OS31A-1225
Long-term Variation of Oceanic Carbon Dioxide and Possible Acidification in the Indian Sector of the Southern Ocean
Long-term monitoring of the partial pressure of CO2 in the surface seawater (pCO2) on-board Japanese icebreaker SHIRASE has been carried out in the Indian sector of the Southern Ocean since 1987 as a part of Japanese Antarctic Research Expedition (JARE). Meridional distributions of pCO2 along 110°E in early December and along 150°E in late March clearly show steep changes at such fronts as Subtropical Front, Subantarctic Front, and Polar Front. pCO2 of each zone divided by the fronts can be distinguished from the others by the difference of averaged pCO2 in the zone. Although pCO2 of each zone shows interannual variation, secular trend is detectable. For example, estimated increasing rate of pCO2 in the permanent open ocean zone between polar front (around 53°S) and northern edge of winter ice cover (63°S) is about 1.5 μatm/y which is almost same as the increasing rate of atmospheric CO2 concentration. Oceanic acidification corresponding to pCO2 increase is one of the most direct effects of increasing atmospheric carbon dioxide. Preliminary analysis of pH which has been observed on board SHIARASE and her predecessor FUJI shows gradual decrease from 1980 to 2005.
OS31A-1226
Understanding the interannual variability of the oceanic carbon cycle: Results from the coupled biogeochemical-physical global model PISCES-NEMO
The last advances on the development of coupled biogeochemical-physical model has brought new insights on the oceanic carbon cycle and particularly on the sources of interannual and decadal variability. In the frame of the CARBOOCEAN and MERCATOR-Vert projects, we seek to analyze the ocean carbon cycle through the biogeochemical-ecosystem model PISCES coupled with the OGCM NEMO in global configurations developed in our project. Here we use the PISCES model at different resolutions (2, 1/2, 1/4) to investigate the interannual changes of the carbon sink in the ocean and to study the impact of resolution on simulations. Then we discuss the model simulations of the last decade using the 2001-2005 integration MERCATOR reanalysis, the french GODAE operational oceanography system, coupled offline to the PISCES model in a 1/4 global configuration. We compare here the results obtained using dynamical fields with and without assimilation of physical data (SST,...) and we try to identify the processes that are responsible for the interannual variability in the oceanic carbon cycle as captured by the operational system.
OS31A-1227
Decadal Changes in DIC Along P16 in the Pacific Ocean
As a result of CLIVAR CO2 Repeat Hydrographic program, P16S was reoccupied in 2005, and P16N in 2006. The P16 is a WOCE cruise line along longitude 150°W in the Pacific Ocean. The section selected for this analysis in the North Pacific includes data collected between Hawaii and Kodiak Island. In the South Pacific, only stations along 150°W occupied during 1991 as P16S and during 1992 as P16A are selected. These are stations that were reoccupied in 2005 as Repeat Hydro P16S. The DIC values are normalized to salinity of 35 after correction for AOU. Comparisons of these DIC values are made along isopycnal surfaces. The DIC increases vary in each isopycnal horizon, with 2.4 µmol/kg along 27.2 isopycnal, and 18.0 µmol/kg along 26.4 isopycnal. With the estimated thickness of isopycnal horizons between 15°N and 45°N, we obtained a mean DIC inventory change of 0.58 µmol/kg/yr in the North Pacifc between 1991 and 2006 (or 0.50 mol/m2/yr). For the South Pacific, DIC increase is apparently faster than that in the North Pacific. We have seen significant DIC increases along each isopycnal horizons, on the order of 20 µmol/kg for isopycnals 26.0 to 27.2. Based on the estimated thickness of isopycnal layers between 15°S and 40°S, the mean DIC inventory change is estimated to be 1.61 µmol/kg/yr from 1991 to 2005, which is equivalent to 1.57 mol/m2/yr.
OS31A-1228
Development Of Small Drifting Buoy System With Sea Surface pCO2 Sensor
Many observations to clarify the fate of CO2 in the atmosphere, related with long term climate change, have been carried out in the world. However, the sea surface pCO2 observations on volunteer observation ships and research vessels concentrated in the North Atlantic and North Pacific. To assess the spatial and temporal variations of surface pCO2 in the global ocean, new automated pCO2 sensor which can be used in platform systems such as buoys or moorings is strongly desired. We have been developing the small drifting buoy system (diameter 250-340 mm, length 470 mm, weight 15 kg) for pCO2 measurement using spectro-photometric technique. The pCO2 is calculated from pH indicator solution equilibrated with seawater through a gas permeable membrane. In our system, an amorphous fluoropolymer tubing form (Teflon AF 2400) with high gas permeability was used as a gas permeable membrane. The measurement side of buoy system consisted mainly of a LED light source, optical fibers, a CCD detector, a micro pump, and a downsized PC. The measured data were transmitted to the shore-based laboratory by satellite communication (Argos system). In the laboratory experiment, we obtained a high response time (less than 2 minutes) and a precision within 3 µatm. Our first deployment of drifting buoy system was made in the east Labrador Sea in May 2008, with the support of the Bedford Institute of Oceanography. The buoy system is measuring sea-surface pCO2 four times a day (also collecting temperature and salinity data) and every six days intervals, because of limited battery capacity. The planned lifetime of buoy system, is about 1 year.
OS31A-1229
Interannual Variability of Southern Ocean Diatom Production in a Model With Flexible Si:N:C:Chl Ratios
The requirements of phytoplankton for nutrients strongly couple the fluxes of different biologically important elements, such as carbon, nitrogen, silicon and iron. In the Southern Ocean this coupling manifests itself in iron limitation controlling the primary production, but also influencing the Si:N-ratio of nutrient drawdown and biomass export. We investigate the interaction between nutrient inputs, biomass composition and the biological pump in the Southern Ocean with the help of a global biogeochemical model that allows for physiological variations of the Si:N:C:Chl ratio in diatom biomass. Our model results agree with observations of enhanced Si:N drawdown ratios around Antarctica and show a belt of strong vertical flux of biogenic Si over the location of the 'opal belt' in the sediment. In the model, the export of BSi around Antarctica is increased through high Si:C ratios in phytoplankton caused by iron limitation. Integration of the model forced by meteorological reanalysis fields leads to interannual variability of sea ice extent, nutrient upwelling and sea surface temperature. We investigate the consequences of this variability for the biological pump in different regions of the Southern Ocean, comparing the model with JGOFS data, and analyse the role of stoichiometric variations.
OS31A-1230
Effect of ocean chemistry changes on particle dynamics and sinking fluxes in the ocean
In an effort to understand the impact of anthropogenic changes in surface ocean chemistry on carbon and nutrient transport to the deep ocean we have developed an Eulerian model of particles sinking to the deep ocean. Increasing atmospheric CO2 affects the production of organic carbon, CaCO3 (Riebesell et al., 2002), which serves as ballast and transparent exopolymer particles (TEP) (Engel, 2002), which increase stickiness and aids aggregation (Alldredge et al., 1993). Our model tracks a statistical sample of the particles throughout the water column, with aggregation and disaggregation occurring stochastically according to probabilities calculated theoretically. We show new measurements of the particle size spectrum from the East coast shelf and slope, which finds a similar power law distribution and slope for surface and subsurface, on and off shore samples. Reproducing this observation requires active disaggregation in subsurface waters, possibly associated with decomposition of TEP. The observed profiles of CaCO3 sinking fluxes call for dissolution in the top kilometer or so of the water column, which we model to take place in the guts of zooplankton. Organic carbon fluxes seem to require a slowing degradation rate with depth, which inspired us to implement multiple age groups with slowing degradation rates, as also inferred for sediment organic geochemistry. We will explore the nature of the flux under varying initial conditions and the potential of feedback mechanisms associated with changing surface conditions.
OS31A-1231
Temporal variability of carbon fluxes in the subtropical North Atlantic at 24.5° N
The Atlantic meriodinal overturning circulation (AMOC) carries warm upper waters north where they cool and sink before returning as cold deep water. The associated ocean-atmosphere heat flux is responsible for northwest Europe's mild climate. A transatlantic hydrographic section including carbon measurements has been occupied at 24.5° N in 1992, 1998 and 2004, allowing us to examine decadal changes in the circulation and fluxes of heat, salt and carbon. In 1998 and 2004 the inferred AMOC is smaller than in 1992, compensated by an increase in southward thermocline circulation shallower than 1000 m. The northward transport of inorganic carbon/anthropogenic carbon in the Gulf Stream is increasing with time and is compensated by an increasing southward flux in the interior. The net flux of inorganic carbon is southward, driven by the large atmosphere to ocean flux of carbon to the north, while net flux of anthropogenic carbon is northward a result higher surface concentrations and lower deepwater concentrations. Most of the carbon flux variability in different years is determined by circulation variability. Although both inorganic and anthropogenic carbon concentrations are increasing with time so that the northward flux in the Gulf Stream is increasing as are the southward return fluxes in the Deep Western Boundary Current or thermocline the overall net carbon fluxes through 24.5° N appear constant. The mean net flux of the 1992, 1998 and 2004 inorganic carbon is -0.89 ± 0.17 Pg C/yr while the net anthropogenic carbon flux is 0.27 ± 0.04 Pg C/yr (both relative to the Bering Strait). A recent study of ten different Ocean General Circulation Models (OGCMs) suggest they underestimate the southward inorganic carbon fluxes by a factor of two in the subtropical North Atlantic region. Intra-annual changes of carbon fluxes are studied using RAPID mooring array and available carbon data in North Atlantic.
OS31A-1232
Long-Term Observations of Ocean Biogeochemistry with Nitrate and Oxygen Sensors in Apex Profiling Floats
Long-term, autonomous observations of ocean biogeochemical cycles are now feasible with chemical
sensors in profiling floats. These sensors will enable decadal-scale observations of trends in global ocean
biogeochemical cycles. Here, we focus on measurements on nitrate and dissolved oxygen. The ISUS (In Situ
Ultraviolet Spectrophotometer) optical nitrate sensor has been adapted to operate in a Webb Research,
Apex profiling float. The Apex float is of the type used in the Argo array and is designed for multi-year,
expendable deployments in the ocean. Floats park at 1000 m depth and make 60 nitrate and oxygen
measurements at depth intervals ranging from 50 m below 400 m to 5 m in the upper 100 m as they profile to
the surface. All data are transmitted to shore using the Iridium telemetry system and they are available on
the Internet in near-real time. Floats equipped with ISUS and an Aanderaa oxygen sensor are capable of
making 280 vertical profiles from 1000 m. At a 5 day cycle time, the floats should have nearly a four year
endurance. Three floats have now been deployed at the Hawaii Ocean Time series station (HOT), Ocean
Station Papa (OSP) in the Gulf of Alaska and at 50 South, 30 East in the Southern Ocean. Two additional
floats are designated for deployment at the Bermuda Atlantic Time Series station (BATS) and in the Drake
Passage. The HOT float has made 56 profiles over 260 days and should continue operating for 3 more
years. Nitrate concentrations are in excellent agreement with the long-term mean observed at HOT. No
significant long-term drift in sensor response has occurred. A variety of features have been observed in the
HOT nitrate data that are linked to contemporaneous changes in oxygen production and mesoscale
dynamics. The impacts of these features will be briefly described. The Southern Ocean float has operated
for 200 days and is now observing reinjection of nitrate into surface waters as winter mixing occurs(surface
nitrate > 24 micromolar). We expect that the OSP and Southern Ocean floats will provide a quantitative
measurement of the timing and magnitude of the spring bloom via the drawdown of surface nitrate. We are
funded through NSF and NOPP to continue float deployments at HOT, BATS, OSP and the Southern Ocean
for the next 3 years and to refine the sensor so it can be offered as a commercial option for all float users.
New sensors in development for float deployments include a stable ISFET pH sensor.
http://www.mbari.org/chemsensor/floatviz.htm
OS31A-1233
Decadal Trend of Dissolved Inorganic Carbon in the Subarctic Western North Pacific Ocean
The dissolved inorganic carbon (DIC) and related chemical species have been measured from 1992 to 2007 at Stations KNOT (44°N, 155°E) and K2 (47°N, 160°E) in the subarctic western North Pacific. DIC corrected for the contribution of the biological activity significantly increased at rates of 0.8-1.6 μmol kg-1 y-1 in the temperature minimum (~100m), which is the remnant of the mixed layer from the previous winter, and intermediate waters (120-380m). This rate decreased with increasing density. To estimate the CO2 uptake rate in this region, we calculated the water column inventory of corrected DIC increase. The water column inventory of CO2 increase was estimated to be 0.42 ± 0.09 mol m-2 y-1, which is almost the same as that previously reported in the subarctic western North Pacific (0.66 ± 0.22 mol m-2 y-1; 1973-1993 [Ono et al., 2000]) and in the global ocean (0.51 ± 0.09 mol m-2 y-1; 1990s [Bindoff et al., 2007]). The corrected DIC increase in the temperature minimum and upper intermediate waters (100-200m, 1.3-1.5 μmol kg-1 y-1) was controlled not only by that expected from oceanic equilibration with increasing anthropogenic CO2 in the atmosphere (0.7 μmol kg-1 y-1), but by the decrease in the difference between atmospheric and oceanic CO2 in the winter mixed layer (0.6-0.8 μmol kg-1 y-1). This decrease, in turn, is caused by the accumulation of CO2 due to the reduction of CO2 emission in winter. If the atmospheric and oceanic CO2 in winter continue to increase at the same rates as during 1992-2007, the calculated CO2 in the temperature minimum will reach that of the atmosphere by 2030- 2040. This would result in the western North Pacific acting as a sink for atmospheric CO2 throughout the year and could accelerate the decrease of pH in surface waters.
OS31A-1234 [WITHDRAWN]
Multidecadal CO2 uptake variability of the North Atlantic
The multi-decadal variability of air-sea CO2 fluxes in the North Atlantic under pre-industrial conditions is simulated using a coupled biogeochemical/circulation model driven by long term surface forcing surface fluxes reconstructed from the leading modes of sea level pressure observations from 1850 to 2000. The magnitude of the simulated multi-decadal CO2 uptake changes is on the order of 0.02 Pg C/yr and amounts to 10 to 15% of the estimated annual anthropogenic CO2 uptake of the North Atlantic. In contrast to the interannual variability, for the multi-decadal CO2 fluctuations heat fluxes are of great importance and they even dominate the effect of wind stress. Another difference compared to interannual time scales is the dominance of the North Atlantic Oscillation (NAO) in driving the variability of the air-sea CO2 fluxes. The multi-decadal CO2 fluctuations are driven by all possible factors: variations in Temperature, Salinity and DIC-content at the sea surface. However, the influence of the different factors compensates to a large extent.
OS31A-1235
Observations of Acidification in the Weddell Sea on a Decadal Time Scale
The amount of anthropogenic CO2 (Cant) that entered the Weddell Sea between 1992 and 2008 is assessed using the extended multiple linear regression method (eMLR). Two multiple linear regressions for total dissolved inorganic carbon (CT) were conducted independently for two data sets from 1992 and 2008. Subtracting these two relations leads to an estimate of anthropogenic carbon (ΔCT) accumulated in the considered time span assuming that the underlying natural correlations between the input parameters and CT do not change with time. In the Warm Deep Water (WDW) and in the Weddell Sea Bottom Water (WSBW) ΔCT values are insignificant, whereas values as high as 8~μmol kg-1 are observed at the shelf. ΔCT concentrations in the surface layer vary with latitude between 2 and 11~μmol kg-1. Weak intrusion of anthropogenic CO2 into Weddell Sea Deep Water (WSDW) was demonstrated, ΔCT yields 1.5 - 2~μmol kg-1. That more Cant is found in the WSDW than in the WSBW is surprising, but can be explained by the more intense ventilation of the WSDW from east of the Weddell Gyre. The invasion of Cant provokes a shift in the equilibria of the carbonate system, resulting in acidification and reduction of CO32-. The mean decrease of pH in the upper 200~m layer is 0.016. Further effects are decrease of the calcite and aragonite saturation states. Our results indicate a slower decrease of aragonite of surface waters in the Weddell Sea than recent model-based estimates.
OS31A-1236
Role of Eddies in Controlling the Uptake of Anthropogenic Ocean in the Southern Ocean and Implications for Recent Claims of a Decline in the Ocean Carbon Sink
There is increasing recognition of the importance of the Southern Ocean in the global carbon cycle. Indeed, a recent reconstruction shows that over 40% of the anthropogenic CO2 in the ocean in 2007 entered south of 50 deg S. However, recent work by Le Quéré et al. (Science, 2007) and Lovenduski et al (GBC, 2007) suggests that CO2 uptake by the Southern Ocean has declined over the last few decades due to an increase in the westerlies. They attribute the decrease to an increase in the strength of the Southern Ocean meridional overturning circulation (MOC) and an associated increase in the outgassing of "natural" CO2. If true, this is cause for considerable concern since most IPCC models predict a continued strengthening of the zonal winds in the Southern Ocean, implying a further weakening of the ocean CO2 sink, with important implications for future climate. Central to the above claims is the assumption that the MOC responds in this simple manner to the winds. However, there are theoretical reasons to suspect otherwise. Here, we use a combination of theory and numerical simulation to better understand and quantify the dynamical response of the Southern Ocean MOC to changes in the zonal wind. Our results show that this response is extremely sensitive to the treatment of baroclinic eddies in ocean models. Specifically, the use of a "Gent- McWilliams-type" parameterization with a constant eddy-transfer coefficient, as is conventional in the coarse resolution ocean models of the type used by the above studies, significantly overestimates the sensitivity-- and perhaps even its sign--of the overturning to changes in the strength of the westerlies. In contrast, the use of a more realistic eddy-parameterization with a spatially variable eddy-transfer coefficient renders the strength of the overturning -- and hence the anthropogenic CO2 uptake -- largely insensitive to changes in the zonal wind. These results, which are supported by idealized eddy-resolving calculations, contradict recent claims of a decline in the ocean carbon sink.
OS31A-1237
Response of plankton ecology and the carbon cycle to climate change over the 21st century
Here we analyze the impact of climate change on ocean plankton ecology and the carbon cycle over the 21st century using a multi-decadal (1880-2100) experiment conducted with the latest version of the Community Climate System Model (CCSM-3.1) coupled ocean-atmosphere-land-ice model. The oceanic ecosystem model component includes three classes of phytoplankton (diatoms, pico/nano plankton, diazotrophs) and one class of zooplankton which grazes differentially on the phytoplankton groups. The competition between phytoplankton groups is altered by climate-induced changes in nutrients, light and zooplankton. Here we connect the resulting changes in the ecosystem structure to changes in the air-sea CO2 fluxes and the global ocean sink. Long-term trends due to anthropogenic changes are compared to the natural variability of the system. Increasing stratification in the northern hemisphere oceans decreases the nutrient supply to the ocean surface and decreases the relative and absolute diatom abundance. The northern hemisphere shift from diatoms to small phytoplankton results in decreases in total primary production, export production and export ratio, and a shift to a more efficiently recycled, lower biomass euphotic layer. By contrast, an increase in Southern Ocean westerlies acts against increasing temperature and freshwater flux to destratify the water-column. Additionally, the wind-driven poleward shift in the Southern Ocean subpolar-subtropical front results in a southward shift and increase in the largest oceanic diatom bloom. In the Southern Ocean diatoms are favored over small phytoplankton on average, acting to increase total chlorophyll, primary production and export production. The impact of these ecological shifts on the global oceanic carbon sink is complex, with northern and southern hemisphere effects partially compensating each other. In the net, total chlorophyll, primary and export production decrease, but less than previous modeling studies have suggested, indicating a small positive feedback on atmospheric pCO2.
OS31A-1238
Multi-Decadal Variations in Calcareous Holozooplankton in the California Current System: Thecosome Pteropods and Foraminifera from CalCOFI
We examine long-term (58 year) variability of two major taxa of calcareous holozooplankton (thecosome pteropods and planktonic foraminifera) in light of recent interest in the impingement of waters undersaturated with respect to aragonite onto shelf depths in the California Current. We utilize the extraordinary CalCOFI zooplankton record from two regions: the Southern California sector (lines 80-93), which is the current location of the California Current Ecosystem-LTER site, and the Central California sector (lines 60-70). We address interannual variability in springtime abundances of net-collected zooplankton sampled in the epipelagic layer from 1951 through 2008. Although the net mesh excluded most of the foram population, the mesh size was held constant through time and is suitable for the largest foram individuals. Sampling was quantitative for the thecosome pteropods, which were typically identified to genus or species from 4 major families. Despite significant long-term trends in density stratification, chlorophyll concentration, and other characteristics of the water column, as of this writing the calcareous holozooplankton do not yet show evidence of significant declines in abundance. Sustained in situ measurement programs are vital to the detection of ecosystem responses, including thresholds and other nonlinear dynamics
OS31A-1239
Carbon Biomass Budgets for the 2008 North Atlantic Spring Bloom Microplankton
The carbon flux dynamics of the North Atlantic spring bloom in 2008 was studied using robotic Lagrangian floats and Seagliders between early April and late June. The instruments on these platforms were ground truthed during four cruises. One goal of the project was to determine how well optical measurements such as backscattering and beam attenuation could be used as proxies for organic carbon biomass in planktonic systems. Critical to this question is whether, and by how much, the proportion of plankton-to-total organic carbon changes over the course of the bloom. Samples were taken in the vicinity of the floats and gliders for plankton biomass determination using flow cytometry and imaging-in-flow technology. In early May the surface plankton community was dominated by chains of the diatom, Chaetoceros. As the bloom progressed the community shifted to a mix of large mixotrophic ciliates (cf. Laboea), dinoflagellates (Ceratium) and different genera of chain diatoms. The Synechococcus and phototrophic nanoplankton abundances increased over this succession as well. This surface succession coincided with a flux of Chaetoceros resting stages to subsurface depths. We estimate the carbon biomass of phototrophic and heterotrophic pico-, nano- and microplankton groups using cell abundances, forward light scatter calibrated to size, and established cell carbon density (carbon mass per unit biovolume) conversions to create a carbon budget during different stages of community succession. The difference between total particulate organic carbon, measured chemically, and the total plankton biomass is inferred to be detrital carbon. Our budget is compared to previously determined carbon budgets for North Atlantic spring bloom communities.
OS31A-1240
The 2008 North Atlantic Spring Bloom Experiment: Radiometric Measurements and Inversions
As part of the 2008 North Atlantic Spring Bloom Experiment (NAB08), hyperspectral radiometric and bio- optical measurements were taken from two Lagrangian Floats every 40-200 s. Measured quantities included downwelling irradiance Ed(λ,z) and upwelling radiance Lu(λ,z) from 320-950 nm as well as oxygen, chlorophyll fluorescence and two inherent optical properties (IOPs), optical backscattering bb(470 nm) and beam attenuation c(662 nm). In addition to Lagrangian drift mode sampling of mixed layers from 5-20 m during more stratified conditions and 20-150 m during pre-bloom conditions and storms, daily high resolution profiles with radiometric and bio-optical samples approx. every 2 m to 250 m were also performed within an hour of solar noon. Calibration profiles of hyperspectral radiometric and bio- optical properties (including attenuation a(λ,z) and c(λ,z) at nine wavelengths) were performed within sight of the floats during a 21-day R/V Knorr cruise. Combined with measurements of oxygen and chlorophyll fluorescence, we have a detailed bio-optical record of both the mixed layer and the entire euphotic zone during the 51 days of the floats' deployment spanning pre-bloom, chain diatom- dominated bloom, and diverse diatom / dinoflagellate / ciliate "post-peak bloom" conditions. We are developing computational inversions of the hyperspectral radiometric measurements, using IOPs as constraints and other bio-optical measurements as checks, to directly diagnose phytoplankton absorption aφ(λ,z) and from this the evolution of the phytoplankton community during the bloom. The hyperspectral resolution allows inquiry into spectral control of the bloom. The high resolution spatio-temporal sampling allows multiple realizations of data that increase the quality of inversions, while the mixed layer excursions allow us to compare inversion of direct radiometric measurements versus inversion of derived apparent optical properties of diffuse attenuation Kd(λ,z) = (d ln Ed(λ,z)/dz) and radiance reflectance RL(λ,z) = Lu(λ,z)/ Ed(λ,z). Our hope is that this technique could provide detailed measurements of phytoplankton community evolution and productivity from autonomous platforms.
OS31A-1241
The 2008 North Atlantic Spring Bloom Experiment II: Autonomous Platforms and Mixed Layer Evolution
The 2008 North Atlantic Spring Bloom Experiment (NAB08) employed a system of drifting floats, mobile
gliders and ship-based measurements to resolve patch-scale physical and biological variability over the 3-
month course of an entire bloom. Although both autonomous and ship-based elements were essential to
achieving NAB08 goals, the autonomous system provided a novel perspective by employing long-range
gliders to repeatedly survey the volume surrounding a drifting Lagrangian float, thus characterizing patch-
scale bloom evolution. Integration of physical and biogeochemical sensors (temperature, conductivity,
dissolved oxygen, chlorophyll and CDOM fluorescence, light transmission, optical backscatter, spectral light,
and nitrate) and development of in situ calibration techniques were required to support this new autonomous
approach.
Energetic, small-scale eddy activity at the experiment site (southeast of Iceland, near the Joint Global Ocean
Flux Study and Marine Light Mixed Layer sites) produced a swift, heterogeneous velocity field that challenged
the gliders" operational abilities and drove refinements to the piloting techniques used
to maintain float-following surveys. Although intentionally deployed outside of energetic eddies, floats and
gliders were rapidly entrained into these features. Floats circulated within eddies near the start and end of
the experiment, drifting generally northwest, across the basin, in-between. An eddy sampled late in the
deployment provided particularly interesting signatures, with elevated biological signals manifest consistently
in one quadrant. As measurements were collected in a parcel-following Lagrangian frame, this suggests
energetic small-scale exchange process (such as vertical or lateral mixing) paired with fast-acting biological
processes capable of modifying the newly entrained water as it navigates its path around the eddy. Despite
this energetic kilometer-scale heterogeneity, broadly distributed platforms appeared to experience similar
broad, long-timescale trends.
Initial mixed layer depths exceeded 200 m, with gradual shoaling punctuated by periods of rapid, storm-driven
deepening. In mid-April, a period of calm weather, rapid restratification and exponentially growing
chlorophyll fluorescence marks the bloom's start. Although one-dimensional
processes (e.g. diapycnal mixing and solar warming) clearly play important roles in producing the spring
bloom, the rate and vertical extent of upper ocean restratification indicate that lateral mixing, perhaps wind-
or eddy-driven exchange or the slumping of lateral density contrasts, play a more important role in
restratifying the upper ocean. These important trigger events present a severe observational challenge as
they take place at small (kilometers) spatial scales, are fully three-dimensional and episodic in time. The
NAB08 efforts demonstrate how mobile, autonomous platforms can be exploited to resolve these events and
their impact over the course of an entire bloom cycle.
http://bloom.apl.washington.edu
OS31A-1242
Signal or Noise? Spikes in Subsurface Optical Measurements During the North Atlantic Bloom
Optical measurements were taken in the North Atlantic southwest of Iceland before and after the onset of the spring phytoplankton bloom by multiple sensors on multiple platforms, including ships, Seagliders and Lagrangian floats. The observation period spanned a total of three months as part of the NAB08 project. Measurements of chlorophyll fluorescence and optical backscatter, proxies for phytoplankton biomass and organic particle concentration, both showed a similar pattern in subsurface waters. After the onset of the bloom we observed a low baseline in these measurements that was punctuated by spikes in depths below the mixed layer. These spikes were present through to our deepest observations at 900m. We interpret these spikes as sinking phytoplankton and other organic particles. Several different methods were used in identifying and quantifying the magnitude of these optical spikes. We compared spikes from different instruments on the same platform as well as spikes from different platforms at varying distances in order to produce initial estimates of the density, distribution and chlorophyll content of the particles responsible for these spikes. The depth distribution of spikes and the relative magnitude of chlorophyll fluorescence vs. backscatter systematically changed over time.
OS31A-1243
Mixed layer restratificaton by submesoscale processes and the initiation of the North Atlantic bloom
Rapid restratification of the deep mixed layers in the spring are crucial for the initiation of the North Atlantic bloom. Air-sea fluxes of heat are too weak to account for the restratification of the mixed layer on the time scales at which the bloom propagates. Our modeling results suggest that when surface cooling and strong westerly winds weaken, the onset of baroclinic instability within the mixed layer generates submesoscale eddies. These eddies convert north-south density gradients to the vertical direction, rapidly stratifying the mixed layer at the meridional rate of progression of a few kilometers per day. The development of stratification in the upper ocean facilitates prolonged exposure of phytoplankton and nutrients to light and brings on a bloom. We compare model results with field observations to ascertain whether this mechanism is at play in the initiation of the North Atlantic bloom.
OS31A-1244 INVITED
The North Atlantic Oscillation: Climatic Significance and Environmental Impact
Marine ecosystems are undergoing rapid change at local and global scales. To understand these changes, including the relative roles of natural variability and anthropogenic effects, and to predict the future state of marine ecosystems requires quantitative understanding of the physics, biogeochemistry and ecology of oceanic systems at mechanistic levels. Central to this understanding is the role played by dominant patterns or modes of atmospheric and oceanic variability, which orchestrate coherent variations in climate over large regions with profound impacts on ecosystems. A leading pattern of weather and climate variability over the Northern Hemisphere is the North Atlantic Oscillation (NAO). The NAO refers to a redistribution of atmospheric mass between the Arctic and the subtropical Atlantic, and swings from one phase to another produce large changes in surface air temperature, winds, storminess and precipitation over the Atlantic as well as the adjacent continents. The NAO also affects the ocean through changes in heat content, gyre circulations, mixed layer depth, salinity, high latitude deep water formation and sea ice cover. Thus, indices of the NAO have become widely used to document and understand how this mode of variability alters the structure and functioning of marine ecosystems. There is no unique way, however, to define the NAO, and so several approaches will be discussed including nonlinear techniques that reveal spatial asymmetries between different phases of the NAO that are likely important for ecological studies. It also follows that there is no universally accepted index to describe the temporal evolution of the NAO. Several of the most common measures are presented and compared. All reveal that there is no preferred time scale of variability for the NAO: large changes occur from one winter to the next and from one decade to the next. There is also a large amount of within-season variability in the patterns of atmospheric circulation of the North Atlantic, so that most winters cannot be characterized solely by a canonical NAO structure. A better understanding of how the NAO responds to external forcing, including sea surface temperature changes in the tropics, stratospheric influences, and increasing greenhouse gas concentrations, is crucial to the current debate on climate variability and change.
OS31A-1245 INVITED
Direct Measurements of Particulate Organic Carbon Flux Through the Twilight Zone During the North Atlantic Bloom Using Neutrally Buoyant Sediment Traps
The spring phytoplankton bloom in the North Atlantic is thought to contribute a significant proportion of the global export of particulate organic carbon (POC) out of the euphotic zone and through the mesopelagic zone. Primary production in this region has been previously estimated at 200 g C m-2 year-1 and higher. POC fluxes were measured between 150 m and 700 m from four deployments of neutrally buoyant sediment traps (PELAGRA) made during the initial development of the North Atlantic spring bloom in May 2008. In addition, fluxes of particulate organic nitrogen, inorganic carbon, and biogenic silica were measured. These data show that the sinking of the bloom is characterised by sedimentation of discrete pulses of material such that at certain times deeper traps collect several-fold more material than shallower traps, reflecting the passage of particle pulses through the water column. During the time period of maximal flux, increasing concentrations of suspended particles at depth were observed in CTD and Seaglider profiles of chlorophyll fluorescence and optical backscatter. We have applied a range of depth-normalisations, from published b-values of the 'Martin-curve', to estimate the export flux at 100 m depth. Daily fluxes during these pulsed events contributed between 1 and 8 % of the annual export reported in the literature using very different approaches (36 g C m-2 year-1). Data from concomitant ship-board measurements of total 234Th disequilibrium during the cruise provide a further constraint on the magnitude of export production.
OS31A-1246
The 2008 North Atlantic Spring Bloom Experiment III: Bloom dynamics
As part of the 2008 North Atlantic Spring Bloom Experiment (NAB08), 2 Lagrangian Floats and 4 Seagliders were deployed in early April before the bloom began. Mixed-layer depth exceeded 150m; chlorophyll concentrations were low and dissolved nutrient concentrations were high; CTD profiles of chlorophyll fluorescence and optical backscattering (proxies for phytoplankton and particle concentrations, respectively) were low and uniform throughout the mixed layer. By mid April autonomously-sensed chlorophyll fluorescence and particle backscattering began to exponentially increase, with a concomitant increase in dissolved oxygen. Phytoplankton biomass and oxygen concentrations were still increasing in early May when the R/V Knorr arrived on station for a 3-week process cruise. Satellite imagery indicated that the bloom was spatially patchy in early May. Large diatom cells, including chain-forming species, dominated the phytoplankton community in early May. By mid-May, community dominance shifted to dinoflagellates and picoeukaryotes as nutrient concentrations decreased. Discrete spikes in chlorophyll fluorescence, optical backscatter, and beam transmission began to appear below the mixed layer, first extending to 250m and then progressively deeper (deepest depth sampled was 900m, by Seagliders). The deep optical spikes were associated with sinking organic particles; the largest catch of sinking cells, based on material collected by PELAGRA floating sediments traps, coincided with a horizontally and vertically wide-spread distribution of optical spikes observed both with the ship's CTD and by Seagliders. Autonomous measurements continued until the end of June, with brief ship-based sampling in mid and late June. Phytoplankton and particle biomass oscillated but remained higher than pre-bloom conditions, as did concentrations of dissolved oxygen. The frequency of subsurface optical spikes diminished, suggesting that the major carbon flux event occurred in mid May. These observations show the power of coordinated autonomous and ship-based sampling, using complementary bio-optical and biogeochemical measurements to measure coupled biophysical processes controlling ocean carbon fluxes.
OS31A-1247 INVITED
Particle Size Distributions and Optical Backscatter Measured in situ With the SOLOPC During the 2008 North Atlantic Bloom Experiment
At the start of the 2008 North Atlantic Bloom Experiment (NABE08), we deployed two SOLOPCs, autonomous profiling floats which measured vertical distributions of particle sizes and abundances, along with temperature, salinity, depth and optical backscatter (Checkley, D.M., Jr., R.E. Davis, A.W. Herman, G.A. Jackson, B. Beanlands, and L.A. Regier. 2008. Assessing plankton and other particles in situ with the SOLOPC. Limnol. Oceanogr. 53: 2123-2136). Our goal has been to relate changes in particle size distributions as a result of bloom dynamics to measurements of water-leaving radiance. We hypothesized that the particle assemblage would shift over time from small particles more efficient at scattering to larger particles, especially aggregates, and that this would be manifest in the particle size distribution. One float collected 13 profiles over a depth range of 5-150 m during the first day. Particles had the characteristic volumetric distribution we have previously observed in the Pacific Ocean off California. The normalized particle volume spectrum (nVd) was unimodal, with its peak between 200 and 400 microns equivalent spherical diameter. The mean size of particles increased with depth. The total particle volume was greatest at a depth of about 20-40 m. A non-linear relationship was observed between total particle volume and the average backscatter in shallower waters, where backscatter was constant while the total particle volume increased with depth. In deeper waters, the two were linearly related. Our results are consistent with a model in which small particles, produced and evenly distributed in the mixed layer, cause most of the backscatter, while large particles form from small particles due to aggregation and increase in size and abundance with depth due to sinking. The linear relationship of abundances of large and small particles deeper remains unexplained. Our results indicate that new SOLOPCs, under construction and with LOPC and chl a and optical backscatter sensors, will allow a test of our hypothesis in the future. Analysis of transmissometer data may also allow a test of this hypothesis for the NABE08. This work is supported by NASA and NSF.
OS31A-1248
Contribution of Phytoplankton Blooms to Export Production
The oceans absorb one quarter of the fossil fuel CO2 emitted to the atmosphere every year. The interannual and decadal variations of this sink are of strong interest, yet they are poorly constrained by observations and we have low understanding of the physical and biological processes that control the surface ocean partial pressure (pCO2). An important drawdown of pCO2 is driven by the sinking of organic matter following the intense phytoplankton spring blooms. The current assumption is that phytoplankton blooms will continue into the future with the same level of activity as today. We estimate the contribution of phytoplankton blooms to the biological export production (EP) using satellite chlorophyll data from CZCS and SeaWiFS sensors and carbon export fluxes from an inverse model based on climatological nutrient concentrations. Phytoplankton blooms are characterized by their initiation date, amplitude and duration. In the North Atlantic Ocean, long blooms of 10-15 weeks occur in the subtropics and are associated with low amplitudes and low EP. Short blooms of less than five weeks occur in the subpolar region and are associated with high amplitudes and high EP. The relation between bloom duration and EP can be characterized by an exponential inverse function. Using this function, we estimate changes in dissolved inorganic carbon and derive the associated changes in surface oceanic pCO2. Interannual and decadal patterns of EP and pCO2 distribution are evidenced for the entire basin.
OS31A-1249
Temporal Evolution of Small-Scale Bio-Optical Variability During the 2008 North Atlantic Bloom Experiment: A Satellite-Based Perspective
Understanding carbon flux in the north Atlantic requires the resolution and sampling of processes at a variety of spatial and temporal scales. Satellite ocean color data helped provide a larger horizontal context for in situ physical and optical measurements collected using a combination of gliders, mixed-layer floats, and ship-based surveys during the 2008 North Atlantic Bloom Experiment (NAB08; April-June and 58.5-61.5 °N). Similar increases during the spring bloom were observed in both the in-water and satellite-derived proxies for phytoplankton biomass. However, satellite observations were unique in their ability to provide synoptic snapshots of the intense mesoscale (O(10-250 km)) and sub-mesoscale (O(<10 km)) variability which was characteristic of the productive waters north of 60 °N throughout the experiment. While the sequence of satellite imagery is incomplete due to variable cloud cover, a sufficient number of images were collected and used to quantify the temporal evolution of small-scale spatial variability (<100 km) in satellite- derived estimates of chlorophyll a, optical backscatter by particles (bbp), and sea-surface temperature. Variogram analysis is a commonly used statistical method of characterizing spatial structure and dependencies in geographically referenced datasets. After removing large-scale trends (>100 km) from individual scenes by applying a 2D kernel-based high pass filter, the spatial structure of the residual variability was characterized by estimating variograms from data collected in the vicinity of the in-water assets. Using satellite data to quantify near-surface spatial structure at scales <100 km helps to constrain and simplify the interpretation of variability observed in the analogous in-water products which may be highly resolved in depth, but less-so in the horizontal dimensions (thereby making local changes difficult to distinguish from spatial patchiness). This work is a first step towards the optimization of future sampling efforts in the north Atlantic, especially those that will rely heavily on the use of new autonomous sampling platforms.
OS31A-1250
Application of variational data assimilation to coupled physical-biological models of the North Atlantic Bloom
Lagrangian floats and seagliders were deployed in the North Atlantic region south of Iceland from late March to early July 2008 and provided 3-D coverage of the spring bloom over time. The measured physical, chemical and bio-optical data, calibrated with data collected on three supporting cruises, is used to develop an ecosystem model describing the bloom and its associated carbon fluxes. The model's physical framework is based on the 1-D General Ocean Turbulence Model (GOTM) which is set up for a North Atlantic site at 60°N, 20°W and forced with data on wind speed, air pressure, air temperature and humidity. Initially, we coupled this physical model to an NPZD biological model with dynamic chlorophyll added. The model was then extended to include a silica cycle and diatoms as a separate phytoplankton group in order to simulate the presumed effect of silicate depletion on diatom sinking. We determined the biological parameters that are most important for model behavior through a sensitivity analysis and will apply variational data assimilation to optimize these model parameters and systematically assess the model-data consistency of the different model variants. The final model will be used to estimate carbon fluxes and export during the bloom.
OS31A-1251
Bio-optical observations of the North Atlantic Spring Bloom from continuous underway surface measurements
The evolution of the North Atlantic Spring Bloom was observed using continuous, underway measurements of particle absorption and scattering indices in a ~10000 km2 area centered on 61N, 26W. Spectral particulate beam attenuation (cp), absorption (ap), and backscattering (bbp) were measured simultaneously using a novel flow-through system with an actuated valve and timer to intermittently direct seawater flow through 0.2 um filters. Initially, the phytoplankton community was dominated by large chain-forming diatoms, which were then succeeded by dinoflagellates and ultimately by picoeukaryotes, as revealed by flow cytometry, microscopy, and HPLC pigment analysis. Despite the wide range in taxonomic variability, consistent relationships between cp and bbp were observed. Further, the slope of this relationship was similar to that obtained in very different oceanic environments (from oligotrophic to mesotrophic). This covariability of the two scattering indices suggests that they are, in large part, sensitive to the same pool of particles or that the particle size distribution is relatively conserved across large gradients in bulk chlorophyll (and presumably chlorophyll degradation products, as well). Coulter counter-based particle size distributions ranged from perfectly Junge-like to having pronounced peaks over the size range investigated (2-60um). In contrast to previous studies in the region, the bio-optical imprint of coccolithophorids was not observed in any of the quantities measured. Empirical relationships between optical measurements (cp, ap, bbp) with coincident biogeochemical measurements such as chlorophyll, particulate organic carbon, and biovolume-estimated phytoplankton specific carbon will also be discussed.
OS31A-1252
The 2008 North Atlantic Spring Bloom Experiment: Observations of a Stationary Eddy on the Eastern Flank of the Reykjanes Ridge
Four Seagliders and two Lagrangian floats were deployed before the onset of the North Atlantic Bloom (NAB) at 59.0°N 20.5°W in early April 2008. The Seagliders surveyed around the Langrangian floats, which were designed to drift in the mixed layer and meandered northwest across the Icelandic Basin during the development of the bloom, finally becoming entrained in an anticyclonic eddy on the flank of the Reykjanes ridge, near 61.5°N 25.5°W, during the month of May. A persistent 10-km scale chlorophyll patch in the northeastern quadrant of the eddy is clearly visible in measurements by the in-situ optics sensors on the autonomous platforms and in imagery from the Aqua MODIS satellite on 12 May 2008. From 2 - 22 May, a process cruise by the R/V Knorr provided CTD casts and water samples necessary to calibrate the in-situ sensors, as well as continuous flow-through and ship-mounted RDI 75 and 150 kHz ADCP measurements in bow-tie survey patterns in the study region to complement the floats and Seagliders. The physical and biological structure of the eddy is described using the integrated dataset. A quantitative understanding will be developed from this description in future work, with the ultimate goal of identifying biophysical couplings that affect the spatial and temporal development of the phytoplankton community.
OS31A-1253
The 2008 North Atlantic Spring Bloom Experiment: Spectral Particulate Absorption Coefficients
During the 2008 North Atlantic Spring Bloom Experiment (NAB08), spectral absorption coefficients of particulate material were measured by two techniques: the Quantitative Filter Technique (QFT) on water samples collected from the CTD/Rosette and in-water measurements collected with a single ac-9 as the difference between filtered and unfiltered profiles. Phytoplankton and non-phytoplankton absorption spectra from the QFT were determined by the Kishino methanol extraction procedure; these coefficients provide the basis for enhanced separation of ac-9 particulate absorption coefficients into phytoplankton and non- phytoplankton components. The blue-to-red ratio (440nm:676nm) of phytoplankton absorption, a diagnostic of photoprotective pigmentation, varied by approximately 25%. Visible and UV light penetration was measured by two radiometric profiling systems: a free-fall hyperspectral radiometer (350-800nm) deployed from the ship and a hyperspectral radiometer mounted on a Lagrangian mixed-layer float (320-950nm). Despite persistent cloud cover, phytoplankton absorption at wavelengths typically dominated by microsporin-like amino acids (MAAs, ~ 320nm) were observed in all spectra in the upper 50m. Phytoplankton community composition changed over the course of the experiment, from a diatom-dominated assemblage to a mixed community of dinoflagellates, picoeukaryotes, and heterotrophic flagellates.
OS31A-1254
Return from the depths: examination of diatom resting cysts from the North Atlantic Bloom Experiment
Mass sinking of cells following spring diatom blooms is known to be a major source of carbon flux to the ocean's interior. The timing of flux events is thought to correlate with nutrient depletion of surface waters, inducing physiological stress and robbing silicified diatoms of their ability to maintain buoyancy. Here, we examined sediment trap material collected during the 2008 North Atlantic Bloom Experiment to determine the physiological status of a sinking spring diatom bloom and evaluate its potential to "seed" future blooms. Sediment trap material was collected from 24hr deployments of floating sediment traps set at 300, 600 and 750m. Diatom cells were examined immediately after trap recovery and the assemblage was dominated by resting cysts of the diatom genus Chaetoceros. Resting cysts were strongly fluorescent under excitation at 470-490nm indicating that accessory pigments were intact and that cells were photosynthetically competent. Variable fluorescence measurements of bulk sediment trap material were high indicating the presence of live and potentially active cells. To test the hypothesis that cysts were viable and could re-seed a future bloom, sediment trap material from each depth was inoculated into nutrient-amended seawater and incubated at 9°C and 50 μmol photons m2 s1 on a 12:12 light:dark cycle. Within 38hrs, the heavily silicified valves of Chaetoceros sp. cysts were cast off and replaced by vegetative valves indicating that excystment had occurred. Initial excystment was followed by rapid cell division, with growth rates of Chaetoceros sp. reaching 1 doubling day1. This phenomenon was observed in samples collected on May 15. Sediment trap materials collected from deployments on May 11 and 19 were not dominated by resting cysts. Diatom resting stages, such as cysts, are known to form within 6 hrs and can sink up to 35 times faster than vegetative cells, potentially explaining the brief pulse of cysts we observed and highlighting the ephemeral nature of important C flux events in the open ocean.
OS31A-1255
Response of phytoplankton biomass and primary production to changes in community size structure during the 2008 North Atlantic Spring Bloom
Oceanic primary production plays a critical role in the global carbon cycle. The flux of carbon material from the upper ocean, however, is dependent on the quality and quantity of the carbon biomass produced. In the North Atlantic, large diatom chains dominate the phytoplankton community structure early in the spring bloom and can promote a significant sinking particulate organic carbon flux. As the spring bloom progresses and stratification sets in, changing phytoplankton community structure results in more recycled and regenerated forms of organic carbon prohibiting efficient carbon flux from the upper ocean. During the May 2-21, 2008 North Atlantic Bloom Experiment southwest of Iceland, size-fractionated (total, <20um) fluorometric chl a concentration and photosynthetic parameters from short-term 14C-uptake incubations were measured between 10am - 2pm at the surface (10m) and at depth (usually 30m) to determine the evolution of phytoplankton population size structure and primary productivity during the spring bloom. During the first two weeks of the cruise, large phytoplankton (>20um, mostly large diatom chains) contributed over ~80% to the total chl a signal at both depths. Over the final week, however, the surface phytoplankton community transitioned to relatively smaller cells (<20um) that contributed at least 60% to total Chl a indicating a shift away from a diatom chain based community structure. A similar phytoplankton community structure shift appeared to begin at depth as the cruise ended with similar implications for total Chl a contribution. Quantification of photosynthetic parameters demonstrated a high degree of spatial and temporal variability at both 10 and 30m which appeared to be correlated with variability in chl a concentration over the course of the cruise. Pmax varied from approximately 20 to 90 mg C m-3 h-1 at the surface and 10 to 60 mg C m-3 h-1 at depth. When maximal values were observed, large cells dominated the production (>80%); when low values were observed small cells made a significant (30-80%) contribution. The pattern suggests that the smaller size fraction provides a relatively constant background level of biomass and production while the dynamic bloom is driven by the biomass and production of the large cells, in this case, chain-forming diatoms.