N in the
Pacific
N in the
Pacific David Archer1 Jim Aiken2 William Balch3 Dick Barber4 John Dunne5 Pierre Flament6 Wilford Gardner7 Chris Garside3 Catherine Goyet8 Eric Johnson9 David Kirchman10 Michael McPhaden11 Jan Newton5 Edward Peltzer8 Leigh Welling12 Jacques White13 James Yoder14submitted to Deep Sea Research II August 1996
revised January 1997
1Department of the Geophysical Sciences, University of Chicago, Ill 60637 (corresponding author). 2Plymouth Marine Laboratory, Plymouth PL13DH, UK. 3Bigelow Laboratory for Ocean Science, McKown Point, West Boothbay Harbor, ME 04575. 4Duke University Marine Laboratory, Pivers Island, Beaufort, NC 28516. 5School of Oceanography, WB-10, University of Washington, Seattle, WA 98195. 6Department of Oceanography, University of Hawaii, Manoa, 1000 Pope Rd., Honolulu, HI 96822. 7Department of Oceanography, Texas A&M University, College Station, TX 77843. 8Department of Chemistry, Woods Hole Oceanographic Institution, Woods Hole, MA 02543. 9Joint Institute for the Study of the Atmosphere and Oceans, University of Washington, Seattle, WA 98195. 10College of Marine Sciences, University of Delaware, Lewes, DE 19958. 11Pacific Marine Environmental Laboratory, NOAA, 7600 Sand Point Way, NE, Seattle, WA 98115. 12College of Oceanography, Oregon State University, Corvallis, OR 97331. 13People for Puget Sound, 1326 5th Ave., Seattle, WA 98101. 14Graduate School of Oceanography, University of Rhode Island, Narragansett Bay Campus, Narragansett, RI 02882.
N, 140W and its
surrounding environment. The front was a component of a tropical
instability wave generated by the convergence of cold equatorial
waters from the south and warmer equatorial counter current water to
the north. Surface waters on the cold side were undersaturated with
oxygen, which suggests that the water had only been exposed at the sea
surface for a period of a few weeks. Although the atmospheric
exposure time was short, the effects of biological activity could be
detected in enhanced concentrations of dissolved organic carbon
concentration, proving that DOC can be produced quickly in response to
changing environmental conditions. The front itself was dominated by
the accumulation of a "patch" of buoyant diatoms Rhizosolenia
castracanei concentrated in the top centimeters of the warm
surface water north of the front, and elevated chlorophyll
concentrations were observed from the air over a spatial scale of
order 10-20 km northward from the front. The nitrogen budget and
thorium data suggest that a significant fraction of the elevated POC,
and virtually all of the PON, arrived in the patch waters as imported
particles rather than in situ photosynthesis. Photosynthetic uptake
of carbon appears to have occurred in patch waters, but without
corresponding uptake of fixed nitrogen (an uncoupling of the usual
Redfield stoichiometry). Solute chemistry of the patch appears to be
controlled by turbulent mixing, which flushes out patch waters on a
time scale of days (faster than atmospheric ventilation). The
subduction of nutrient-rich equatorial surface water below the front
was detected 100 km north of the front in the signatures of
temperature, salinity and ammonium.
36' North Latitude, and 85 West Longitude, which
placed us about two hundred miles southeast of Cocos Island.When I approached within the possibility of more accurate examination, I saw that the line, which stretched from horizon to horizon, extended in a northeast and southwest direction. On our side, the south, the water showed dark and rough, but much lighter and smoother to the north. When the Arcturus was at last actually astraddle of the rip, I saw it as a narrow line of foam, zigzagging across the placid sea, with spouting white-caps shooting up through the froth that marked the meeting place of the great ocean currents."
The journey of the Arcturus was sponsored by the National Zoological
Society of America and took place in 1925. William Beebe's description of a
convergent front in April at 2 N, 85 W in the
eastern Pacific was first published in 1926 [Beebe, 1926] and yet sounds
astonishingly familiar to participants in the JGOFS Fall Survey expedition of
August-September, 1992. A photograph published in Beebe's account is
indistinguishable from a photograph taken on the JGOFS cruise [Yoder et
al., 1994] (Figure 1).
As in Beebe's time, our initial detection of the front came from the officers
on the bridge, in our case via radar where the front appeared as a distinct
line of high reflection. Our position was 140 W, 2 N,
considerably west of Beebe's report. The radar reflection from the front arose
from waves breaking upon the density stratification of the front. In addition
to the breaking waves the front was marked by a surface "patch" of buoyant
diatoms Rhizosolenia castracanei [Yoder et al., 1994]. The
Rhizosolenia patch generated a striking contrast in sea surface color
across the front. The south (cold) side was dark blue, whereas the north
(warm) side (containing the Rhizosolenia patch) was a yellowish-green
mixture of the colors of hay and split pea soup. The Rhizosolenia patch
was quite shallow; aft of the moving ship, the patch was swept aside by the
wake revealing blue water underneath.
The dynamics of the front were intimately associated with upwelling at the equator and shear between the westward flowing South Equatorial Current and the eastward North Equatorial Counter Current. Shear instability generates a sawtooth-shaped fluctuation in the boundary between the two currents, which propagates westward with a period of 25-30 days [Legeckis, 1977; Kessler and McPhaden, 1995; Flament et al., 1996; Johnson, 1996]. Superimposed upon the instability wave is the divergence-driven upwelling of water at the equator, which drives subduction to the north. The convergent front was located on the westward (leading) edge of the instability wave.
Here we present a summary of numerous investigators' shipboard and
remote observations of the wave and front to reconstruct its impact on
the chemistry and biology of the equatorial surface ocean. Shipboard
data comes from two 140 W Survey
legs of the U.S. JGOFS EqPac expeditions [Murray et al., 1995],
which encountered ENSO conditions during Survey I in February-March of
1992, and cooler conditions, and the wave and front, during Survey II
in August-September of the same year. We also show data from a
PROTEUS mooring at 0 N, 140
W [Kessler and McPhaden, 1995] and from NASA aircraft
[Yoder et al., 1994], contemporary with the shipboard
observations of the front. The reader is directed to a series of
primary papers, mostly contained in the first two Equatorial Pacific
JGOFS Deep-Sea Research special issues, for details and methodology
for the data presented here.
N and heading south toward the mooring at the equator. The
ground-based shipboard and mooring data are combined with the
satellite image of the westward propagating instability wave by
transposing the ground data eastward from the 21 August position at a
rate of 0.5
d-1 (time is indicated at the bottom of the front face of
Figure
2) [Flament et al., 1996; Johnson, 1996]. The
technique uses time-dependent observations at a single line of
longitude as a proxy for zonal structure, based on the assumption that
the instability wave is an unchanging feature propagating at a
constant velocity. Any real E-W slope of the thermocline will be
neglected by the technique, as will time variability of the wave
structure. The satellite image is cut away to reveal high-resolution subsurface temperature data from a towed Undulating Optical Recorder (UOR), which profiles from near the surface to 100 meters depth on a spatial scale of ~3 km while the ship is underway. The sea surface front is revealed to be a wedge of warm water to the north. The velocity structure of the water column was characterized by two accoustic doppler current profilers (ADCPs), one aboard the Thompson (blue arrows) and the other mounted on the mooring [Kessler and McPhaden, 1995] (red arrows). Both shipboard and mooring data are time averaged, and are displaced longitudinally as a function of time. The instability wave is itself propagating to the west, as indicated by the yellow arrow on the left side of the figure. The velocity vectors from shipboard and moored ADCP do not include this component of the wave's velocity; that is to say, they are velocities relative to the earth's surface, rather than relative to the moving front.
North of the front the water flows in a jet roughly parallel to the
front, toward the southwest. On the south side of the front the flow
is northward, subducting under the warmer waters to the north
[Johnson, 1996]. The equatorial undercurrent can be seen as
eastward subsurface flow on the front plane of the figure (the
equator). At the equatorial sea surface, the coldest waters are seen
to the east of the front as the outcropping of the 24 isotherm.
The relationships between the water masses associated with the front
can be seen in a temperature / salinity diagram (Figure
3) (see also Johnson [1996] and Balch and Kilpatrick
[1996]). A T-S trace of a vertical cast on the warm side of the front
begins in warm surface water at relatively low salinity. With
increasing depth the trace ventures toward the T/S signature of
equatorial surface water with higher salinity, and below that, the
salinity trend reverses, decreasing into deep water. This distinctive
salinity reversal "cusp" signature can be seen in subsurface waters to
the north, suggestive of a relationship between cold surface waters at
2 N and
subsurface waters at 3 N. A
similar T - S cusp is visible in Survey I data, but the cusp water
during Survey I never outcropped. In T-S space, the cusp water
appears to be related to equatorial surface water, although neither
cusp water nor waters from the T-S "saddle" beneath the cusp water can
be found at the equator, presumably due to dominance of the equatorial
undercurrent on the provenance of waters in the 100-150 m depth range.
and 2 N. A
sea-surface expression of the 1 - 2 N feature
was detected in the underway sea surface chlorophyll determinations by
Balch and Kilpatrick [1996], but missed by the CTD sampling
program. Shipboard oxygen [Murray et al., 1995] (Figure
4b) and pCO2 data (Figure
4c) [Archer et al., 1996; Goyet and Peltzer,
submitted] reveal the profound and abrupt transition in water
chemistry between the warm waters to the north and the colder southern
waters.
Cold waters south of the front. The cold surface waters south
of the front are denser than equatorial surface water, and are in fact
the densest surface waters observed during either JGOFS Survey
transect. Consistent with the general increase in nutrient
concentration and pCO2 with density in the ocean, this
water predictably had high concentrations of nutrients and high
pCO2 (Figure
5). The exposure of fertile subsurface water to the atmosphere is
presumably the cause of the highest measured rate of primary
production and calculated rate of air-sea gas exchange from the
transects. Above and beyond this observation, we can say from
relationship between oxygen and CO2 at the sea surface (Figure
6) that this water has only been exposed to the atmosphere for a
short time [Archer et al., 1996]. A simple model for
simultaneous gas exchange of oxygen and CO2, beginning with
initial conditions from subsurface waters, constrains the maximum
atmospheric exposure time of these waters to 10 to 20 days.
Apparently the mixed layer is either subject to intense mixing with
subsurface waters, or else the entire mixed layer is flushed through
by ventilation along an outcropping isopycnal surface. Current
velocities from ADCP are consistent with this short atmospheric
exposure time: the average northward velocity of surface waters
between the equator and 2 N was 45 cm
sec-1, which should cover the 220 km distance in 6 days.
Exposure of thermocline waters to the atmosphere was clearly responsible for high values of pCO2 and CO2 evasion rates from the ocean. Beyond this, however, subduction of nutrient-rich waters is a process that has special significance to understanding the dynamics of the carbon cycle in the thermocline. Consider a parcel of water at the sea surface, with zero nutrients and in atmospheric equilibrium in CO2. The parcel is subducted and gains nutrients and CO2 by oxidation of biological particles sinking from above. If this parcel is brought to the surface again and held there until biological activity has depleted the nutrient stock completely, the excess CO2 typical of thermocline water is incorporated into biogenic particles and pumped to depth, resulting in zero net flux to the atmosphere (neglecting any temperature change and its effect on CO2 solubility). If on the other hand the parcel is exposed to the atmosphere and allowed to degas, but subducted before its nutrients are completely utilized, a net flux of CO2 from the thermocline to the atmosphere is allowed to occur. In other words, ventilation and subduction of high-nutrient waters may constitute a "leak" in the biological pump, which the intense dynamics of the front have revealed in our data.
DOM production resulting from surface exposure. A major
uncertainty in our understanding of the carbon cycle in the upper
ocean is the role of dissolved organic matter (DOM), the rate of
production of which is poorly known. DOM concentration variability
was characterized by measuring total organic carbon (TOC) by injecting
unfiltered seawater samples into a high temperature organic carbon
analyser [Peltzer and Hayward, 1996]. Figure
7 shows the relationship between the dissolved plus particulate
organic carbon concentration, TOC, and 
between the equator and 5 N from both
Survey I and II transects. At
between 22.5 and 24 (characteristic of the outcropping water on the
cold side of the front during Survey II, see shading in Figure
7) the TOC concentration is elevated relative to Survey I data.
We surmise that the TOC concentration responded to biological activity
at the sea surface, within the 10-20 day time frame of atmospheric
exposure (Figure
6). In contrast, the pCO2 and nutrient concentrations
when plotted against density are indistinguishable between Survey I
and Survey II, except for waters within the Rhisosolenia patch
[Archer et al., 1996]. One definition of a long-lived chemical
species in the ocean would be one whose production and decay times are
comparable to the circulation time of the ocean, whereas a transient
tracer might by definition have a lifetime which is shorter than the
ocean circulation. Here we show a component of TOC which is clearly
transient by this definition.
Signatures of ventilation in subsurface waters. In addition to
the subsurface signatures found in surface water south of the front,
we detected the signature of surface exposure within the waters
beneath the sea surface north of the front. These include a maximum
in NH4 [Garside and Garside, 1995; McCarthy et
al., 1996] found in "cusp" water (refer to Figure
3) which was present in Survey II but absent in Survey I (Figure
8). This signal might be attributed to local production north of
the front, but rates of primary production and particle export at
3 and
5 N
were similar between Survey I and II. The more plausible explanation
is that the NH4 maximum was generated by subduction and
oxidation of suspended particulate and dissolved organic matter
(i.e. TOC) that was produced at or south of the convergent front. The
mesozooplankton biomass was elevated at depth in the ammonium maximum
at 3 N
relative to Survey I observations [Zhang et al., 1995],
consistent with local subsurface organic carbon regeneration. In
addition, the signature of the radiolarian population assemblage
observed at depth at 3 N was
characteristic of the cold surface water [Welling et al.,
1996].
Mean regional impact of instability waves. The net mean effect of instability waves on the geochemistry of the equatorial Pacific is difficult to judge. The mean rate of equatorial upwelling and poleward subduction is probably determined by forcing from the winds, while the impact of the instability waves would be to modulate the upwelling rate about the wind-determined mean value [Harrison, 1996]. A guess about the effect of the waves on the ventilation of the thermocline would therefore require a comparison of an ocean with intermittant sea surface exposure of extremely high pCO2 waters against an ocean with steady exposures of moderately high pCO2 waters. The net impact of the waves could be either to increase or decrease the ventilation of the thermocline relative to a more sluggish ocean. The effect of the waves on plankton ecology is more definite; an ocean with a high degree of outcropping variability would probably have a more dynamic "bloom selected" ecosystem structure than would a more sluggish ocean. Our geochemical data indicate that the plankton communities observed between the equator and the front must have been only recently exposed to sunlight.
Subsurface, the slope of the deepening isothermal surfaces north of
the front is not well resolved, but we estimate from the UOR data a
slope of approximately 1 meter depth per kilometer distance. The
incoming UOR transect from the north is estimated to end at roughly 11
km north of the front based on the analysis of front movement by
Johnson. The proximity of the outgoing transect to the front is more
difficult to estimate, because the location of the front with time
became less predictable toward the end of our occupation of Station 6
(Eric Johnson, personal communication). Cross-front velocities were
to the north, at the surface south of the front and continuing
subsurface north of the front, at a rate of roughly 20 cm
s-1 averaged over a depth interval of about 100 meters.
From this we can calculate a subduction rate of 20 m3
s-1 per linear meter of the front; if this flow were
characteristic of values along a homogeneous 500 km front, we would
calculate a subduction rate of approximately 10 Sv over a single
wavelength of the tropical instability wave train. This can be
compared with an estimate of a 50 Sv of upwelling and equatorial
divergence over the entire equatorial region, 5 N - 5 S, dateline
to 90 E
[Wyrtki, 1981].
N, and all
data are plotted relative to this. The CTD fluorescence data are
plotted relative to the distance from the front using the cruise track
geometry from Johnson [1996] (Table
1). Fluorescence chlorophyll concentrations are five times higher
than typical values away from the front (corroborated by analysis of
bottle data, not shown). CTD data suggest that the chlorophyll
anomaly extended to the depth of the mixed layer north of the front.
Flow-through seawater line chlorophyll data, while not reaching the
same high values as were observed in the CTD data, show high values
extending to order 10 km to the north of the front. Laser induced
chlorophyll fluorescence data from NASA aircraft [Yoder et al.,
1994], which should reflect the sea surface Rhizosolenia patch,
revealed elevated values in a sharp front located at 2.5 N on August
25th, when most of the cold water CTD deployments were made (the warm
side of the front was mostly sampled on the 26th).
Profiles of chemical and biological properties from across the front
are presented in Figure
11. The profiles are resolved into warm and cold endmembers (from
3
N and 1
N, stations 5 and 7, respectively) and casts from near the
front, divided by the sea surface temperature into warm and cold front
casts (all from station 6 near 2N). Most of
the chemical impact of the front is seen on the warm side; the only
apparent distinction between cold front and cold endmember casts in Figure
11 is that is
somewhat higher at the sea surface close to the front (as noted
previously, the cold front waters were the densest outcrop observed
during the entire JGOFS Survey expeditions). On the warm side, some
of the observed chemical properties from the front casts, notably and
NO3, resemble a hybrid transition between the cold casts
and the warm endmember, reflecting the layered structure of the water
column north of the front. However other properties are anomalous
near the front. Specifically, warm frontal waters had extremely
depressed pCO2 [Archer et al., 1996] and AOU
(denoting oxygen supersaturation), and elevated values of TOC
[Peltzer and Hayward, 1996] and POC relative to the warm
endmember. These geochemical tracers are all indicative of high rates
of primary productivity.
Production within the warm frontal waters. Oxygen in the warm
frontal mixed layer was roughly 45 umol kg-1
supersaturated, while values in the warm endmember sea surface were
near saturation. Using the Redfield respiration ratio, this oxygen
signal translates to a removal of ~35 uM 
CO2
by photosynthesis in a closed system. The analogous calculation can
be done using the patch anomaly in pCO2, which was 65 uatm
lower in the patch than in warm endmember casts. The deficit of
dissolved CO2 required to generate this observed deficit of
pCO2 can be approximated by
where 
is the Revelle buffer factor (assumed for the warm patch waters to be
8 [Archer et al., 1996]), and taking 
pCO2
= 65 uatm, pCO2 = 380 uatm, and CO2
= 1960 umol kg-1 (warm endmember values), we can calculate
that approximately 40 umol kg-1 CO2
must have been removed from the warm front water to the warm
endmember, consistent with the oxygen anomaly in a closed system.
Oxygen supersaturation in surface waters has often been used as a
measure of the rate of biological production, as for example, by
Spitzer and Jenkins [1989]. Assuming a gas exchange piston
velocity of 3 m day-1, the oxygen evasion rate from the
warm frontal water to the atmosphere (neglecting the narrow patch
zone) was approximately 140 mmol O2 m-2
day-1. However, it would be deceptive to make the usual
assumption that the oxygen degassing flux is equivalent to the
biological source flux. Gas exchange would fractionate O2
and CO2 if it were the dominant balance for biological
production (see Archer et al. [1996] and caption to Figure
6). Here we observed no fractionation; both O2 and
CO2 anomalies from warm end member chemistry indicate a
removal of ~35-40 uM CO2.
We propose two potential explanations. The first possibility is that
the time scale of the biological perturbation was shorter than the gas
exchange time. For example perhaps a day / night cycle in
photosynthesis and respiration generated oxygen supersaturation only
during the daytime. However, this hypothesis would predict a greater
day / night cycle in pCO2 than was detected in the
continuous pCO2 data [Goyet and Peltzer, submitted].
The second possibility is that the waters in the warm frontal waters are quickly flushed by turbulent mixing and subduction, a process which would not fractionate between O2 and CO2 (the flushing hypothesis). The gas exchange time scale was of order 5 days, assuming a mixed layer thickness of 16 meters and a gas exchange rate constant of 3 m d-1; the flushing hypothesis would imply that the residence time for warm frontal water must be shorter than this. A physical analysis of the front support the flushing hypothesis [Johnson, 1996]. The dynamics of gravity flows predict mixing at a turbulent hydraulic jump just to the warm side of the front. Also, the residence time of warm frontal water can be estimated from ADCP and CTD data on either side of the front. On the south (cold) side, northward water flow into the front is restricted to a very narrow density range, reflecting the fact that the surface waters are well mixed. On the north side, the mass flux nearly balances the flux south of the front, but the density range of the flowing water has widened, presumably by mixing and subduction of warm waters into the cold flow beneath. Heat balance requires a southward flux of warm water of approximately 1.0 m3 s-1 per meter length of the front. This warm water flux implies a residence time for water in a patch 16 m deep and 8 km wide (the distance from the front of the CTD profile used by Johnson) of about two days.
The implication of the short residence time for warm frontal water coupled with the oxygen and CO2 anomalies is that oxygen production rate within the warm frontal water may have been many times higher than the 140 mmol m-2 d-1 calculated above. Assuming a mixing time of 2 days, and a water slab thickness of 16 meters, a production rate of very roughly 300 mmol m-2 d-1 within the patch would be required to maintain the observed oxygen or CO2 anomalies. However, this estimate is sensitive to the depth distributions of oxygen anomaly and water residence time (flow), neither of which are well resolved in the data.
Particles in the warm frontal water. The concentration of particulate organic carbon (POC) in the bulk mixed layer on the warm side of the front was elevated to approximately 13 - 26 umol kg-1 (two replicates) relative to typical sea surface values away from the front of 2 - 3 umol kg-1 . Thus a considerable extent of the biological perturbation of O2 and CO2 could be accounted for simply by production of organic matter which remained suspended in the patch mixed layer. The problems with this simple scenario are the NO3, PON, and 234Th data. The biological production implied by the CO2 and O2 data corresponds to approximately 5 umol kg-1 of nitrogen (the measured particle C/N within the warm frontal water was 5.64). However, we notice that the warm end member surface contained only ~1.5 umol kg-1 NO3, the dominant form of available dissolved nitrogen, and that in fact values in the patch waters were actually somewhat higher than this (2-3 umol kg-1). Since the required nitrogen to produce the PON is greater than the total dissolved concentration in warm endmember waters, we must conclude there are other ways of importing fixed N to that water parcel. One possibility is nitrogen fixation within the patch, but this would require very high rates in the presence of dissolved NO3, which seems unlikely. Another possibility is that the nitrogen was imported in particulate form, floating out of subducting cold waters from the south, as concluded by [Yoder et al., 1994]. Rhizosolenia sp. mats in the north Pacific have in fact been observed vertically transporting NO3 [Villareal et al., 1993; Villareal et al., 1996], although this was in nutrient depleted conditions.
At the same time, although uptake of NO3 appears to be an insignificant term in the PON budget, the anomalies in dissolved oxygen and CO2 argue that a considerable excess of photosynthesis over respiration must have taken place within the waters of the patch. Our conclusion must be that the Rhizosolenia carry enough metabolic nitrate from their previous residence in the nitrate-rich equatorial waters that exposure to sunlight allows considerable uptake of CO2 and release of oxygen by photosynthesis without further uptake of nitrogen (again consistent with [Villareal et al., 1993; Villareal et al., 1996]). Regardless of the uncertain time-dependent behavior and turbulent mixing regime, the observed anomalies in solute chemistry insist that this uncoupling between carbon and nitrogen must have occurred.
The conclusion that particles were imported to front waters is also
supported by measurements of 234Th within the waters of the
front. 234Th is produced by the decay of
238U, which has an activity of 2.40 dpm kg-1 in
ocean waters. In a closed system steady state, the activity of
234Th would equal the activity of 238U (secular
equilibrium). In surface waters, the activity of 234Th is
typically observed to be lower than this, because some Th is scavanged
by sinking particles. Surface waters away from the front between
1 -
3 N
contained 1.4 - 2.15 dpm kg-1 of 234Th total
(dissolved plus particulate), while waters from a CTD cast
approximately 3 km on the warm side of the front contained a total of
3.3 dpm kg-1: actually higher than the secular equilibrium
value. This is strong evidence for the import of particles bearing
234Th to the front water. Also, dissolved
234Th, a signature of redissolving particles, was not
anomalously high in the front. This is consistent with the conclusion
based on dissolved O2 and CO2, above, that a
water parcel within the front had only a short residence time (< 3
days) before subduction.
Using the assumption that a major fraction of the chlorophyll in the patch was imported to the patch in particles floating out of the convergent flow, we would like to estimate the accumulation timescale of the patch based on the budget of chlorophyll. The inventory of excess chlorophyll in the patch was approximately 125 g Chl per meter length of the front (taking the dimensions of the high chlorophyll patch to be 10 meters thick by 5 km wide, and the excess chlorophyll concentration to be 2.5 mg m
N
during August, 1992. The front was the leading edge of a tropical
instability wave and separated cold equatorial water to the south from
the warmer equatorial counter current water to the north. Signatures
of subduction can be seen below the warm water north of the front, and
a patch of buoyant phytoplankton accumulated in the convergence zone.
Disequilibrium of surface waters with atmospheric oxygen concentration was seen on both sides of the front: undersaturation in cold waters to the south and supersaturation in patch waters adjacent to the front to the north. The undersaturated waters to the south were apparently the product of recent exposure of subsurface waters to the atmosphere. Based on the depth of mixing and the rate of gas exchange, we estimate that this water had only been exposed for 10-20 days. These waters had elevated concentrations of dissolved organic carbon relative to waters of the same density but below the surface during Survey I, indicating that DOC production occurred quickly in response to exposure to sunlight. The T / S signature of surface waters south of the front could be traced subsurface 100 km north of the front; these waters contained high concentrations of NH4 relative to Survey I results, presumably the product of subduction and oxidation of suspended organic matter.
The warm waters at the line of convergence were filled with buoyant Rhizosolenia sp. phytoplankton. The dynamics of fluid and particles within the Rhizosolenia patch are constrained by observations and budget calculations for the biological tracers O2, CO2, chlorophyll, and fixed nitrogen. A nitrate budget and thorium data indicate that much of the standing stock of particulates must have been imported as particles rather than growing in place. However, the concentrations of O2 and CO2 were both perturbed from source values, indicating that primary production had occurred within patch waters, but apparently without corresponding NO3 uptake (an uncoupling of normal Redfield stoichiometry). The oxygen and CO2 anomalies are of comparable (Redfield) magnitudes, which probably indicates that the biological source for O2 by photosynthesis was balanced by turbulent mixing and subduction of front waters (which would not fractionate the two gases) rather than gas exchange (which would fractionate). This conclusion is supported by dissolved 234Th. The implication of this is that the residence time for patch water was only a few days.
Aiken, J., and I. Bellan, Optical oceanography: an assessment of a towed method, in Light and Life in the Sea, edited by P. J. Herring, A. K. Campbell, M. Whitfield and L. Maddock, pp. 39-57, Cambridge University Press, Cambridge, 1990.
Archer, D. E., E. Peltzer, and D. Kirchman, A time scale for DOC production in equatorial Pacific surface waters, Global Biogeochem. Cycles in press.
Archer, D. E., T. Takahashi, S. Sutherland, J. Goddard, D. Chipman, K. Rodgers, and H. Ogura, Daily, seasonal, and interannual variability of sea surface carbon and nutrient concentration in the Equatorial Pacific Ocean, Deep-Sea Res. II 43, 1155-1180, 1996.
Balch, W. M., and K. Kilpatrick, Calcification rates in the equatorial Pacific along 140o W, Deep-Sea Res. II 43, 971-994, 1996.
Barber, R. T., M. P. Sanderson, S. T. Lindley, F. Chai, J. Newton, C. C. Trees, D. G. Foley, and F. P. Chavez, Primary production in the equatorial Pacific along 140o W, Deep Sea Research II 43, 933-970, 1996.
Beebe, W., The Arcturus Adventure. The Knickerbocher Press, New York, 1926.
Flament, P., S. Kennnan, R. Knox, P. Niiler, and R. Bernstein, The three-dimensional structure of a tropical instability wave, Nature 383, 610-613, 1996.
Garside, C., and J. C. Garside, Euphotic zone nutrient algorithms for the NABE and Eqpac study sites, Deep-Sea Res. II 42, 335-348, 1995.
Goyet, C., and E. T. Peltzer, Variation of CO2 partial pressure in surface seawater in the equatorial Pacific Ocean., Deep-Sea Res. II submitted.
Harrison, D. E., Vertical velocity in the central tropical Pacific -- a circulation model prespective, Deep-Sea Res. II 43, 687-706, 1996.
Johnson, E. S., A convergent instability wave front in the central tropical Pacific, Deep-Sea Res. II 43, 753-778, 1996.
Kessler, W. S., and M. J. McPhaden, The 1991-1993 El Nino in the central Pacific, Deep-Sea Res. II 42, 295-334, 1995.
Legeckis, R., Long waves in the eastern equatorial Pacific: A view from a geostationary satellite, Science 197, 1177-1181, 1977.
McCarthy, J. J., C. Garside, J. Nevins, and R. T. Barber, New production along 140 W in the equatorial Pacific during and following the 1992 El Nino event, Deep-Sea Res. II 43, 1065-1094, 1996.
Murray, J. W., E. Johnson, and C. Garside, A U.S. JGOFS process study in the equatorial Pacific (EQPAC): Introduction, Deep-sea Res. II 42, 275-294, 1995.
Peltzer, E. T., and N. A. Hayward, Spatial and temporal variability of total organic carbon along 140 W in the equatorial Pacific ocean in 1992, Deep-Sea Res. II 43, 1155-1180, 1996.
Spitzer, W. S., and W. J. Jenkins, Rates of vertical mixing, gas exchange, and new production: estimates from seasonal gas cycles in the upper ocean near Bermuda, J. Mar. Res. 47, 169-196, 1989.
Villareal, T. A., M. A. Altabet, and K. Culver-Rymsza, Nitrogen transport by vertically migrating diatom mats in the North Pacific Ocean, Nature 363, 709-712, 1993.
Villareal, T. A., S. Woods, J. K. Moore, and K. Culver-Rymsza, Vertical migration of Rhizosolenia mats and their significance to NO3 fluxes in the central North Pacific gyre, J. Plankton Res. 18, 1103-1121, 1996.
Welling, L. A., N. G. Pisias, E. S. Johnson, and J. R. White, Distribution of polycystine radiolaria and their relation to the physical environment during the 1992 El Nino and following cold event, Deep-Sea Res. II 43, 1413-1434, 1996.
Wyrtki, K., An estimate of the equatorial upwelling in the Pacific, J. Phys. Oceanography 11, 1205-1214, 1981.
Yoder, J. A., S. G. Ackleson, R. T. Barber, and P. Flament, A line in the sea, Nature 371, 689-693, 1994.
Zhang, X., H. D. Dam, J. R. White, and M. R. Roman, Latitudinal variations in mesozooplankton grazing and metabolism in the central tropical Pacific during the EqPac study, Deep Sea Res. II 42, 695-714, 1995.