.0.CO;2.
Shi, W., and M. Wang (2007), Observations of a Hurricane Katrina –
induced phytoplankton bloom in the Gulf of Mexico, Geophys. Res.
Lett., 34, L11607, doi:10.1029/2007GL029724.
Vukovich, F. M. (1988), On the formation of elongated cold perturbations
off the Dry Tortugas, J. Phys. Oceanogr., 18, 1051 – 1059, doi:10.1175/
1520-0485(1988)018<1051:OTFOEC>2.0.CO;2.
Walker, N. D., R. R. Leben, and S. Balasubramanian (2005), Hurricaneforced upwelling and chlorophyll a enhancement within cold-core cyclones in the Gulf of Mexico, Geophys. Res. Lett., 32, L18610,
doi:10.1029/2005GL023716.
Wang, M. (2007), Remote sensing of the ocean contributions from ultraviolet to near-infrared using the shortwave infrared bands: Simulations,
Appl. Opt., 46, 1535 – 1547, doi:10.1364/AO.46.001535.
Wang, M., and W. Shi (2005), Estimation of ocean contribution at the
MODIS near-infrared wavelengths along the east coast of the U.S.:
Two case studies, Geophys. Res. Lett., 32, L13606, doi:10.1029/
2005GL022917.
Wang, M., J. Tang, and W. Shi (2007), MODIS-derived ocean color products along the China east coastal region, Geophys. Res. Lett., 34,
L06611, doi:10.1029/2006GL028599.
Wang, M., S. Son, and W. Shi (2009), Evaluation of MODIS SWIR and
NIR-SWIR atmospheric correction algorithms using SeaBASS data,
Remote Sens. Environ., 113, 635 – 644, doi:10.1016/j.rse.2008.11.005.
Xie, L., X. Liu, and L. J. Pietrafesa (2007), Effect of bathymetric curvature
on Gulf Stream instability in the vicinity of the Charleston Bump, J. Phys.
Oceanogr., 37, 452 – 475, doi:10.1175/JPO2995.1.
X. Liu, W. Shi, and M. Wang, Center for Satellite Applications and
Research, National Environmental Satellite, Data, and Information Service,
NOAA, E/RA3, Room 102, 5200 Auth Road, Camp Springs, MD 20746,
USA. ([email protected])
12 of 12