2012 Sevadjian et al - Oregon State University

Continental Shelf Research 50–51 (2012) 1–15
Contents lists available at SciVerse ScienceDirect
Continental Shelf Research
journal homepage: www.elsevier.com/locate/csr
Research papers
Shoreward advection of phytoplankton and vertical re-distribution
of zooplankton by episodic near-bottom water pulses on an insular
shelf: Oahu, Hawaii
J.C. Sevadjian a,n, M.A. McManus a, K.J. Benoit-Bird b, K.E. Selph a
a
b
Department of Oceanography, University of Hawaii at Manoa, 1000 Pope Road, Honolulu, Hawaii 96822, USA
College of Earth, Ocean, and Atmospheric Sciences, Oregon State University, 104 CEOAS Administration Building, Corvallis, Oregon 97331, USA
a r t i c l e i n f o
abstract
Article history:
Received 18 January 2012
Received in revised form
7 July 2012
Accepted 13 September 2012
Available online 19 September 2012
Shoreward pulses of cold, high salinity, phytoplankton-rich bottom water represent short time scale
changes in nearshore hydrography and biological community structure off the leeward coast of Oahu,
Hawaii. A three-week mooring deployment in Spring 2009 revealed that ‘cold pulses’ occurred over all
phases of the semidiurnal surface tide, but that a statistically greater proportion occurred at low slack
tide, when the typically along-shore flow rotated and briefly exhibited a more dominant onshore
component. Cold pulses were more frequent and propagated farther shoreward when background
water-column stratification increased. Targeted shipboard sampling of a cold pulse in Spring 2010
revealed that chlorophyll fluorescence within the cold pulse was 7 standard deviations higher than the
11-h mean outside the cold pulse, phytoplankton concentrations (cells mL 1) were up to a factor of
3 higher within the cold pulse, and phytoplankton entrained within the cold pulse were adapted to
habitats with lower light levels and higher nutrient concentrations compared to ambient waters.
Analysis of multi-frequency acoustic data collected during two shipboard surveys in 2009 and 2010
indicated that acoustic scattering during cold pulses was predominantly biological, dominated by 1.0–
1.5 mm spheroid fluid-like scatterers, both in the waters above the cold pulse and in a strong-scattering
feature at the cold–pulse interface. These aggregations of larger organisms at the cold-pulse interface
did not appear to migrate downwards into the phytoplankton-rich water during the active passage of
the cold pulse. Observations of similar temperature events throughout the tropical Pacific, combined
with our multidisciplinary findings, suggest that pulsed deliveries of phytoplankton-rich water to
nearshore habitats may be regular occurrences throughout the North Pacific Subtropical Gyre.
& 2012 Elsevier Ltd. All rights reserved.
Keywords:
Cold pulse
Internal wave
Zooplankton
Phytoplankton
1. Introduction
Intermittent near-bottom pulses of colder, deeper water to the
nearshore are common phenomena at a variety of coastal regions
(Pineda, 1995). Cold pulses are characterized by rapid changes in
temperature and increases in velocity, in some cases up to 5 1C and
30 cm s 1 over 1–20 min, and likely represent the elevation of
subthermocline water by internal waves (Leichter et al., 1996;
Pineda, 1991; Storlazzi et al., 2003). The most commonly studied
mechanism for cold pulses of this type is a tidally generated internal
wave. While internal waves in the deep ocean can locally impact the
shallow water ecosystem by temporarily introducing high-nutrient
water from below the euphotic zone, deep-ocean internal waves are
not linked with significant horizontal transport. Near the shoreline,
however, nonlinear interactions with a shoaling sea floor can cause
internal waves to become unstable and break, resulting in high near-
n
Corresponding author. Tel.: þ1 8083479949.
E-mail address: [email protected] (J.C. Sevadjian).
0278-4343/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.csr.2012.09.006
bottom velocities and turbulent motions, ultimately providing a
means of advection of nutrients, larvae, phytoplankton and zooplankton to the nearshore (Leichter et al., 1996, 1998; Pineda, 1991).
Residence times of pulsed water on shallow (o30 m) environments
can vary from 1 h to several days (Leichter et al., 1996; Pineda,
1995), thus there is a finite, but substantially larger amount of time
in which nutrients and plankton can be taken up by organisms that
reside in the shallow environments. The introduction of these
waters to the nearshore environment has several potential biological
implications: increased food availability may result in aggregations
of organisms of higher trophic levels (Owen, 1981; Woodson and
McManus, 2007); increased nutrient levels can represent a significant fraction of the overall nutrient flux onto coral reefs,
allowing nutrient limited communities to prosper (Leichter et al.,
1996); and internal motions of this type may represent one of few
processes that result in shoreward transport of larvae (Pineda,
1991).
Although their characteristics vary greatly, internal waves
exist in some form throughout most of the world’s coastal regions
(Jackson, 2004). Rapid changes in near-bottom temperature
2
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
caused by shoreward-propagating internal waves have been
studied at continental shelves (Cairns, 1967; Pineda, 1995), shelf
breaks (Holloway, 1987; Small et al., 1999), coral reefs (Leichter
et al., 1996), in bays (Haury et al., 1979; Scotti and Pineda, 2004;
Storlazzi et al., 2003), and in straits (Morozov et al., 2002).
Shoreward pulses of cold bottom water also have been observed
on inner shelves around the Pacific Islands (Gove et al., 2006;
Storlazzi and Jaffe, 2008 PIFSC Cruise Report, 2011), and on all
coasts of the island of Oahu, Hawaii (McManus et al., 2008;
Merrifield et al., unpubl. data; Sevadjian et al., 2010), although
these coastal Hawaiian observations have been anecdotal and
have not been examined in detail prior to this study. Because
internal-wave pulses are widespread phenomena near coasts, it
follows that episodic cross-shore exchange of organisms and
nutrients by internal wave pulses may also be widespread. Even
small temperature decreases can be associated with significant
increases in plankton concentrations (Leichter et al., 1998). In
spite of this apparent ubiquity in occurrence, only a handful of
locations have been studied with the tools necessary to resolve
physical, biological, and chemical characteristics of cold pulses, with
notable exceptions including locations off Southern California
(Pineda, 1991, 1999), in the Florida Keys (Leichter et al., 1996,
1998), and in Massachusetts Bay (Haury et al., 1979, 1983).
Our study was located off the leeward coast of Oahu, the third
largest island of the Hawaiian Archipelago, which is located near
the center of the North Pacific Subtropical Gyre (NPSG). The
surface ocean ecosystem of the NPSG is oligotrophic, persistently
low in nutrients and biomass and supportive of a low production
rate of organic matter (Karl and Lukas, 1996). The pycnocline
within the NPSG is relatively deep year round (Karl and Lukas,
1996), and the shallow ( o50 m) coastal waters surrounding the
Hawaiian Islands are often very weakly thermally stratified
(Hamilton et al., 1995). On the inner shelf of Oahu, periods of
vertical stratification in density and plankton distributions often
persist only a few minutes to hours at a time (McManus et al.,
2012; Sevadjian et al., 2010). Within the context of this oligotrophic and relatively homogeneous environment, episodic intrusions of cold, high salinity, phytoplankton-rich water to the inner
shelf may be particularly important in providing nutrients and
food necessary to sustain shallow water ecosystems.
The arrival of internal waves near the coast is sometimes
associated with specific phases or amplitudes of the semidiurnal
surface tide (Leichter et al., 1996; Ramp et al., 2004; Storlazzi
et al., 2003), although the extent of this association is likely in
part a function of the degree of exposure of a given site to internal
tidal energy (Shroyer et al., 2011). Wave arrival time, frequency of
occurrence, and amplitude can also be modulated by longer timescale forcing such as ocean–atmosphere interactions. Pineda and
Lopez (2002) found that a shallowing of the thermocline caused
by a change in low-frequency wind forcing resulted in an increase
in the number of warm surface bores (e.g., fronts that form when
cold, deeper water from internal bores recedes offshore and
warmer water returns at the surface) in Southern California.
Because a given disturbance will be more energetic in a more
strongly stratified water column (Pineda and Lopez, 2002), a
common observation is that both the number of internal wave
events and the extent of shoreward propagation increase during
the spring and summer months, when background stratification is
higher (Cairns and Nelson, 1970; Pineda, 1991; Storlazzi et al.,
2003). A better understanding of the timing and magnitudes of
cold pulses may lead to insight into how organisms may actively
respond to exploit the temporary influxes of nutrients or food.
We took a multidisciplinary approach to describe the physical
and biological characteristics of episodic cold pulses off leeward
Oahu, Hawaii. A synthesis of physical, biological, optical, chemical
and acoustical data was undertaken in an effort to formulate
a more complete picture of the dynamics associated with these
events.
2. Methods
2.1. Study area and sampling overview
The study was part of a four-year research effort that investigated the relationships between the distribution of phytoplankton, zooplankton, mesopelagic micronekton, and spinner dolphins
off the leeward coast of Oahu, with respect to local and regional
physical circulation patterns. McManus et al. (2008) reported the
first observations of cold bottom water pulses occurring off
the west shore of Oahu, in a study based on the first two years
of the project (2005 and 2006). In the 2008 study, investigation
into the dynamics of cold pulses was cursory. We returned to the
field site in 2009 and 2010 to substantially build upon these early
observations.
The 2009 and 2010 studies were located at 1581W, about 1 km
offshore from Makua Beach (Fig. 1). This section of coastline runs
north to south with little variation, and isobaths are roughly
parallel to the shoreline. The bottom slope is gradual (3.91) from
the coastline to the 110 m isobath, a linear distance of 1.6 km.
Between the 110 m and 2000 m isobath, the slope averages 9.21.
Diver surveys conducted during the study revealed that the sea
floor near the 10 m isobath is composed primarily of living coral
features on the order of meters in lateral size, above a bed of fine
sand. At progressively greater depths, larger coral structures give
way to smaller patchy fragments of coral rubble, and finally to a
sandy expanse at depths greater than 40 m.
Over three-week periods in the Springs of 2009 and 2010, we
deployed moored sensors to measure physical and acoustical
properties on the inner shelf in this region. Additionally, during
each deployment period we conducted a shipboard survey to
measure changes in the physical, biological, optical, chemical, and
acoustical attributes of the water column associated with individual cold-pulse events.
2.2. Moorings
Between 20 April and 12 May 2009, four moorings were
deployed at the 46, 25, 20, and 10-m isobaths. Between 11 and
30 April 2010, three moorings were deployed at the 25, 20, and
10-m isobaths (Fig. 1b, Table 1). Prior studies had shown significant nightly cross-shore migrations of micronekton (specifically, myctophid fish) at this site and noted the importance of
physical and biological gradients along this axis (Benoit-Bird
et al., 2008; McManus et al., 2008). Thus, the mooring locations
were oriented to resolve processes occurring in the cross-shore
direction. The mooring array consisted of thermistors, a dual
conductivity–temperature sensor, a near-bottom current meter,
an acoustic Doppler current profiler (ADCP), and a 200 kHz
echosounder. Because the additional mooring at 46 m was
deployed only during the 2009 study, the time-series analysis
was focused on the 2009 deployment period.
Prior to deployment, all thermistors were factory-calibrated; the
internal compasses of the current meter and ADCP were calibrated,
including accounting for local magnetic declination; and the
200 kHz echosounder was calibrated using the standard sphere
method (Foote et al., 1987). The 200 kHz echosounder had a 101
beam and used a 256 ms pulse. The conductivity cell of the Sea-Bird
Electronics (SBE) model 37 dual conductivity–temperature sensor
was not pumped, which likely resulted in errors in the salinity
calculation due to cell thermal mass during periods of rapid
temperature change. While the typical amount of time required
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
3
21.54°N
22°N
Makua
Range
KAUAI
21.52°N
Kaena Pt
OAHU
20
10
200
100
2000
Kaena Ridge 1000
AB C D
Makua
Beach
Makapu’u Pt
Mamala Bay
MOLOKAI
21.5°N
50 km
21°N
159°W
158°W
157°W
1 km
21.48°N
158.24°W
158.22°W
Fig. 1. (a) Study area (boxed region) off the leeward coast of the island of Oahu (depth contours in 1000 m intervals; circles indicate locations of HOT deep-water CTD
casts); and (b) mooring locations (depth contours in 10 m intervals; dark contours are 100 m, 200 m, 300 m, and 400 m isobaths).
Table 1
Instrumentation deployed during the 2009 and 2010 study periods.
Site name
Water depth (m)
Instrumentation
Manufacturer/specifications
Measurement depth (m)
Sample interval
Resolution
Moorings
Aa
46
25
11.5, 22.5, 33.5, 44.5
44.7
0–44.5
2–22
15 s
2 min
4s
2 min
C-T sensora
Sea-bird electronics (SBE) 39
Nortek
ASL water column profiler
RD instruments (RDI)
workhorse
SBE-37
0.002 1C
1.5 cm s 1
B
Thermistors
Current meter
200 kHz echosounder
300 kHz ADCP
24.5
1 min
Thermistors
Thermistors
SBE-39
SBE-39
5.5, 12, 18.5
5.5, 7, 8.5
15 s
15 s
0.002 1C
0.003 S m 1
0.002 1C
0.002 1C
Profiler package (C, T, p, O2,
chl, ac-9b, TAPS)
600 kHz ADCP
Echosounder package
Various; see Section 2.4 in
text
RDI workhorse
Simrad EK60 at 38, 70, 120,
200, and 710 kHz
5L
0.5 m opening/closing ring;
150 mm mesh
1–24
6–8 Hz
2–23
1–23
2 min
8–10 Hz
20, 14, 20.4
0–24, acoustic target
10–37 min
15 min
C
D
20
10
Shipboard surveys
B
25
Bottle samplesb
Zooplankton net towsa
a
b
2.8 cm s 1
3.4 cm s 1
2009 only.
2010 only.
for the conductivity cell of an un-pumped SBE-37 to equilibrate to a
change in temperature surrounding it is 100 s (C. Janzen, pers.
comm.), we found that during cold-pulse events the median
elapsed time before salinity returned to pre-event levels was
10 min, indicating that although cell thermal mass may have
contributed to the initial spikes, the persistent high salinity values
found several min after the onset were indicative of the true salinity
of the pulsed water. For further analysis of the moored salinity data,
we used a sliding median filter to eliminate spiking potentially
caused by cell thermal mass that removed all salinity values
0.02 psu greater than the 15 min median.
In addition to the deployed moorings, hourly wind and air
temperature data were obtained from a weather station located
on the coastline of Makua Range, 2.5 km to the north–northeast of
our mooring array (Horel et al., 2002). Data from monthly CTD
casts collected as part of the Hawaii Ocean Time Series (HOT)
program (http://hahana.soest.hawaii.edu/hot/hot-dogs/interface.
html) were utilized to locate the depths of the thermocline and
chlorophyll maximum. Casts were taken by the HOT program at
stations to the north–northwest and south of the study site in
water depths of 1500 m and 2500 m, respectively (Fig. 1a).
2.3. Identification of cold-pulse events
‘Cold-pulse’ events on the west shore of Oahu were first observed
by McManus et al. (2008). Visual inspection of the time series of
4
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
where for the ith measurement, n¼1, 2, 3,y, m.
Temperature data from the deepest thermistor at each mooring were tested for these criteria. Beginning with the first
measurement in the time series (i¼1), the temperature gradient
DT/Dt was calculated using Eq. (1) with n¼1. If (1) was satisfied
over these two data points, n was increased by 1, and (1) was
tested again. This iterative process continued until (1) failed
(n m). If for any 1rn rm, both (1) and (2) were satisfied, the
data indices i, iþ1,y, iþm were considered representative of a
cold pulse event. This entire process was then repeated beginning
with the next measurement (i¼mþ1) in the time series.
Following the identification of cold-pulse events, the following
characteristics were tabulated for each event: (1) arrival time; (2)
magnitude of temperature decrease; (3) time elapsed before
reaching minimum temperature; and (4) time elapsed before
temperature increased to within 0.5 times the initial decrease
(referred to here as ‘recovery time’). All physical and acoustical
data were then analyzed using these demarcated time indices to
compare event conditions with non-event conditions.
described in Holliday et al. (2009). Separate intakes were installed
for (1) the conductivity, temperature, oxygen, and WETStar fluorometer sensors; and (2) the ac-9 sensor. To avoid contamination
of water parcels, both plumbing intakes were installed 5 cm below
the bottom-most section of the profile cage, and only data collected
on downcasts were considered. All instruments sampled at rates of
6 Hz or 8 Hz, resulting in a vertical resolution of approximately
2 cm. In-situ data collected by the profiler were relayed to the
vessel in real-time, and the profile characteristics were used to
target plankton sampling with bottles and nets.
A vessel-mounted, downward-looking 600 kHz ADCP pinged
approximately once every 2 s throughout the shipboard survey.
Water-column velocity measurements were obtained by subtracting
ship motion recorded by the ADCP bottom-tracking feature. A
pressure sensor on the ADCP recorded transducer depth (which
was modulated by vessel motion), and 2-min velocity averages were
calculated in post-processing in fixed 0.5 m depth bins between 1 m
depth and 2 m.a.b. The resolution in velocity measurements under
this configuration is 3.4 cm s 1. A five-frequency, downwardlooking echosounder system (Simrad EK60 at 38 kHz, 70 kHz,
120 kHz, 200 kHz, and 710 kHz) was used during the surveys to
provide nearly continuous information on a wide size-range of
meso-zooplankton and fish. Transducers were mounted 1 m below
the surface on a rigid mount on the vessel’s starboard side. The
centers of each echosounder transducer were no more than 35 cm
apart to maximize spatial comparability of the data. All echosounders used a 256 ms outgoing pulse, with a pulse rate of approximately 8–10 Hz. The 38 kHz echosounder had a 121 conical splitbeam; the 70 kHz, 120 kHz, and 200 kHz echosounders each had a
71 conical split-beam; and the 710 kHz echosounder had a 51 single
beam. Interference between the acoustic sensors was prevented by
using frequencies that were not harmonics of one another and by
synchronizing outgoing pulses (Korneliussen et al., 2004). The
echosounder system was calibrated just prior to sampling using
the settings and physical set up used for data collection, using the
standard sphere method as prescribed by Foote et al. (1987).
2.4. Shipboard surveys
2.5. Targeted sampling
Several shipboard surveys were conducted during each of the
2009 and 2010 mooring deployment periods at the 25-m mooring
site. We present results from two of these surveys, in which a cold
pulse was detected while on station. During these surveys,
repeated profiles were taken from 1 m depth to 1 m above the
bottom (m.a.b.) over the side of an anchored vessel (the 10 m F/V
Alyce C.). This sampling method resulted in one profile approximately every 4 min, over a 4-h period (a total of 46 profiles) on
6 May 2009 and an 11-h period (142 profiles) on 18 April 2010.
The profiler cage was a 0.5 0.5 1.2 m steel frame fitted with
small floats in order to achieve a slight negative buoyancy, which
resulted in an average descent rate of 15 cm s 1. Cables running
between the profiler and the vessel were kept slack, allowing the
profiler to drop smoothly, decoupled from ship motion. A SBE-4
on the profiler measured conductivity, a SBE-3 sensor measured
temperature, a SBE-29 measured pressure, a SBE-43 measured
dissolved oxygen concentration, a WETLabs WETStar fluorometer
measured chlorophyll-a fluorescence, a WETLabs ac-9 measured
optical absorption and attenuation at nine wavelengths, and a
Tracor Acoustic Profiling System (TAPS) measured acoustic backscatter at six frequencies between 265 kHz and 3000 kHz (Table 1).
In addition to fluorescence measured by the WETStar, which
provides an estimate of chlorophyll-a concentration, an independent estimate of chlorophyll-a concentration was obtained using
absorption measurements from the ac-9, following the baselinesubtraction method of Davis et al. (1997) and a chlorophyll in-vivo
specific absorption of 0.014 m2 mg 1 (Bricaud et al., 1995), as
Zooplankton were sampled twice during the 2009 deployment
period using an opening–closing, 0.5 m diameter ring net with a
150 mm mesh and a remote pressure sensor. Each time, the net
was used to obtain a vertically integrated tow as well as one
targeted on an observed acoustic scattering layer. The net was
pulled upward at a rate of approximately 30 m min 1. Samples
were preserved with 5% formalin in buffered seawater. After
returning to the lab, individual zooplankton in the samples were
identified and sized following Benoit-Bird et al. (2008).
When a cold pulse was detected in the profiler’s temperature,
salinity, and fluorescence data during the 2010 survey, a total of
three bottle samples were taken—one from within and two from
above the pulsed water mass. ‘Bottle’ samples were taken using a
5 L Aquatic Research Instruments Horizontal Discrete Point Water
Sampler fixed with a Simrad pressure sensor (PI 60). Phytoplankton population concentrations within the bottle samples were
estimated using flow cytometry. Flow cytometry samples (2 mL)
were preserved (0.5% paraformaldehyde, v/v, final concentration),
then frozen at 80 1C until analyzed. Thawed samples were
stained with Hoechst 33342 (1 mg mL 1, v/v, final concentration)
at room temperature in the dark for 1 h (Campbell and Vaulot,
1993; Monger and Landry, 1993). Stained aliquots (100 mL) were
analyzed using a Beckman-Coulter EPICS Altra flow cytometer
mated to a Harvard Apparatus syringe pump for volumetric
sample delivery. Simultaneous (co-linear) excitation of plankton
was provided by two argon ion lasers, tuned to 488 nm (1 W) and
the UV range (200 mW). The optical configuration distinguished
near-bottom temperature at the 46-m mooring revealed recurring,
rapid decreases in temperature on the order of 0.5 1C over 3 min.
These temperature signals were distinguishing features in the time
series, as variability in bottom temperature was otherwise very low
(9mean9¼ 0.029 1C min 1). In order to isolate and further study
these events, a fixed set of criteria was established to objectively
identify individual events throughout the time series.
To be considered a cold-pulse event, two conditions had to be
met: (1) the rate of change of temperature (T) with respect to
time (t), (DT/Dt), must maintain a gradient of r 0.3 1C
(5 min) 1 (the equivalent of 0.015 1C sample 1); and (2) the
total decrease must be r 0.3 1C.
T i þ n T i
r 0:3 1Cð5 minÞ1
t i þ n t i
ð1Þ
T i þ n T i r0:3 1C
ð2Þ
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
populations on the basis of chlorophyll a (red fluorescence,
680 nm), phycoerythrin (orange fluorescence, 575 nm), DNA (blue
fluorescence, 450 nm) and forward and 901 side-scatter
signatures. Calibration beads (0.5 mm and 1.0 mm yellow–green
beads and 0.5 mm UV beads) were used as fluorescence standards.
Raw data (listmode files) were processed using the
software FlowJo (Treestar Inc., http://www.flowjo.com). Water samples were also filtered through acid-washed GHP filters and analyzed on a Technicon Model AAII to determine nitrate concentration.
5
2.6. Analysis of acoustic data
As an instructive guide to whether cold pulses were collectively associated with the development of turbulent microstructure or horizontal re-positioning of scattering objects, we
compared vertically-integrated water-column scattering prior to
cold-pulse events with that during cold-pulse events. Volume
scattering strength from the 200 kHz echosounder at the 46-m
mooring was integrated from 0.5 m below the surface to 1.5 m
Ti O1 K1 M2
frequency * PSD
1
temperature
0.75
0.5
0.25
0
10−2
10−1
100
cross−shore
along−shore
1
frequency * PSD
101
0.75
0.5
0.25
0
10−2
10−1
100
101
−1
frequency (hr )
20
150
23.3
125
23.1
75
50
0
v (cm s-1)
100
T (ºC)
# cold pulses
10
-10
25
22.9
-20
0
4
2
3
4
5
4
3
2
7
4
2
M2 phase
Fig. 2. (a) Variance-preserving power spectral densities of temperature (upper panel) and near-bottom currents (lower panel) normalized by the respective maximum
values, with the local inertial period and three most energetic tidal constituents labeled for reference; and (b) phase-averaged temperature (squares) and along-shore
current velocity (‘v’, circles), and total number of cold-pulse occurrences (gray shaded boxes) with respect to M2 tidal phase (relative amplitude shown as gray cosine
curve). Dashed lines indicate upper and lower 95% confidence levels for cold-pulse occurrences, determined from chi-square tests for homogeneity over high slack, ebb,
low slack, and flood tides.
6
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
above the bottom over 3 min (45 acoustic pulses; typical duration
of temperature decreases during events) and 15 min (225 acoustic pulses; typical recovery time of observed events) intervals
before each identified cold pulse as well as 3 min and 15 min
starting at the onset of the pulse. Paired t-tests of water columnintegrated acoustic scattering and variance water-column integrated in acoustic scattering were used to determine whether changes
in these parameters as the cold pulses progressed were significant.
In addition, data collected by the CTD profiler and multi-frequency
echosounders during the two shipboard surveys were employed to
determine whether some of the measured acoustic scattering could
be attributed to turbulent microstructure. We considered two methods of predicting estimates of backscattering strength from the CTD
data collected by the profiler. The first method followed an indirect
approach, in which temperature data were first used to calculate I, a
measure of temperature microstructure (Gargett, 1985).
e 3=4
I¼
ð3Þ
2
N n
The data were then further cataloged by defining cutoff points
for high slack, ebb, low slack, and flood tide. Slack tide is
considered here as the phases of the tide associated with low or
reduced flow, thus our demarcations were based on current
velocities rather than M2 phase alone. To define slack tide, we
first averaged M2 tidal current magnitude over p/16 rad
( 23 min) intervals (as in Figs. 2b, 3). We then located the two
peaks in current magnitude associated with maximum ebb tidal
flow and maximum flood tidal flow, and defined high and low
slack tide as the phases j of the M2 tide when currents were less
than one-half those peak magnitudes. The resulting cutoff points
were (j expressed in rad of the 2p rad M2 cycle, with high tide
corresponding to 0 or 2p rad): high slack (30p/16r j o3p/16);
ebb (3p/16r j o12p/16); low slack (12p/16 r j o16p/16); and
flood (16p/16r j o30p/16). Spring and neap tidal phase demarcations were based on the local moon phase.
2.8. Buoyancy frequency and shear
2
where e is the rate of dissipation of turbulent kinetic energy, N is the
buoyancy frequency, and n is the kinematic viscosity of seawater.
Turbulent dissipation rates were estimated from overturning scales
(Thorpe, 1977) in the shipboard CTD casts, following Dillon (1982):
e ¼ 0:64 Nov 3 LT 2
Time series measurements of buoyancy frequency were calculated to quantify changes in background stratification averaged
over time scales of several days. Buoyancy frequency at the 46-m
ð4Þ
where Nov is the buoyancy frequency (N) computed from the
re-ordered (stable) st profile, and LT is the Thorpe scale:
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
g dst
ð5Þ
N¼
r0 dz
20
where g is the acceleration due to gravity, r0 is the mean watercolumn density, st is r(T, S, p¼0) 1000, S is salinity, and z is
depth, with dz¼50 cm. False inversions caused by measurement
noise were eliminated following Galbraith and Kelley (1996).
These values for I were then used to estimate acoustic backscattering levels using the classical model of Goodman (1990).
A second estimate of backscatter from microstructure was
obtained by explicitly following the methods of Goodman (1990)
and Warren et al. (2003). A full derivation of this approach can be
found in Warren et al. (2003). To summarize briefly, in this
approach temperature, salinity, and acoustic scattering data were
combined with models of acoustic scattering to conduct a leastsquares inversion to predict the level of acoustic backscatter
attributable to microstructure. It is important to note that our
profiler did not measure turbulent dissipation directly, thus our
calculations of turbulence are approximations.
Finally, time-series measurements of water column-integrated
acoustic scattering during the shipboard cold pulses were combined with inspection of the acoustic frequency responses for
selected regions both outside the cold pulse and in a strongscattering feature at the cold-pulse interface to determine
whether the passage of the cold pulses were associated with a
change in the type or quantity of scatterers.
15
2.7. Tidal analysis
Spectral analysis of temperature, currents, and pressure fluctuations over the 21 d mooring deployment in 2009 revealed a high
degree of variability at the M2 frequency (Fig. 2a). To focus on the
influence of the M2 tidal constituent, we performed harmonic
analysis on pressure data from the frame-mounted ADCP at the
25-m mooring to extract the amplitude and frequency of individual
constituents using the t-tide Matlab toolbox (Pawlowicz et al., 2002),
then calculated the M2 tidal phase at each time-index in the nearbottom temperature record at the 46-m mooring.
EBB
v (cm s-1)
10
5
LOW
SLACK
0
HIGH
SLACK
-5
-10
FLOOD
-15
-10
-5
0
u (cm s-1)
5
Fig. 3. Hodograph of near-bottom cross-shore (‘u’) and along-shore (‘v’) M2
currents at the 46-m mooring. Each arrow represents a p/16 rad average of the
2p rad M2 cycle.
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
mooring was calculated by combining temperature data at the
four thermistors on the 46-m thermistor chain with salinity data
at the 25-m mooring, using Eq. (5), with dz¼11 m, and assuming
dS/dz¼0. There was no significant rainfall in the area during the
study, nor are there significant land-based sources of fresh-water
during the spring and summer, thus the observed salinities fell
within a narrow range (35.03–35.24, D0.21). Shear during coldpulse events at the 25-m mooring was calculated using data from
the 300 kHz ADCP as
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2
du
dv
S¼
þ
ð7Þ
dz
dz
where u(z) and v(z) were cross-shore and along-shore (respectively)
current velocities as functions of depth (z), and dz¼5 m.
3. Results
3.1. Physical oceanographic setting of study area
Inspection of 13 months of wind data between 2009 and 2010
revealed that winds at Makua Range were most often 5–8 m s 1
out of the northeast. However, these relatively strong and persistent trade winds were often replaced by weaker, diurnal land/sea
breezes for days to weeks at a time. We observed a distinct change
in wind patterns beginning 01 May at the study site during the
2009 deployment. Between 20 and 30 April 2009, winds were
strong and consistently out of the east at speeds of 4.471.7 m s 1
(mean7SD), with gusts up to 15.2 m s 1. Between 01 and 12 May
2009, winds weakened and transitioned to a diurnal pattern.
During this period, wind speeds averaged 2.570.8 m s 1 with
gusts of up to 9.8 m s 1, with daytime sea breezes from the
northwest and nighttime land breezes from the southeast.
Stratification is generally very low in the shallow (o50 m)
coastal waters surrounding the island of Oahu (Hamilton et al.,
1995). Monthly CTD casts taken as part of the Hawaii Ocean Time
Series program show a year-round thermocline at 70–90 m. The
shift to weaker, diurnal winds beginning 01 May 2009 coincided
with increased inner-shelf stratification at the study site. Between
20 and 30 April 2009, the mean buoyancy frequency (N2) was
4.77 10 5 rad2 s 2, while between 01 and 12 May 2009, the mean
buoyancy frequency increased to 9.41 10 5 rad2 s 2. This increase
7
in stratification was due in part to increased surface warming, which
was likely a reflection of the increase in spring air temperatures (in
addition to decreased wind speeds) that occurred at this time.
The semidiurnal surface tide propagates towards the Hawaiian
Islands from the north–northeast (Flament et al., 1996). The
surface tide near Oahu is mixed semidiurnal, with two highs
and two lows of unequal height each day. Tidal height ranged
from 0.04 m to 0.63 m during spring tide and from 0.02 m to
0.55 m during neap tide. Despite a relatively small tidal range, the
power spectral density of near-bottom temperature, currents, and
pressure at the 46-m mooring all showed maximum energy at
semidiurnal frequencies (Fig. 2a). The spring–neap cycle was
evident by modulations in along-shore current speeds. The 2009
study spanned two full spring tides and one full neap tide; the
average daily maximum current magnitude was 36 cm s 1 over
the two spring tidal periods and 30 cm s 1 over the neap tidal
period.
Nearshore currents off the leeward coast of Oahu were largely
directed along isobaths, oscillating between southerly (flood tide)
and northerly (ebb tide) with the M2 cycle (Figs. 2b, 3). At the
46-m isobath, the maximum cross-covariance of sea-surface
height and near-bottom along-shore current velocity occurred at
a lag of 3.3 h (951 out of phase), indicating that the peak in
northerly currents occurred 3.3 h after high tide. Near-bottom
temperature was higher on average during ebb tide, when
currents were directed towards the north; and lower on average
during flood tide, when currents were directed towards the south
(Fig. 2b). As slack tide approached and the flow reversed, currents
rotated either towards shore or away from shore, briefly exhibiting an increase in the cross-shore component of flow (Fig. 3). The
direction of rotation varied, but was much more predictable
during the transition from flood to ebb (high slack) tide. Over
the 40 transitions from flood to ebb observed during the 2009
study period, near-bottom currents rotated away from shore 39
times (97.5%) and towards shore one time (2.5%). Over the same
number of transitions from ebb to flood (low slack), currents
rotated away from shore 23 times (57.5%) and towards shore 17
times (42.5%).
Whereas along-shore currents were very consistent and predictable, rotating only twice per M2 cycle, cross-shore currents
were highly variable at shorter time scales. The power spectra of
near-bottom along-shore currents at 46 m showed maximum
tidal height
(m)
0.5
0.3
(cm s−1)
0.1
30
near−bottom currents
0
−30
( ºC)
24
near−bottom temperature
23
22
00:00 02:00 04:00 06:00 08:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 00:00
Fig. 4. One-day time series of tidal height, near-bottom currents, and near-bottom temperature at the 46-m mooring, illustrating several cold-pulse events (gray lines).
8
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
energy at 12.4 h, whereas that of cross-shore currents showed a
broad peak ranging between 0.8 h and 2.5 h. Although brief and
relatively weak in magnitude, these cross-shore flows coincided
with marked changes in bottom temperature, with onshore
currents resulting in decreasing temperatures, and offshore currents resulting in increasing temperatures (Fig. 4).
T (°C)
22.8
23.15
23.5
1547 h
3.2. Changes to the physical environment during cold-pulse events
1612 h
1645 h
1705 h
1728 h
depth (m)
Over the 21-d study period in 2009, a total of 314 cold pulseevents were identified at the 46-m mooring using the criteria
outlined in Section 2.3. Following the rapid temperature decrease,
events were most typically characterized by a gradual recovery
lasting 5–15 min, which was often overlaid by higher frequency
saw-tooth oscillations (Fig. 4). However, some events had a very
rapid recovery (as fast as 15 s, the sampling interval of the
thermistor) followed by several more temperature oscillations
of period 4–20 min. On rare occasions, the leading signal of an
event was an abrupt increase in temperature. Another type of
event involved an abrupt decrease in temperature that remained
low for 1.5 h to 3 h, before temperature gradually recovered over
10–30 min. Over all 314 events, the average time elapsed
between the initial decrease in temperature and the minimum
temperature was 3.1 min, with 80% of events recovering within
20 min. Cold-pulse ‘groups’ (defined as any event occurring
within 20 min of a previous event) consisted of between one
and six individual pulses, with the majority of groups consisting
of one pulse.
For each cold-pulse event identified at the 46-m mooring,
near-bottom currents were calculated over three periods: (1) a
fixed time interval prior to the onset of the event; (2) the time
between the initial temperature decrease and the temperature
minimum; and (3) the time between the temperature minimum
and the temperature recovery. Between 13.5 min and 10.4 min
prior to the onset of events, the lagged cross-covariance function
of bottom temperature and cross-shore currents was below the
95% significance level; over this time interval the average crossshore current velocity was 2.2 cm s 1 (offshore). Between the
initial decrease and the temperature minimum, cross-shore
currents averaged þ5.4 cm s 1 (onshore), whereas between the
temperature minimum and temperature recovery cross-shore
currents weakened to (on average) þ0.3 cm s 1. The overall mean
cross-shore current when no cold-pulse event was present averaged
2.3 cm s 1 (offshore). The vertical component of currents averaged
1.1 cm s 1 upwards during events and 0.3 cm s 1 upwards when no
events were present.
Visual inspection of the velocity profiles from the 25-m ADCP
revealed a high degree of shear in the cross-shore direction during
cold pulses, with bottom currents flowing onshore and surface
currents flowing offshore, consistent with a shoreward-propagating
mode-1 internal wave. Over the 2009 study, 91 cold-pulse events
were identified at the 25 m site. For each event, shear was calculated
between the onset of the temperature decrease and the temperature
recovery. The average of the maximum shear between onset and
recovery for all 91 cold pulses was 0.05170.018 s 1.
Of the 314 cold pulses observed at the 46-m isobath, 39%
reached the 20 m isobath and 5% reached the 10 m isobath. An
example of the cross-shore progression of one cold pulse event is
shown in Fig. 5. Initially, a pocket of cold bottom water was
present at 46 m depth (panel a). As the feature began to develop,
offshore isotherms were raised (a and b). Cold bottom water was
then advected shoreward by onshore bottom currents of 5–
10 cm s 1 (b and c). Nearshore isotherms then raised while
offshore isotherms lowered (c–e). The cold water mass propagated along the bottom as far shoreward as the 20 m isobath,
where it bulged upward, occupying approximately half of the
0
10 cm s−1
10
20
30
40
50
1200
1000
800
distance from shoreline (m)
1804 h
600
Fig. 5. Cross-shelf view of the progression of one cold-pulse event. Temperature
contours were derived from data from thermistors at the 46, 20, and 10-m
isobaths (black circles, panel e). Cross-shore current velocity vectors are from the
frame-mounted ADCP at the 25 m isobath, and represent 6th degree polynomial
fits. Velocity scale is given in panel f. Local time of day is indicated to the right of
each panel.
water column (d–e). Bottom currents then weakened or reversed,
the cold water moved back offshore, and warmer waters returned
near the surface (f). These observations are consistent with the
shoreward-propagating internal tidal bores observed by Pineda
(1994) off southern California.
The near-bottom isotherm at the 46 m site was displaced
upward during events by an average of 9.14 m, or 20% of the full
water-column depth. The largest isotherm displacement was
30.2 m, or 65% of the full water-column depth. Six of the 314
events occurred when the overlying water column was weakly
thermally stratified, making it difficult to infer the displacement
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
40
30
−70
20
−80
−90
10
−15
SV (dB)
−60
m above bottom
of the bottom isotherm at the onset of the cold pulse; these six
events were not included in this analysis.
Cold-pulse characteristics were similar between the two
deployment periods. Over the 2009 deployment, the average
temperature decrease over each cold-pulse event at the 10-m
(25-m) mooring was 0.50 1C (0.46 1C), and the maximum decrease
was 1.52 1C (1.30 1C). Over the 2010 deployment, the average
temperature decrease at the 25-m mooring was 0.46 1C, and the
maximum decrease was 1.08 1C.
Using the monthly deep CTD casts from the HOT program as a
reference, the average depth at which temperature was 0.50 1C
cooler than the temperature at the same cast at 46 m depth, was
at a depth of 68 m. This implies, assuming strictly cross-shore
propagation, that water masses pulsing to the 46-m site originated from approximately the 68 m isobath, a linear distance of
146 m. Assuming a constant particle velocity of þ5.4 cm s 1, it
would take 45 min for the cooler water to reach the 46 m isobath.
Prior studies indicated that cold pulses on the south shore of
Oahu were higher in salinity than ambient water (Sevadjian et al.,
2010), thus the 25-m mooring data were utilized to investigate
whether the same was true on the west shore of Oahu. The mean
change in salinity over all events identified at the 25-m site in 2009
was þ0.026 psu; the mean change in st was þ0.152 kg m 3. Coldpulse events occurred at all phases of the M2 surface tide; however,
a significantly (495% CI) greater proportion of events occurred
during low slack tide (12p/16r j o16p/16) (see Fig. 2b, gray bars).
Significance levels were determined from chi-square tests for
homogeneity, which determined the likeliness that a given number
of cold pulses would have occurred over a defined period of
time—in this case, the demarcated phases of the M2 tide. During
the transition from ebb to flood tide, near-bottom currents often
rotated towards shore and, as a result, a greater proportion of cold
pulse events occurred during low slack tide. In contrast, a significantly lower proportion of cold pulses occurred during flood tide,
when currents were most often directed within a small azimuthal
spread to the south.
We found no significant difference in the number of events
that occurred during spring and neap tides, suggesting the fortnightly modulation in along-shore-current magnitude was not
directly related to cold pulse occurrence. Likewise, the average
temperature decrease and recovery time of events were not
significantly different between spring and neap tide.
Beginning 01 May 2009, wind speeds decreased and shifted to
a diurnal pattern, and water column stratification (N2) increased.
The ratio of DT/Dz during events to DT/Dz when no events were
present was higher prior to 01 May, indicating that events that
occurred during this period were more pronounced relative to
background conditions. Between 20 and 30 April 2009, there was
an average of 11.3 events per day, whereas between 01 and 12
May 2009 there was an average of 18.8 events per day. Events
also appeared to propagate farther shoreward after 01 May. The
ratio of events that occurred at 46 m, 20 m and 10 m prior to 01
May was 1:0.27:0.03, whereas after 01 May this ratio was
1:0.46:0.06. Thus, after 01 May not only did more temperature
decrease events occur at all depths, but also a greater proportion
of events seen at the offshore mooring were also seen farther
shoreward. Temperature decreases were of similar magnitude
during the period of weaker stratification (mean¼0.47 1C) and
during the period of higher stratification (mean ¼0.51 1C).
9
−10
−5
0
5
10
time (min)
15
20
Fig. 6. Example of acoustic backscatter signal during a cold-pulse event. Time axis
origin is referenced to the onset of the event as determined from the bottom
thermistor data at the 46-m mooring, as outlined in Section 2.3. (Because the
echosounder was 90 m south of the 46-m thermistor mooring, the scattering
signature of a given event did not occur simultaneously with the temperature
decrease).
ending at 400–600 h the following morning, as a result of the diel
horizontal migration of the mesopelagic scattering layer (BenoitBird and Au, 2001, 2004; McManus et al., 2008). Thus, further
analyses of acoustic data during cold pulses were restricted to
daytime events. Inspection of the daytime acoustic scattering data
revealed a consistent set of characteristics during the more
energetic cold-pulse events (e.g., during events in which the
bottom isotherm displacement, magnitude of temperature decrease,
and time rate of change of temperature were all greater than the
average event), which were most pronounced during the onset of the
leading cold pulse in a series (Fig. 6). These more energetic events
were characterized by (1) an abrupt increase in volume-scattering
strength (SV) near the sea floor at the onset of the cold pulse;
(2) upward movement of high-intensity scattering, in concert with
the upward movement of the bottom isotherm; (3) development of
layers of reduced scattering directly above and below the highscattering water; (4) in some cases, an increase in scattering at the
bottom of the displacement resembling a trapped core-like feature;
and (5) a breaking-up and disorganization of the high intensity
scattering region following its maximum vertical displacement (typically 3–4 min following the onset).
Over the combined pool of the 123 cold-pulse events that
occurred during daytime hours, paired t-tests on the watercolumn integrated acoustic scattering and the variance in
water-column integrated acoustic scattering showed no significant differences in scattering before and during the cold pulses
(pc0:05 for all comparisons, b between 0.86 and 0.91 for all
comparisons). Thus, there was no systematic increase in the total
amount of scattering in the water column after the onset of cold
pulses, as well as no change in the variability in water-column
integrated acoustic scattering.
3.4. Shipboard surveys: high-resolution measurements of physics,
biology, optics, chemistry and acoustics
3.3. Characteristics of acoustic scattering during cold-pulse events
Each night during the 21-d mooring deployment in 2009,
acoustic scattering was dominated by the presence of highly
mobile, micronektonic animals beginning at approximately
1900–2100 h (all times herein are reported in local time) and
Changes in temperature, salinity, fluorescence, and acoustic
scattering were similar over the cold pulses sampled during the
shipboard surveys in 2009 and 2010. During the 2009 survey, the
cold pulse sampled occurred at low slack tide, as bottom currents
rotated towards shore from the north to the south. At the onset of
10
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
0.3
0
24.04
15
23.75
30
0
23.46
35.16
15
35.13
30
0
35.1
23.92
15
23.815
30
0
23.71
3.29
15
1.72
30
0
0.15
4.44
15
4.39
30
0
4.34
50
15
0
30
0
−50
50
15
0
u (cm s−1)
O2 (mL L−1) Fluor (mg m−3)
σt
S
T (°C)
0
v (cm s−1)
depth (m)
tidal height (m)
18 Apr 2010
0.6
−50
30
13:00
15:00
17:00
19:00
21:00
23:00
Fig. 7. Profiles of physical, optical, and chemical properties of the water column during the shipboard survey 18 April 2010: (a) tidal height; (b) temperature; (c) salinity;
(d) st; (e) chlorophyll a fluorescence; (f) dissolved oxygen; (g) cross-shore current velocity (‘u’, positive towards shore); (h) along-shore current velocity (‘v’, positive to the
north).
the cold pulse, between 1743 h and 1929 h, near-bottom temperature decreased by 0.65 1C, salinity increased by 0.007 psu,
and fluorescence increased by 1.0 mg m 3. These increases
represent 4, 2, and 3 standard deviations from the means of
temperature, salinity, and fluorescence outside the cold pulse,
respectively. During the 2010 shipboard survey, the cold pulse
sampled occurred at high slack tide, again as bottom currents
rotated towards shore (Fig. 7). At the onset of this cold pulse,
between 1859 h and 1908 h, near-bottom temperature decreased
by 1.10 1C, salinity increased by 0.115 psu, and fluorescence
increased by 3.4 mg m 3. These increases represent 14, 14, and
7 standard deviations from the means of temperature, salinity,
and fluorescence outside the cold pulse, respectively. A second
estimate of chlorophyll a concentration within and above the cold
pulse obtained from the ac-9 absorption spectra was in agreement with the fluorometer data. There was no noticeable change
in oxygen concentration associated with either cold pulse.
Three bottle samples were taken during the 2010 survey while
the pulsed water was evident at our survey station; one bottle
from within the cold pulse and two bottles from directly above the
cold pulse. Phytoplankton abundances (cells mL 1) were higher
within the cold pulse for Prochlorococcus, Synechococcus, and
photosynthetic eukaryotes by up to factors of 1.3, 1.5, and 3.1,
respectively, compared to that of waters above the cold pulse
(Table 2). In addition, average chlorophyll fluorescence per cell was
higher within the cold pulse for Prochlorococcus, Synechococcus, and
photosynthetic eukaryotes by up to factors of 1.5, 3.8, and 1.3,
(respectively), compared to that of waters above the cold pulse.
Non-pigmented bacteria levels were similar for waters within and
above the cold pulse. Changes in the shape of the ac-9 absorption
spectra during the survey suggested that the pulsed water
consisted of relatively less detritus and more phytoplankton
relative to waters outside the cold pulse. Nitrate concentration
within the pulsed water (0.17 mmol L 1) was a factor of two higher
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
11
Table 2
Phytoplankton abundances (cells mL 1) and population-specific average chlorophyll fluorescence per cell (fluor cell 1) estimated using flow cytometry, and nitrate
concentrations for samples taken within and above the cold pulse observed during the shipboard survey 18 April 2010. Phytoplankton populations distinguished by this
methodology are Prochlorococcus (PRO), Synechococcus (SYN), and Photosynthetic Eukaryotes (PEUK). Also enumerated are non-pigmented bacteria (HBACT). Values
reported for fluor cell 1 represent the average cell-specific chlorophyll fluorescence recorded for each population by the flow cytometer.
Sample time
19:23
19:33
20:10
Sample depth
20 m (within)
14 m (above)
20.4 m (above)
PRO
SYN
PEUK
HBACT
[NO3 ]
Cells mL 1
( 104)
Fluor cell 1
Cells mL 1
( 103)
Fluor cell 1
Cells mL 1
( 103)
Fluor cell 1
Cells mL 1
( 105)
mmol L 1
6.0
4.5
4.9
4.1
2.7
3.2
6.1
4.0
5.2
16.9
4.4
5.6
4.3
1.4
1.8
73.1
55.9
46.4
2.2
2.1
2.2
0.17
0.09
0.08
than that of waters above the cold pulse (0.09 mmol L 1 and
0.08 mmol L 1).
Measurements of acoustic backscatter were taken during both
shipboard surveys using a multi-frequency downward-looking
echosounder package, as well as a TAPS mounted on the profiler.
In both cases, the appearance of shoreward-migrating micronekton could clearly be identified at the site and did not occur until
well after the passage of the cold pulse. The temporal pattern of
variation in backscatter was similar to that measured by the
moored echosounder at the 46 m isobath during cold pulses, as
described in Section 3.3. Volume scattering during both shipboard
survey events intensified near the sea floor at the beginning of the
cold pulse, moved upward in concert with the isotherm displacement, and became less organized after the cold pulse, before
reorganizing and returning to the sea floor during the recovery
stage (Fig. 8). As was observed in the moored echosounder data
during the 2009 deployment, high levels of acoustic scattering
associated with the cold pulses were restricted to the upper
boundary of the cold water mass, while scattering just beneath
the cold pulse interface and in the overlying water was relatively
low. Although vertical displacements of scattering objects were
clearly apparent in the echograms, the mean water-column
volume scattering strength at 200 kHz was 62 dB both before
and during each of the shipboard cold-pulse events. The spectral
response of scattering from TAPS data from within the highscattering feature at the cold pulse interfaces showed a similar
pattern to that from waters both above the cold pulse and
throughout the water column prior to the arrival of the cold
pulse, suggesting there was no change in the type of scattering
objects (Fig. 8, lower panels). Inversion of the TAPS data (Holliday
and Pieper, 1995) indicated that scattering at the cold pulse
interface, above the cold pulse, and throughout the water column
prior to the cold pulse, was dominated by spheroid, fluid-like
scatterers with an equivalent spherical diameter of approximately
1.0–1.5 mm. Zooplankton net tows conducted during the 2009
deployment at the 25 m mooring site – both those that covered
the entire water column and those that targeted acoustic features
– were dominated in both number and biomass by calanoid
copepods between 0.5 mm and 2.0 mm diameter, similar in size
and shape to the dominant scatterers predicted by the TAPS. The
most abundant group was Clausocalanus spp. A separate band of
high scattering appeared after the final observations of the 2010
cold pulse at approximately 2050 h, but this scattering was likely
associated with along-shore advection of a separate water mass,
based on strong northerly currents and a distinct oxygen signal at
this time, but no corresponding signal in temperature, salinity, or
fluorescence.
We followed two approaches to determine whether scattering
during the cold-pulse events measured from the shipboard
sampling in 2009 and 2010 could have been dominated by
turbulent microstructure. In the first approach, we calculated I
(a measure of temperature microstructure), then estimated
scattering levels predicted for these values of I following
Goodman (1990). Goodman (1990) provides examples of I under
various ocean conditions, classifying I¼11 as the average background open ocean turbulence level, and I¼ 3000 as ‘very intense’
turbulence. All calculated values for I during the shipboard
surveys at depths greater than 2 m were o3.5. Goodman
(1990) predicts that where Io3.5, scattering within the frequency
range of the moored and ship-mounted echosounders used in this
study should be less than 110 dB. This is three orders of
magnitude lower than the 80 dB threshold of the moored
echosounder and four to five orders of magnitude lower than
the median volume-scattering strength of 70 dB to 60 dB
observed during cold-pulse events.
In the second approach, we combined temperature, salinity,
and acoustic scattering data with models of acoustic scattering to
predict the level of scattering attributable to microstructure.
Using this method, the highest predicted backscatter value at
any of the frequencies of the ship-mounted echosounders was
119 dB, seven orders of magnitude weaker than the scattering
observed at the leading edge of each cold pulse. In addition, the
frequency response for microstructure during these cold pulses is
predicted to decrease with increasing frequency over the range
measured here (Lavery et al., 2007), opposite to the observed
spectral response during both events.
4. Discussion
4.1. Cold pulses and internal tides
The observed pulses of cold bottom water appear to be the
manifestation of shoreward propagating internal waves. The
existence of internal waves on shelves has been attributed almost
exclusively to shoaling tidally-generated waves, generated offshore at regions of abrupt bathymetry. Other generation mechanisms, such as buoyant plume propagation (Nash and Moum, 2005;
Woodson et al., 2011) have been proposed recently, but implicate
conditions (in particular, the presence of a convergent front) not
typically found off leeward Oahu.
During the 2010 shipboard survey, we observed a decrease in
temperature of 1.10 1C and increase in fluorescence of 3.4 mg m 3
associated with one cold pulse. Using the monthly deep CTD casts from
the HOT program as a reference, the average depth at which
temperature was 1.10 1C cooler than the temperature at the same cast
at 46 m depth, was at a depth of 102 m. Fluorometry data from these
deep casts indicate a persistent chlorophyll maximum between 105 m
and 130 m. Thus, it is likely that the observed increase in fluorescence
at the 25 m station was a reflection of an elevation of isopycnals and
pulsed delivery of water from the deep chlorophyll maximum.
In some coastal locations, shoaling internal waves can give rise
to temperature changes of 5 1C over 1–20 min (Cairns, 1967;
Leichter et al., 1996; Winant, 1974). The magnitude of
12
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
Fig. 8. Acoustic scattering during the shipboard survey 18 April 2010: (upper panel) echogram of volume scattering strength (SV) from the 200 kHz echosounder, overlaid
with st ¼ 23.78 and 24.06 isopycnals. The lower panels show volume scattering for three regions highlighted with black boxes in the 200 kHz echogram: (A) before the cold
pulse, (B) at the same depth near the beginning of the cold pulse, and (C) within the high-scattering feature at the cold pulse interface. In all three panels, circles indicate
mean volume scattering measured from the ship-mounted echosounders over a 30 s by 1 m depth interval while crosses represent mean volume scattering measured
using TAPS data from a single profile integrated over 1 m vertically. The dashed line represents the predicted volume scattering response from 1.25 mm diameter spheroid
fluid-like scatterers relative to the volume scattering measured using the 200 kHz shipboard echosounder.
temperature decreases in the tropical waters off the leeward coast
of Oahu, Hawaii was small in comparison to these locations.
However, changes in isotherm depth, acoustic backscatter distribution, and phytoplankton distributions were substantial. Thus
it appears that the relative change in temperature with respect to
baseline stratification at a particular coastal site is more important than the magnitude of the temperature change. This finding
suggests that in many locations, the full impact of these types of
events may commonly be overlooked, either because too little
significance is placed on changes to the physical environment that
appear relatively small in magnitude, or because interdisciplinary
data is not available. The ubiquity and significance of shorewardpropagating cold pulses may be dramatically underestimated for
these reasons.
In contrast to some continental shelf studies (Leichter et al.,
1996; Ramp et al., 2004; Storlazzi et al., 2003), rapid bottom
temperature changes on the leeward Oahu insular shelf occurred
at all phases of the M2 tide, with only a slight bias towards low
slack tide, as opposed to occurring predictably within a small
phase window. Similarly, Storlazzi and Jaffe (2008) observed
rapid bottom temperature changes on the West Maui insular
shelf that occurred roughly once each day, but found no relationship between occurrence and diurnal tidal phase. These findings
are likely a reflection of the lack of a direct propagation path of
internal tidal energy between the known internal tide generation
sites and the study area. The Hawaiian Ridge is a site of
substantial energy loss from the surface tide to the internal tide
(Carter et al., 2008; Egbert and Ray, 2001; Rudnick et al., 2003).
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
Models of internal wave energy throughout the Hawaiian Ridge
suggest that large semi-diurnal energy fluxes radiate away from
Kaena Ridge ( 45 km west–northwest of the study site) to the
south–southeast and north–northwest, with the vast majority of
internal tidal energy found away from the Oahu coastline, in
depths 41000 m (Merrifield et al., 2001; Merrifield and
Holloway, 2002; Rudnick et al., 2003). A second, weaker internal
tide generation site is located off Makapu’u Point on the southeast
tip of Oahu. Energy flux vectors here are oriented clockwise from
the generation site, with energy radiating in towards Mamala Bay
on the south shore of Oahu, and interacting at variable longitudes
with energy generated near Kaena Point (Alford et al., 2005; Eich
et al., 2004). This interference from multiple sources, as well as
the complex bathymetry of this region, makes the internal tide
field south of the Hawaiian Ridge highly variable (Rainville et al.,
2010). Thus, the degree of coherence between cold pulses and
surface tidal phase on insular shelves throughout the Hawaiian
Islands is also likely variable.
4.2. Internal wave characteristics
Cold pulses varied in shoreward extent of propagation and in
recovery time. The cross-shore progression of one cold pulse
event is shown in Fig. 5. The evolution of the water mass shown in
this figure is typical of events that had a relatively long recovery
time (1.5–3 h), such as the event we sampled during the 2010
shipboard survey, and appears to be consistent with the set of
dynamics Pineda (1994) attributed to a shoreward propagating
internal tidal bore. Pineda (1994) eloquently summarized the
dynamics of shoreward-propagating internal tidal bores in southern California. As the author describes, in the first phase, cold
subsurface water is advected shoreward by internal tidal bores.
The advection of dense water to the nearshore is then responsible
for setting up a horizontal pressure gradient in relation to the
lighter offshore water. In the second phase, this in turn causes
bottom waters to return offshore as a gravity current, setting up
an onshore movement of warm surface waters. The downward
tilting of isotherms behind the cold water front is further
suggestive of an internal bore rather than a solitary wave, which
would have more symmetrical isotherm displacements (Moum
et al., 2007).
Acoustic scattering from the moored 200 kHz echosounder
during cold-pulse events had the characteristic of upward movement of scatterers in concert with the upward movement of the
bottom isotherm. This is likely a reflection of the waves propagating as waves of elevation, as opposed to waves of depression.
When the pycnocline is closer to the sea floor than the sea surface
(as is often the case in shallow Hawaiian waters), internal tidal
waves propagate as waves of elevation (Scotti and Pineda, 2004).
The most studied generation mechanism for a wave of elevation
in the nearshore is a wave of depression propagating along a
shoaling bottom (Scotti and Pineda, 2004).
During the more energetic events, high levels of scattering
were found at the boundary between the cold pulse and the
surrounding water. Similar observations of acoustic scattering
during internal waves were reported by Haury et al. (1979) and
Moum et al. (2007). Haury et al. (1979) found strong density
inversions associated with acoustically-observed overturns in
Massachusetts Bay using a 200 kHz echosounder. Moum et al.
(2007) measured potential density and backscatter at 120 kHz,
and calculated turbulent dissipation from a turbulence profiler as
an internal wave propagated along the Oregon continental shelf.
Moum et al. (2007) showed strong agreement between high
scattering regions and high turbulence in the wake of the trailing
edge of the wave, although it was not clear whether scattering in
those instances was due to turbulence directly, or due to a
13
re-positioning of biological scatterers by the turbulent wave. We
find that although the time course of cold-pulse events could be
approximated from measurements from the moored and shipmounted echosounders, cold-pulse events did not significantly
increase water-column integrated acoustic scattering as they
passed. In addition, predicted acoustic scattering levels due to
turbulent dissipation, calculated from CTD data collected during
the shipboard surveys were 4–7 orders of magnitude weaker than
the observed scattering. Although our profiler did not measure
turbulent dissipation directly, and thus our calculations of turbulence are approximations, these findings suggest that turbulent
microstructure was not a dominant source of acoustic scattering
during cold-pulse events.
Examination of the spectral response of the scattering during
the shipboard cold-pulse events indicated that the primary
scatterers within the high-scattering features at the cold-pulse
interfaces, above the cold pulses, and throughout the water
column prior to the arrival of the cold pulses, were 1.0–1.5 mm
diameter fluid-like spheroids. This is consistent with net tows
conducted at the study site that showed the dominant mesozooplankton to be 0.5–2.0 mm diameter calanoid copepods. It is
possible that bottom shear associated with cold pulses resulted in
sediment re-suspension, however, typical sediment diameters on
the south shore of Oahu are between 100 mm and 300 mm
(J. Fram, pers. comm.). This size is too small to have been detected
at the relatively low frequencies used by the ship-mounted and
moored echosounders (Holliday and Pieper, 1995). Because the
frequency response and intensity of acoustic backscatter before
and during cold pulses were similar, if a separate zooplankton
community was entrained within the cold pulse as the phytoplankton were, they would have to have been of the same type
and quantity as the surrounding water. Based upon this evidence
we conclude that, during cold-pulse events, zooplankton scatterers were vertically re-distributed at the onset of the cold pulse,
but were not entrained within the cold pulse.
It does not appear that these organisms at the interface
actively migrated downwards into the phytoplankton-rich pulse
waters during the active passage of the cold pulse. It has been
noted that foraging zooplankton, such as copepods, may associate
physical gradients with food presence, and use this association as
a cue for directed search patterns (Woodson et al., 2005, Woodson
and McManus 2007). Woodson et al. (2005) showed in a laboratory study of two copepod species that these animals tended to
swim away from or along density gradients, rather than across
them. The density interface at the edge of the pulsed water may
have acted as a physical or behavioral barrier to movement into
these relatively rich waters (Harder, 1968; Woodson et al., 2005).
A similar pattern was observed on the south shore of Oahu during
a cold pulse in 2008 (Sevadjian et al., 2010). Here, increased levels
of acoustic backscatter associated with zooplankton were also
found in a thin layer at the interface between cold, salty, high
fluorescence pulsed water and the surrounding water.
4.3. Internal motions and stratification
In agreement with studies along the coast of California and in
the Florida Keys, we found that (1) more events occurred when
the water column was more stratified (Pineda and Lopez, 2002);
and (2) proportionally more events were found closer to shore
when the water column was more stratified (Leichter et al., 1996;
Pineda and Lopez, 2002; Storlazzi et al., 2003). As pointed out by
Pineda and Lopez (2002), for a given isopycnal displacement,
internal motions will be more energetic when the water column
has higher levels of stratification. Supporting evidence for the
correlation between stratification and occurrence of bore-like
events has been found in longer term studies in temperate
14
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
regions (Cairns and Nelson, 1970; Pineda, 1991; Storlazzi et al.,
2003). These studies found that more of these types of events
occurred during spring and summer months when stratification
was higher.
copepods likely feed upon smaller phytoplankton, such as the
photosynthetic eukaryotes brought into shallower water by the
cold pulse. Thus, these cold-pulse events could provide extra
nutrition for the copepods, potentially allowing enhanced reproduction (Paffenhöfer et al., 2006).
4.4. Biological implications
We find that waters within a cold pulse contained higher
concentrations of several phytoplankton populations, with elevated levels of chlorophyll per cell. The amount of phytoplankton
chlorophyll per cell varies with the light environment of the cells,
as well as with nutrient status. Higher chlorophyll fluorescence
per cell indicates cells that are adapted to live deeper in the water
column (at lower light levels) and/or have more nutrients available to them. Indeed, nitrate levels within the entrained water
mass were found to be a factor of two higher than that of the
surrounding waters, and the water mass was estimated to have
originated from a depth of approximately 102 m. If these organisms were light-limited, their re-distribution to shallower waters
with higher light conditions could encourage faster growth rates.
Chlorophyll concentrations within the NPSG begin to increase at
depths shallower than nutrient levels increase (Karl and Lukas,
1996). Conceptually, this suggests a two-tiered threshold for coldpulse events, in which some cold pulses elevate water of increased
chlorophyll but similar nutrients, and more energetic cold pulses
elevate water of both increased chlorophyll and increased nutrients to the nearshore. At the 46-m mooring, events with isotherm
displacements at least as large as those during the 2010 shipboard
survey were not uncommon (n¼38, mean¼ 1.8 d 1), suggesting
deep water intrusions off leeward Oahu may regularly be higher in
phytoplankton and nitrate concentrations.
Evidence of similar bottom events occurring at islands and
atolls in the Northwest Hawaiian Islands (spanning a distance of
42000 km) is building; inspection of temperature data from
stations deployed by the National Oceanic and Atmospheric
Administration Coral Reef Ecosystem Division revealed that three
of four locations show episodic temperature spikes similar to our
observations on Oahu, with some sites regularly experiencing
decreases of 5 1C (PIFSC Cruise Report, 2011; J. Gove pers. comm.).
Focused, multidisciplinary studies at such locations are needed to
test the broadness to which the correlation between temperature
spikes and delivery of high-chlorophyll, high-nutrient water
applies throughout the greater NPSG.
It remains to be determined how organisms in coastal habitats
within the NPSG may respond to these seemingly ubiquitous
events. Leichter et al. (2005) proposed that at Conch Reef, Florida,
internal wave events lasting tens of minutes to a few hours may
be frequent and persistent enough for suspension-feeding corals
and macroalgae to benefit from the enhanced nutrient and
suspended particle fluxes associated with the events.
Variability in the shoreward extent of cold-water pulses, such
as that observed during the transition to weaker, diurnal winds
and increased stratification beginning 01 May 2009, can cause
patchiness in food availability, resulting in the formation of
temporary microhabitats (Leichter et al., 1996). Increases in
occurrence and shoreward propagation of cold pulses at coastal
areas within the NPSG could result in a temporary re-distribution
of phytoplankton populations from deeper water into the inner
shelf and nearshore environment. As concentrations of phytoplankton can attract aggregations of mobile zooplankton (see
review by Woodson and McManus, 2007), this may in turn result
in adapted foraging behavior by larger organisms. The mesozooplankton genus Clausocalanus, which was the most abundant
group observed in net tows in our study area, is one of the most
common zooplankton species in the NPSG (Landry et al., 2001,
2008; McGowan and Walker, 1979). Juvenile stages of these
5. Conclusions
Shoreward pulses of cold, high salinity, phytoplankton-rich
bottom water are common off the leeward coast of Oahu, Hawaii.
The occurrence of ‘cold pulses’ was related to the rotation of M2
tidal currents that occurred at low slack tide. Cold pulses were
more common and propagated farther onshore when overall
stratification increased and winds weakened and became more
diurnal. The pulsed water likely originated from depths near the
chlorophyll maximum. While phytoplankton were entrained
within the pulses, zooplankton appeared to aggregate at the
upper boundary of the pulsed water.
Shoreward pulses of cold bottom water also have been
observed on inner shelves of other islands and atolls around the
Pacific. Until this contribution, however, the planktonic responses
to cold pulses within the NPSG had not yet been described. The
vertical distributions of chlorophyll and nutrients off the leeward
coast of Oahu, Hawaii are representative of many sites in the
NPSG, thus, we hypothesize that cold pulses occurring at other
sites throughout the NPSG also bring water higher in phytoplankton concentration, and possibly nutrient concentrations, to shallow insular shelf environments. These events have the potential
to substantially influence the dynamics of marine ecosystems on
inner shelves of insular islands and atolls around the Pacific.
Acknowledgements
We wish to thank Amanda Timmerman and Sarah Feinman for
collection of water samples during the 2010 shipboard survey and
for arranging for nutrient analysis of these samples; Rebecca
Baltes, Jim Eckman, Linnaea Jasiuk, Jennifer Patterson, Ross
Timmerman, Donn Viviani, Gordon Walker, Chad Waluk, and
Jon Whitney for assistance in the field; Jim Sullivan for interpretation of ac-9 optics data; and Captain Joe Reich of the Alyce C,
as well as Captain Ryan Wagner from Sea Engineering Inc. for
their help with deployment and recovery of moorings as well as
with survey work. This work was funded by the Office of Naval
Research awards N00014-08-1-1212 (M.A.M.), and N00014-08-11210 (K.J.B-B.). We are extremely grateful for this support.
References
Alford, M.H., Gregg, M.C., Merrifield, M.A., 2005. Structure, propagation and mixing
of energetic baroclinic tides in Mamala Bay, Oahu, Hawaii. Journal of Physical
Oceanography PKM-000, 1–19.
Benoit-Bird, K.J., Au, W.W.L., 2001. Target strength measurements of animals from
the Hawaiian mesopelagic boundary community. Journal of the Acoustical
Society of America 110, 812–819.
Benoit-Bird, K.J., Au, W.W.L., 2004. Fine-scale diel migration dynamics of an
island-associated sound-scattering layer. Deep Sea Research Part I: Oceanographic Research Papers 51, 707–719.
Benoit-Bird, K.J., Zirbel, M.J., McManus, M.A., 2008. Zooplankton spatio-temporal
distributions in Hawaiian waters favor diel horizontal migration by midwater
micronekton. Marine Ecology Progress Series 367, 109–123.
Bricaud, A., Babin, M., Morel, A., Claustre, H., 1995. Variability in the chlorophyllspecific absorption coefficients of natural phytoplankton: analysis and parameterization. Journal of Geophysical Research 100, 13321–13332.
Cairns, J.L., 1967. Asymmetry of internal tidal waves in shallow coastal waters.
Journal of Geophysical Research 72, 3563–3565.
Cairns, J.L., Nelson, K.W., 1970. A description of the seasonal thermocline cycle in
shallow coastal water. Journal of Geophysical Research 75, 1127–1131.
J.C. Sevadjian et al. / Continental Shelf Research 50–51 (2012) 1–15
Campbell, L., Vaulot, D., 1993. Photosynthetic picoplankton community structure
in the subtropical North Pacific Ocean near Hawaii (station ALOHA). Deep Sea
Research Part I: Oceanographic Research Papers 40, 2043–2060.
Carter, G.S., Merrifield, M.A., Becker, J.M., Katsumata, K., Gregg, M.C., Luther, D.S.,
Levine, M.D., Boyd, T.J., Firing, Y.L., 2008. Energetics of M2 barotropic-to-baroclinic
tidal conversion at the Hawaiian Islands. Journal of Physical Oceanography 38,
2205–2223.
Davis, R.F., Moore, C.C., Zaneveld, J.R.V., Napp, J.M., 1997. Reducing the effects of
fouling on chlorophyll estimates derived from long-term deployments of
optical instruments. Journal of Geophysical Research 102, 5851–5855.
Dillon, T.M., 1982. Vertical overturns: a comparison of Thorpe and Ozmidov length
scales. Journal of Geophysical Research 87, 9601–9613.
Egbert, G.D., Ray, R.D., 2001. Estimates of M2 tidal energy dissipation from TOPEX/
Poseidon altimeter data. Journal of Geophysical Research 106, 22,475–22,502.
Eich, M.L., Merrifield, M.A., Alford, M.H., 2004. Structure and variability of
semidiurnal internal tides in Mamala Bay, Hawaii. Journal of Geophysical
Research 109, C05010.
Flament, P., Kennan, S., Lumpkin, R., Sawyer, M., and Stroup, E.D., 1996. Ocean atlas
of Hawaii. University of Hawaii, Honolulu, HI. Available at: /http://www.
satlab.hawaii.edu/atlas/S.
Foote, K.G., Vestnes, G., Maclennan, D.N., Simmonds, E.J., 1987. Calibration of
acoustic instruments for fish density estimation: a practical guide. ICES
Cooperative Research Reports 144, 57.
Galbraith, P.S., Kelley, D.E., 1996. Identifying overturns in CTD profiles. Journal of
Atmospheric and Oceanic Technology 13, 688–702.
Gargett, A.E., 1985. Evolution of scalar spectra with the decay of turbulence in a
stratified fluid. Journal of Fluid Mechanics 159, 379–407.
Goodman, L., 1990. Acoustic scattering from ocean microstructure. Journal of
Geophysical Research 95, 11557–11573.
Gove, J.M., Merrifield, M.A., Brainard, R.E., 2006. Temporal variability of currentdriven upwelling at Jarvis island. Journal of Geophysical Research 111, C12011.
Hamilton, P.J., Singer, J., and Waddell, E., 1995. Mamala Bay Study, Ocean Current
Measurements: A Report to the Mamala Bay Commission, HI. Technical report.
Science Applications International Corporation, Raleigh, NC.
Harder, W., 1968. Reactions of plankton organisms to water stratification.
Limnology and Oceanography 13, 156–168.
Haury, L.R., Briscoe, M.G., Orr, M.H., 1979. Tidally generated internal wave packets
in Massachusetts Bay. Nature 278, 312–317.
Haury, L.R., Wiebe, P.H., Orr, M.H., Briscoe, M.G., 1983. Tidally generated highfrequency internal wave packets and their effects on plankton in Massachusetts Bay. Journal of Marine Research 4, 65–112.
Holliday, D.V., Donaghay, P.L., Greenlaw, C.F., Napp, J.M., Sullivan, J.M., 2009. Highfrequency acoustics and bio-optics in ecosystems research. ICES Journal of
Marine Science 66, 974–980.
Holliday, D.V., Pieper, R.E., 1995. Bioacoustical oceanography at high frequencies.
ICES Journal of Marine Science 52, 279–296.
Holloway, P.E., 1987. Internal hydraulic jumps and solitons at a shelf break region on
the Australian north west shelf. Journal of Geophysical Research 92, 5405–5416.
Horel, J., Splitt, M., Dunn, L., Pechmann, J., White, B., Ciliberti, C., Lazarus, S., Slemmer, J.,
Zaff, D., Burks, J., 2002. Mesowest: cooperative mesonets in the western United
States. Bulletin of the American Meteorological Society 83, 211–226.
Jackson, C.R., 2004. An Atlas of Internal Solitary-like Waves and their Properties.
second ed., Global Ocean Assoc., Alexandria, Va. (Available at /http://www.
internalwaveatlas.com)S.
Karl, D.M., Lukas, R., 1996. The Hawaii Ocean Time Series (HOT) program:
background, rational, and field implementation. Deep Sea Research Part II:
Topical Studies in Oceanography 43, 129–156.
Korneliussen, R.J., Diner, N., Ona, E., Fernandes, P.G., 2004. Recommendations for
the collection of multi-frequency acoustic data. ICES CM 2004/R 36, 15.
Landry, M.R., Al-Mutairi, H., Selph, K.E., Christensen, S., Nunnery, S., 2001. Seasonal
patterns of mesozooplankton abundance and biomass at Station ALOHA. Deep
Sea Research Part II: Topical Studies in Oceanography 48 (8-9), 2037–2061.
Landry, M.R., Decima, M., Simmons, M.P., Hannides, C.C.S., Daniels, E., 2008.
Mesozooplankton biomass and grazing responses to Cyclone Opal, a subtropical mesoscale eddy. Deep Sea Research II 55, 1378–1388.
Lavery, A.C., Wiebe, P.H., Stanton, T.K., Lawson, G.L., Benfield, M.C., Copley, N.,
2007. Determining dominant scatterers of sound in mixed zooplankton
populations. Journal of the Acoustical Society of America 122, 3304–3326.
Leichter, J.J., Deane, G.B., Stokes, M.D., 2005. Spatial and temporal variability of
internal wave forcing on a coral reef. Journal of Physical Oceanography 35,
1945–1962.
Leichter, J.J., Shellenbarger, G., Genovese, S.J., Wing, S.R., 1998. Breaking internal
waves on a Florida (USA) coral reef: a plankton pump at work? Marine Ecology
Progress Series 166, 83–97.
Leichter, J.J., Wing, S.R., Miller, S.L., Denny, M.W., 1996. Pulsed delivery of
subthermocline water to Conch Reef (Florida Keys) by internal tidal bores.
Limnology and Oceanography 41, 1490–1501.
McGowan, J.A., Walker, P.W., 1979. Structure in the copepod community of the
North Pacific central gyre. Ecological Monographs 49, 195–226.
McManus, M.A., Benoit-Bird, K.J., Woodson, C.B., 2008. Behavior exceeds physical
forcing in the diel horizontal migration of the midwater sound-scattering layer
in Hawaiian waters. Marine Ecology Progress Series 365, 91–101.
15
McManus, M.A., Sevadjian, J.C., Benoit-Bird, K.J., Cheriton, O.M., Timmerman, A.V.,
Waluk, C.M., 2012. Observations of thin layers in coastal Hawaiian waters.
Estuaries and Coasts 35, 1119–1127.
Merrifield, M.A., Holloway, P.E., 2002. Model estimates of M2 internal tide
energetics at the Hawaiian Ridge. Journal of Geophysical Research 107, 3179.
Merrifield, M.A., Holloway, P.E., Johnston, T.M.S., 2001. The generation of internal
tides at the Hawaiian Ridge. Geophysical Research Letters 28, 559–562.
Monger, B.C., Landry, M.R., 1993. Flow cytometric analysis of marine bacteria with
Hoechst 33342. Applied and Environment Microbiology 59, 905–911.
Morozov, E.G., Trulsen, K., Velarde, M.G., Vlasenko, V.I., 2002. Internal tides in the
strait of Gibraltar. Journal of Physical Oceanography 32, 3193–3206.
Moum, J.N., Farmer, D.M., Shroyer, E.L., Smyth, W.D., Armi, L., 2007. Dissipative
losses in nonlinear internal waves propagating across the continental shelf.
Journal of Physical Oceanography 37, 1989–1995.
Nash, J.D., Moum, J.N., 2005. River plumes as a source of large-amplitude internal
waves in the coastal ocean. Nature 437, 400–403.
Owen, R.W., 1981. Fronts and eddies in the sea: mechanisms, interactions and
biological effects. In: Longhrust, A.R. (Ed.), The Analysis of Marine Ecosystems.
Academic press, London, pp. 233.
Paffenhöfer, G.A., Mazzocchi, M.G., Tzeng, M.W., 2006. Living on the edge: feeding
of subtropical open ocean copepods. Marine Ecology 27, 99–108.
Pawlowicz, R., Beardsley, B., Lentz, S., 2002. Classical tidal harmonic analysis including
error estimates in MATLAB using T_TIDE. Computers and Geosciences 28, 929–937.
PIFSC Cruise Report CR-11-002. Issued 15 April 2011. Pacific Islands Fisheries
Science Center, Honolulu, HI. Available at: /http://www.pifsc.noaa.gov/
library/pubs/cruise/Hiialakai/CRHA1007-KD.pdfS.
Pineda, J., 1991. Predictable upwelling and the shoreward transport of planktonic
larvae by internal tidal bores. Science 253, 548–551.
Pineda, J., 1994. Internal tidal bores in the nearshore: warm-water fronts, seaward
gravity currents and the onshore transport of neustonic larvae. Journal of
Marine Research 52, 427–458.
Pineda, J., 1995. An internal tidal bore regime at nearshore stations along western
U.S.A.: predictable upwelling within the lunar cycle. Continental Shelf
Research 15, 1023–1041.
Pineda, J., 1999. Circulation and larval distribution in internal tidal bore warm
fronts. Limnology and Oceanography 44, 1400–1414.
Pineda, J., Lopez, M., 2002. Temperature, stratification and barnacle larval settlement in two Californian sites. Continental Shelf Research 22, 1183–1198.
Rainville, L., Johnston, T.M.S., Carter, G.S., Merrifield, M.A., Pinkel, R., Worcestoer, P.F.,
Dushaw, B.D., 2010. Interference pattern and propagation of the M2 internal tide
south of the Hawaiian Ridge. Journal of Physical Oceanography 40, 311–325.
Ramp, S.R., Tang, T.-Y., Duda, T.F., Lynch, J.F., Liu, A.K., Chiu, C.-S., Bahr, F., Kim, H.-R.,
Yang, Y.J., 2004. Internal solitons in the northern South China Sea, Part I: Sources
and deep water propagation. IEEE Journal of Oceanic Engineering 29, 1157–1181.
Rudnick, D.L., Boyd, T.J., Brainard, R.E., Carter, G.S., Egbert, G.D., Gregg, M.C.,
Holloway, P.E., Klymak, J.M., Kunze, E., Lee, C.M., Levine, M.D., Luther, D.S.,
Martin, J.P., Merrifield, M.A., Moum, J.N., Nash, J.D., Pinkel, R., Rainville, L.,
Sanford, T.B., 2003. From tides to mixing along the Hawaiian Ridge. Science
301, 355–357.
Scotti, A., Pineda, J., 2004. Observation of very large and steep internal waves of
elevation near the Massachusetts coast. Geophysical Research Letters 31, L22307.
Sevadjian, J.C., McManus, M.A., Pawlak, G., 2010. Effects of physical structure and
processes on thin zooplankton layers in Mamala Bay, Hawaii. Marine Ecology
Progress Series 409, 95–106.
Shroyer, E.L., Moum, J.N., Nash, J.D., 2011. Nonlinear internal waves over New
Jersey’s continental shelf. Journal of Geophysical Research 116, C03022.
Small, J., Sawyer, T.C., Scott, J.C., 1999. The evolution of an internal bore at the
Malin shelf break. Annales Geophysicae 17, 547–565.
Storlazzi, C.D., Jaffe, B.E., 2008. The relative contribution of processes driving
variability in flow, shear, and turbidity over a fringing coral reef: West Maui,
Hawaii. Estuarine. Coastal and Shelf Science 77, 549–564.
Storlazzi, C.D., McManus, M.A., Figurski, J.D., 2003. Long-term, high-frequency
current and temperature measurements along central California: insights into
upwelling/relaxation and internal waves on the inner shelf. Continental Shelf
Research 23, 901–918.
Thorpe, S.A., 1977. Turbulence and mixing in a Scottish loch. Philosophical
Transactions of the Royal Society of London, Series A 286, 125–181.
Warren, J.D., Stanton, T.K., Wiebe, P.H., Seim, H.E., 2003. Inference of biological and
physical parameters in an internal wave using multiple-frequency, acousticscattering data. ICES Journal of Marine Science 60, 1033–1046.
Winant, C.D., 1974. Internal surges in coastal waters. Journal of Geophysical
Research 79, 4523–4526.
Woodson, C.B., Barth, J.A., Cheriton, O.M., McManus, M.A., Ryan, J.P., Washburn, L.,
Carden, K.N., Cheng, B.S., Fernandes, J., Garske, L.E., Gouhier, T.C., Haupt, A.J.,
Honey, K.T., Hubbard, M.F., Iles, A., Kara, L., Lynch, M.C., Mahoney, B., Pfaff, M.,
Pinsky, M.L., Robart, M.J., Stewart, J.S., Teck, S.J., True, A., 2011. Observations of
internal wave packets propagating along-shelf in northern Monterey Bay.
Geophysical Research Letters 38, L01605.
Woodson, C.B., McManus, M.A., 2007. Foraging behavior can influence dispersal of
marine organisms. Limnology and Oceanography 56, 2701–2709.
Woodson, C.B., Webster, D.R., Weissburg, M.J., Yen, J., 2005. Response of copepods
to physical gradients associated with structure in the ocean. Limnology and
Oceanography 50, 1552–1564.