program & abstracts - university of washington · 4:00 guillaume dencausse, rosemary morrow,...

Program & Abstracts

2

Argo and Altimetry WorkshopManchester Grand Hyatt Hotel, San Diego

Tuesday, October 18, 2011

8:30–8:40 Introduction

Session 1: Sea level variations, heat and freshwater storage

8:40 Dean Roemmich and John Gilson: A global view of steric and sea surface height variability, 2004 – 2010

9:00 Don Chambers and Josh Willis: Can a combination of altimetry, Argo, and GRACE detect deep ocean warming?

9:20 Karina von Schuckmann and Pierre-Yves LeTraon: Estimation of global ocean indicators from Argo with focus on regional and deep ocean contributions to global steric sea level

9:40 Josh Willis and Takmeng Wong: Global sea level and the planetary energy balance: What Argo, altimetry and other observing systems tell us about radiative forcing.

10:00 Shigeki Hosoda, Toshio Suga, and Keisuke Mizuno: Recent global changes in surface layer salinity depicted by Argo: Footprint of the global hydrological cycle enhancement

10:20–10:40 Break

10:40 Stephanie Guinehut, Michael Ablain, Marie-Helene Rio, Guillaume Valladeau, Jean-Francois Legais, and Gille Larnicol: Data quality assessment of Argo and altimeter measurements through SSH comparisons

11:00 Paul Barker, Jeff Dunn, Catia Domingues, and Susan Wijffels: Global and regional impacts of Argo pressure drifts on thermosteric sea level (Presented by Neil White)

11:20–11:40 Discussion: Argo data quality needed for SSH and global change

Session 2: Ocean circulation

11:40 Marie-Helene Rio, Stephanie Guinehut, Sandrine Mulet, and Gille Larnicol: Monitoring the ocean from observations

12:00 Steven Jayne, Paul Robbins, and Breck Owens: Time-mean upper ocean circulation of the North Atlantic

3

Argo and Altimetry Workshop

3

12:20–1:30 Lunch

1:30 Kathryn Kelly, LuAnne Thompson, and John Lyman: Estimates of Heat and Mass Budgets in the Atlantic Ocean

1:50 Nathalie Zilberman, Dean Roemmich, and Sarah Gille: The mean and the time-variability of the shallow meridional overturning circulation in the tropical South Pacific Ocean

2:10 Donata Giglio and Dean Roemmich: Wind-driven variability of the subtropical Pacific Ocean

Session 3: Mesoscale dynamics

2:30 Bo Qiu and Shuiming Chen: Effect of decadal Kuroshio Extension variability on the modification of North Pacific Intermediate Water

2:50 Leonid Ivanov, Curt Collins, and Tetyana Margolina: Multi-scale analysis of satellite altimetry observations off California

3:10–4:00 Break and Poster Viewing

4:00 Guillaume Dencausse, Rosemary Morrow, and Francesco D’Ovidio: Simulating (sub-) mesoscales in Argo float-derived sea surface tracer fields using altimeter-derived advection south of Tasmania

4:20 Peter Gaube and Dudley Chelton: Observations of eddy-induced Ekman pumping sustaining phytoplankton blooms in Indian Ocean anticyclones

4:40 Toshio Suga, Kanako Sato, Shigeki Hosoda, Taiyo Kobayashi, Fumiaki Kobashi, Noriyuki Matsuo, Hiroyuki Nakajima, Katsuya Toyama, and Toshiro Saino: Biogeochemical impact of mesoscale disturbance in the subtropical North Pacific

5:00–5:20 Discussion: Deep Argo sampling requirements for understanding SSH

Posters

Melissa Bowen and Philip Sutton: Mean dynamic height in boundary currents: an evaluation of observations from the East Auckland Current

Celine Heuze and Frederic Vivier: Analysis of the interannual variability of the Southern Ocean’s mixed-layer temperature and salinity from hydrographic profiles

Philippe Rogel, David Salas Y Melia, Benoit Meyssignac, Melanie Becker, Emilia Sanchez, Christophe Cassou and Elodie Fernandez: Reliability estimates of decadal sea-level trend hindcasts: a case of predictions with uncertain verification data

44


Posters (cont.)

Stavroula Biri and Detlef Stammer: Temporal variability of eddy kinetic energy and eddy statistics

Ananda Pascual, Simon Ruiz, Bruno Buongiorno Nardelli, Stephanie Guinehut, Yann Drillet, Sandrine Mulet, Marie-Helene Rio, Francesco Bignami, Gille Larnicol, and Joaquin Tintore: Vertical motion estimated from combined ARGO and altimetry observations

Oleg Melnichenko, Nikolai Maximenko, James Potemra, and Peter Hacker: Spatial patterns and variability of near-surface vertical gradients of salinity from historical CTD and Argo float data

Koji Ogawa, Norihisa Usui, Yosuke Fujii, Takahiro Toyoda, Takanori Iwao, Masafumi Kamachi: Impact of Argo data in a data assimilation system for the North Pacific area

William Llovel, Ichiro Fukamori, and Benoit Meyssignac: Quantifying the thermosteric sea level contribution for different oceanic layers at global and basin scales over 1950-2010.

jkvamme

Text Box

For archived workshop information: http://www.aviso.oceanobs.com/en/courses/sci-teams/ostst-2011/

5


5

ABSTRACTS

66


A global view of steric and sea surface height variability, 2004—2010Roemmich, Dean

Dean Roemmich, Scripps Institution of Oceanography UCSDJohn Gilson, Scripps Institution of Oceanography UCSD

The Argo Program achieved a global array of profiling floats in 2004, with spatial coverage continuing to increase until the objective of 3000 floats at approximately 3 degrees x 3 degrees spacing was met in 2007. The globally distributed and sustained nature of the Argo array provides satellite-like coverage, but of subsurface ocean variability, 0 – 2000 m, occurring within the Argo era or in Argo compared with historical datasets. In historical comparisons, multi-decadal increases in steric height have emphasized a strong role of deep warming in the Southern Ocean (e.g. Gille, 2008, Sutton and Roemmich, 2011). Steric change in the ocean extends below 2000 m (e.g. Purkey and Johnson, 2010), and the Argo Program is now planning to increase its depth capability to full-ocean profiling for better understanding of sea level variability and change.

The present work focuses on global seasonal-to-interannual variability during the Argo era, and in particular on comparing the patterns of steric height (SH) and altimetric sea surface height (SSH) anomalies. A challenge for the analysis is the different sampling densities of the satellite and subsurface datasets. Here, we identified over 400,000 Argo profiles having a nearly co-located altimetric height measurement from along-track data, and the analysis is based on these co-located pairs. After removing a common time mean, 2004-2010, the anomalies are averaged in large-scale monthly bins for comparison. Upper-ocean SH is the dominant contributor to SSH on seasonal-to-interannual time-scales. Nevertheless, the small differences between them may represent deep ocean steric variability, the mass-related component of SSH, or errors in either dataset. We consider global maps and zonal means of the binned SH and SSH variance, the seasonal cycle, the ENSO-related interannual variability, the extreme monthly values, and the 6-year trends. Some significant differences between SH and SSH are identified, such as SH having a larger seasonal range in the southern hemisphere than SSH. Argo’s time domain is still too short for decadal and longer-term variability, but the present dataset provides a “first look” at Argo’s capabilities for estimating steric trends. The longer SSH time-series allows us to compare the patterns emerging from the Argo era with those of the full SSH time-series.

7


7

Can a Combination of Altimetry, Argo, and GRACE Detect Deep Ocean Warming?Chambers, Don

Don Chambers, College of Marine Science, University of South FloridaJoshua Willis, Jet Propulsion Laboratory, California Institute of Technology

Sea level will fluctuate due to changes in both the density structure and the mass transported into and out of the region. Altimetry can measure the combined effect of both of these, temperature/salinity measurements from Argo can measure the density variations above 1000 m, while the gravity data from the GRACE mission can be used to infer the mass component. The residual, should, therefore, reflect changes in the density structure below 1000 m. This is generally small over short time-scales, but can be significant over periods longer than a year. Recently, Purkey and Johnson (2010) have used repeat deep hydrographic sections over the Southern Ocean to show that there has been a significant warming of the ocean below 1000 m south of the Subantarctic Front (approximately 50°S), with a trend equivalent to 0.87 ± 0.76 mm/year of steric sea level over the last decade.

Using 6 years of altimetry (primarily Jason-1 and Jason-2), Argo, and GRACE data, we evaluate whether residuals averaged over the area south of 50°S have a trend that is significant and consistent with the results of Purkey and Johnson (2010). Although very preliminary, we do find a residual trend of slightly more than 1 mm/year, although uncertainty is still high due to the short record. The results suggest that with a long record of altimetry, Argo, and GRACE, we can detect deep ocean warming that is of the size that has been occuring in the Southern Ocean over the last decade.

88


Estimation of global ocean indicators from Argo with focus on regional and deep ocean contributions to global steric sea levelvon Schuckmann, Karina

Karina von Schuckmann, CNRS/LOCEAN/IFREMERPierre-Yves Le Traon, IFREMER

Argo temperature and salinity measurements during the period 2005 to 2010 are used to estimate global ocean indicators (GOIs) such as global ocean heat content (GOHC), global ocean freshwater content (GOFC) and global steric sea level (GSSL). A method based on a simple box averaging scheme is developed to estimate GOIs, together with an error estimation due to data sampling, data processing and climatology uncertainties. Over the six year time period, trends of GOHC and GSSL are 0.55±0.1 W/m2 and 0.69±0.14 mm/yr, respectively. The trend of GOFC is barely significant. Interannual variability at global scale can be observed, especially for GOFC. These results are valid under the assumption that no systematic errors remain in the global observing system.

In addition, regional and deep ocean (700-2000m) contributions to the Argo GSSL estimation are analyzed. Results show that previously neglected deep ocean and salinity effects have a significant impact on estimations of GSSL. Space scales of coherent patterns increase with increasing depth due to both temperature and salinity effects. As already shown in previous studies, density changes due to salinity effects compensate temperature changes in large areas of the global ocean. Steric sea level increase due to salinity effects alone occur as well and can reach down to 2000m depth. A comparison of Argo steric sea level to total sea level from satellite altimetry (AVISO) reveals that largest regional contributions to both GSSL and global total sea level can be observed in the global tropical basin.

9


9

Global Sea Level and the Planetary Energy Balance: What Argo, Altimetry and Other Observing Systems Tell us About Radiative ForcingWillis, Josh

Josh Willis, JPLTakmeng Wong, NASA Langley

The oceans absorb the vast majority of any net radiative imbalance in the global climate system, such as the one caused by the addition of anthropogenic greenhouse gases to the atmosphere. For this reason, quantifying the accumulation of heat in the global oceans provides important insight into the accumulated impact of human-caused climate change. The majority of recent changes in global sea level in recent years can be explained by ocean warming observed by Argo and ocean mass increase observed by the gravity mission GRACE. The ocean warming observed by Argo is in reasonably good agreement with estimates of the net radiative imbalance of the Earth as measured by the CERES satellites between about 2005 and the present. Prior to that, however, ocean warming data become dominated by measurements from eXpendable BathyThermograph (XBT) probes, which are known to suffer from a variety of biases. For this reason, the transition period in observations of ocean temperature, between 2003 and 2005, remains problematic for the sea level budget as well as the net radiative imbalance. This work will consider whether observations of the sea level budget from altimetry and GRACE, along with estimates of the net radiative imbalance from CERES can help constrain the pre-Argo estimates of ocean warming.

1010


Recent global changes in surface layer salinity depicted by Argo: Footprint of the global hydrological cycle enhancementHosoda, Shigeki

Shigeki Hosoda, JAMSTEC RIGCToshio Suga, Tohoku University/JAMSTEC RIGCKeisuke Mizuno, JAMSTEC RIGC

Surface layer salinity distributions and those characteristics in spatial and temporal variations on the global ice-free ocean were investigated using in situ observation, including Argo float data. Salinity in the surface layer is one of the most important measures indicating the condition of the ocean and climate. However, due to a lack of temporally and spatially homogeneous salinity data, previous observational studies did not detail global changes in salinity. Since the start of the Argo Project on 2000, the number of active Argo float is increasing and the Argo float array has allowed us to document changes in salinity of the global ocean. In the climatology calculated using historical data for 30 years in 1960-1989, the surface layer salinity is basically lower in the subpolar and tropical regions and higher in the subtropics. In the recent distribution of surface layer salinity, which was obtained from Argo float data in 2003-2007, an enhancement of contrast in lower and higher surface layer salinity distribution, except in the North Atlantic Ocean, was clarified through a comparison with the climatology. The enhancement was continuing through 2003-2007 in the global ocean, and was still on going until 2009. Since direct accurate estimation of freshwater flux (evaporation minus precipitation) from observations is difficult at the sea surface, this result shows that estimating the E-P flux from oceanic salinity is an effective alternative. To suggest a relationship between the surface layer salinity changes and the enhancement of the global hydrological cycle, changes in basin-scale freshwater flux was estimated using a simple box model on the basis of the difference in the surface layer salinity between recent Argo data and the climatology. The result suggests a possibility that the global hydrological cycle was strengthened up to a few percent rather than 30 years ago. This also indicates that surface layer salinity change is reflecting a possibility of a long-time climate trend, such as the global warming.

11


11

Data Quality assessment of Argo and Altimeter measurements through SSH comparisonsGuinehut, Stephanie

Stephanie Guinehut, CLSMichael Ablain, CLSMarie-Helene Rio, CLSGuillaume Valladeau, CLSJean-François Legeais, CLSGilles Larnicol, CLS

Multi-observations CalVal (Calibration/Validation) methods are widely used between in-situ and satellite data to assess the quality of the latest. The stability of the different altimeter missions is, for example, commonly assessed by comparing altimeter sea surface height measurements with those from arrays of independent tide gauges [Mitchum, 2000]. Other examples include the validation of altimeter velocity products with drifter data [Bonjean and Lagerloef, 2002; Pascual et al., 2009], or the systematic validation of satellite SST with in-situ surface temperature measurements from drifting buoys. Comparison of in-situ T/S profiles and altimeter data can also provide an indication of the quality of the in-situ data [Guinehut et al., 2009].

In this study, we present the main activities performed routinely at CLS as part of multi-observations CalVal activities through SSH comparisons between Argo and Altimeter measurements that are separated into four main tasks: (1) Validation of altimeter time series (T/P, Jason-1, Jason-2, Envisat) in order to detect drifts or jumps and for a global consistency check between the different altimeter missions; (2) Validation of new altimeter standards like orbit, geophysical corrections, ground processing; (3) Validation of each Argo floats time series in order to detect drift, spike or bias and to separate rapidly suspicious floats for more careful examination; (4) Validation and contribution of pressure and salinity corrections applied to the global Argo array.

1212


Global and regional impacts of Argo pressure drifts on thermosteric sea levelDomingues, Catia

Paul Barker, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania, AustraliaJeff Dunn, Centre for Australian Weather and Climate Research, CSIRO, Hobart, Tasmania, AustraliaCatia Domingues, Antarctic Climate and Ecosystems Cooperative Research Centre, Hobart, Tasmania, AustraliaSusan Wijffels, Centre for Australian Weather and Climate Research, CSIRO, Hobart, Tasmania, Australia

In recent years, autonomous profiling floats from the Argo program have become the prime component of the in situ ocean observing system. These data are now the dominant input to estimates of the evolution of global ocean heat content and (thermo)steric sea level. Subtle instrumental errors, however, can limit our ability to identify climate and sea level change signals because of the very low signal-to-noise ratio of these calculations. Here, we show that a large percentage of pressure measurements from the dominant type of Argo float, the Autonomous Profiling Explorer (APEX), as of January 2009, were uncorrected for sensor drift. Although positive and negative APEX pressure biases (mostly within 5 dbar but at times as large as 20 dbar) were corrected in this study, the bulk of the corrections were done for positive biases. About 43% of the APEX profiles were uncorrectable due to firmware limitations (negative pressure drifts truncated to zero (“TNPD” floats)) and due to insufficient technical files and/or metadata. Comparison between the corrected and uncorrected Argo datasets reveals that the pressure corrections remove significant regional errors from ocean temperature, salinity and thermosteric sea level fields. In the global mean, as the uncorrectable APEX float profiles appear to largely offset the effect of the correctable APEX float profiles with positive pressure drifts, our uncorrected 0-700 m global mean thermosteric sea level estimate lies within the one standard deviation error bars of our unbiased estimate. This indicates that the Argo pressure errors are too small to be detected in independent estimates of global (thermo)steric sea level inferred from altimetry and GRACE, and should not affect the closure of the sea level budget (within error bars) over 2003/4–2008. Overall, our study underscores the continuous need for careful analyses to detect and remove subtle errors in ocean observations and for complete and accurate technical/metadata information. The Argo national programs have now recovered most of the missing technical files and metadata (>95%) needed for pressure corrections, are properly flagging uncorrectable TNPD floats (which data should be excluded from global change studies), and ensuring the delivery of pressure-corrected APEX floats via the Global Data Assembly Centers (GDACs). Finally, we note that while inter-comparisons with altimeter and GRACE satellite observations are very useful to detect biases in Argo data, these analyses alone are not sufficient. Inter-comparisons with high-quality shipboard CTD data would be similarly important to enhance the calibration of Argo data.

13


13

Monitoring the ocean from observationsRio, Marie-Helene

Marie-Helene Rio, CLSStéphanie Guinehut, CLSSandrine Mulet, CLSGilles Larnicol, CLS

Producing comprehensive information about the ocean has become a top priority to monitor and predict the ocean and climate change. Complementary to modeling/assimilation approaches, an observation-based approach is proposed here. It relies on the combination of remote-sensing (altimetry and sea surface temperature) and in-situ (temperature and salinity profiles) observations through statistical methods.

The method uses first a multiple linear regression method to derive synthetic T/S profiles from the satellite measurements. These synthetic profiles are then combined with all available in-situ T/S profiles using an optimal interpolation method. The thermal wind equation with a reference level at the surface is finally used to combine current fields from satellite altimetry with the thermohaline fields to generate the global 3D current fields. Global temperature, salinity and current fields are thus available at a weekly period from the surface down to 1500-meter depth and a reanalysis is available for the 1993-2009 periods.

To assess the quality of this reanalysis, diagnostics from the common MyOcean WP4 validation protocol have been applied. In particular, volume and heat transports through various sections have been computed and compared to literature. Also, a 1993-2010 long time series of the maximum Atlantic Meridional Overturning Circulation strength has been computed that shows good agreement with results from RAPID-MOCHA monitoring project and GLORYS Mercator-Ocean reanalysis. The 3D velocity field has been further validated through comparison with independent current meter velocities available worldwide. Finally, the variability of the ocean temperature through the 1993-2009 periods has been analyzed. It shows a clear warming that is visible at all depths and for all latitudes. If the variability is baroclinic with strong interannual signals in the tropics, it shows a clear long term trend at high latitude with depth consistent signals.

1414


Time-mean upper ocean circulation of the North AtlanticJayne, Steven

Steven Jayne, Woods Hole Oceanographic InstitutionP. E. Robbins, Woods Hole Oceanographic InstitutionBreck Owens, Woods Hole Oceanographic Institution

We estimate the time-averaged upper ocean circulation using the time-mean sea surface height, the geoid, surface velocity observations from drifters, and Argo float profiles and displacements. These data are combined using an objective mapping technique that has been modified to include constraints for no flow normal to the boundary. The surface circulation for the North Atlantic is estimated from observed sea surface height and drifter velocity. Time-mean dynamic height of the sea surface referenced to 1000 dbar is computed from Argo profile data. Surface velocity, dynamic height, and Argo parking-depth displacement estimates are then combined to create a map of the mid-depth ocean circulation.

15


15

Estimates of Heat and Mass Budgets in the Atlantic OceanKelly, Kathryn

Kathryn Kelly, University of WashingtonLuAnne Thompson, University of WashingtonJohn Lyman, Pacific Marine Environmental Laboratory/NOAA

Anomalies of sea level in the Atlantic Ocean result from changes in the density and the mass of the ocean. Density changes from salinity are relatively small in the North Atlantic, so that the steric contribution to sea level there is primarily from changes in ocean heat content. In the absence of lateral motion (ocean transport convergences) and continental sources (rivers and melting ice sheets), surface heat and freshwater fluxes can be integrated to produce heat content and mass anomalies. Discrepancies between modeled and observed anomalies are the result of lateral convergences or sources, combined with errors in the fields and the model. Using a Kalman filter with an unknown control formulation and high-quality surface fluxes, we model the sea level components, constrained by heat content from in situ observations and gravity anomalies from GRACE, as well as altimetric sea level. From the unknown control we infer heat and mass transport convergences/sources in coast-to-coast latitude bands. The assimilation reconciles the relative contributions of heat and mass anomalies to sea level with the estimates of the errors. The heat budget alone is estimated from 1993; the heat and mass budgets are estimated jointly after 2002, when GRACE data are available. Regional heat convergences are integrated to estimate meridional heat transport (MHT), with suitable assumptions in high latitudes. Estimates are made for the Atlantic north of the Southern Ocean for comparisons with the cross-Atlantic RAPID array MHT near 26N. A more detailed model in the North Atlantic shows that much of the trend in sea level in the subpolar gyre is associated with changes in surface heating, rather than changes in mass or heat transport convergence.

1616


The mean and the time-variability of the shallow meridional overturning circulation in the tropical South Pacific OceanZilberman, Nathalie

Nathalie Zilberman, SIODean Roemmich, SIOSarah Gille, SIO

Gridded Argo temperature and salinity profiles and atmospheric reanalysis surface winds are used to study the time-variability of the meridional transport in the Pacific Ocean subtropical cell between 2004 and 2010. Particular attention is given to transport in the ocean interior, from 170E to the South American coast, but the western boundary west of 170E is also addressed. The poleward Ekman and equatorward geostrophic branches of the subtropical cell exhibit an El Nino-Southern Oscillation (ENSO) signature with strong meridional transport occurring during La Nina and weak meridional transport during El Nino. At 7.5S, mean basin-wide geostrophic transport from 1000 dbar is 48.5 ± 2.5 Sv (standard deviation based on annual mean values), of which 30.3-38.4 Sv returns to the subtropics in the surface Ekman layer, whereas 10.1-18.2 Sv flows northward feeding the Indonesian Throughflow (ITF). The geostrophic transport within the subtropical cell is stronger in the interior and weaker in the western boundary during La Nina. The opposite occurs during El Nino. The major portion of the time-variability in basin-wide geostrophic transport is due to changes in the interior. Interannual variability of altimetry-derived surface geostrophic velocity across 7.5S is consistent with Argo and allows the time-series to be extended into the pre-Argo era. The mean oceanic heat transport across 7.5S compensates half of the air-sea heat fluxes north of 7.5S. The heat transport in the ITF calculated as the residual of the oceanic heat transport across 7.5S and the air-sea heat fluxes north of 7.5S is 0.49 ± 0.35 PW using ECMWF and 0.53 ± 0.17 PW using NCEP. The time-variability of the heat transport in the ITF shows an ENSO signature with the heat transport rate one-third higher during La Nina than El Nino.

17


17

Wind-driven Variability of the Subtropical Pacific OceanGiglio, Donata

Donata Giglio, Scripps Institution of Oceanography, University of California San DiegoDean Roemmich, Scripps Institution of Oceanography, University of California San Diego

Wind-driven fluctuations in the circulation and temperature structure of the oceanic subtropical gyres have impacts on marine ecosystems and potentially on climate through ocean-atmosphere feedbacks in the western boundary current regions. Also, because of the long time-scales of oceanic response, the changes in large-scale ocean circulation are sensitive diagnostics of changing patterns in the noisier atmospheric forcing. Previous studies of subtropical variability have mostly been limited to altimetry and other surface ocean data or to western boundary current regions. The Argo array provides a unique dataset to explore variability of the subsurface ocean interior during the period of good Argo coverage since 2004. ECMWF wind stress provides maps of the atmospheric forcing. Two distinct patterns of surface wind stress are observed during 2004-2010 in the subtropical North Pacific. In 2004-2006, wind stress and Ekman upwelling anomalies are consistent with observed steeper isopycnals in the ocean interior and hence indicate a stronger gyre circulation. In 2007-2009, reversed anomalies in wind forcing suggest a weaker gyre. Evidence is shown in ocean maps of integrated dynamic height and pressure anomaly on isopycnals. Also, a horizontal movement of the gyre is detected in these two different periods. Earlier observations of the wind forcing indicate that similar patterns of surface momentum flux occurred between the strong El Nino event in 1997-98 and 2004. The 18-year time-series of altimetric height allows the analysis to be extended into the pre-Argo period.

1818


Effect of Decadal Kuroshio Extension Variability on the Modification of North Pacific Intermediate WaterQiu, Bo

Bo Qiu, University of Hawaii at ManoaShuiming Chen, University of Hawaii at Manoa

Salinity modifications in the North Pacific Intermediate Water (NPIW) core layer of 26.7-26.8 sigma_theta in the western North Pacific Ocean are investigated using temperature-salinity data from available profiling float and hydrographic measurements in 2002--09. During 2002--05 when the Kuroshio Extension (KE) jet was intense and zonally elongated, coherent positive salinity anomalies appeared along the inflow KE jet southeast of Japan and in the downstream Mixed Water region east of 152E. Broad-scale negative salinity anomalies were detected south of the KE jet and in the upstream Mixed Water region west of 152E. The signs of these observed salinity anomalies were reversed in 2006-09 when the KE jet transitioned to a weakened and zonally-contracted dynamic state. By adopting an isopycnal advection-diffusion model and conducting model runs with the time-dependent advective field inferred from the eddy-resolving, satellite altimeter sea surface height data, it is found that the observed salinity anomalies are oscillatory in nature and are determined not only by the decadally-varying KE jet itself, but also by mesoscale eddy signals that modulate temporally and longitudinally along the path of the KE jet.

19


19

Multi-Scale Analysis of Satellite Altimetry Observations off CaliforniaIvanov, Leonid

Leonid Ivanov, Moss-Landing Marine Laboratories, San Jose state UniversityCurt Collins, Naval Postgraduate SchoolTetyana Margolina, Naval Postgradute School

A new approach for multi-scale analysis of satellite altimetry observations based on double spectral representation of oceanic fields has been developed to understand variability of ocean flows at different spatial and temporal scales. The approach uses a set of spatial basis functions constructed from knowledge of the basin geometry and a filtering procedure based on a discrete wavelet transform. Metrics to distinguish between different types of oceanic flows have been developed through phase-frequency analysis of satellite signals. The approach allows highly-effective detection and separation of coherent structures (jets, waves and eddies) for in-situ as well as satellite ocean observations, development of new methods of non-assimilation fusion of data and models, comparison between satellite data and high-resolution models at different spatio-temporal scales as well as new interpretations of SSH altimetry signals in terms of oceanographic processes.

The approach has been applied to a new interpretation of SSH anomalies propagating off California as a signature of weakly nonlinear Rossby waves rather than nonlinear eddies. Satellite altimetry data are combined with ARGO and RAFOS float data. Different propagation regimes for Rossby waves were identified using propagation speed, wave steepness (a measure of wave nonlinearity) and degree of spatial coherence. The waves typically evolved as weakly nonlinear waves with recurrence time of about 105-120 (195-210) days for the semiannual (annual) component. When wave amplitudes reached maximum possible values (the saturation stage), the spatial coherence of the wave field decreased considerably and wave packets and separated features (interpreted as mesoscale eddies) were clearly observed. Propagation speed also decreased with the degree of nonlinearity of the wave field and westward propagation (zero mean propagation speed) stopped in the saturation stage. Our analysis has also shown that mesoscale flow off California presents a self-organized system driven by near resonance interactions between different flow scales. Quartet (modulation) instability dominated and caused a non-local transfer of energy from Rossby waves and eddies to bi-annual oscillations and quasi-zonal jets.

2020


Simulating (sub-)mesoscales in Argo float-derived sea surface tracer fields using altimeter-derived advection south of TasmaniaDencausse, Guillaume

Guillaume Dencausse, LEGOS/OMP UMR 5566Rosemary Morrow, LEGOS/OMP UMR 5566Francesco D’Ovidio, LOCEAN UPMC UMR 7159

Meso- and submesoscale processes at the ocean surface are believed to play an important role in the global ocean, creating strong horizontal gradients that could impact lateral mixing, as well as water mass formation through up- and downwelling (Sallée et al., 2008). An important next step for realistic ocean models is therefore to better simulate or parameterize mechanisms at these finer-scale ocean fronts. However, such improvements require much higher observational coverage than is currently available.

Surface lagrangian advection with time-evolving altimeter derived SSH geostrophic velocities has been recently shown to simulate quite effectively some submesoscale processes (d’Ovidio et al., 2009). This method thus potentially offers plenty of information on meso- and submesoscale processes on a global scale. However, the method’s limitations – such as the non-divergence of the simulated horizontal flow or the absence of water property modifications during advection – need to be carefully examined on a regional basis.

In this study, we evaluate the altimetry-advection method’s ability to simulate mesoscale and submesoscale activity in the Southern Ocean region south of Tasmania. The initial conditions are derived from large-scale weekly SSS and SST gridded fields, constructed using Argo floats and other in-situ data and an objective analysis (Coriolis). The low-resolution tracer fields are then advected using the time-evolving altimetric currents over a 2-3 week period, and this “stirring” induces smaller scale structures. The advected mesoscale tracer fields are then validated with fine-scale underway surface data from the Survostral repeat cruise section between Tasmania and Antarctica over the period 2002-2007, and data from the Advanced Very High Resolution Radiometer.

Our primary aim is to evaluate whether the Argo float-derived Coriolis products can be combined with altimetry to reproduce horizontal submesoscale structures in this region. The results are very promising. We have analyzed the simulated tracer fields in comparison to the independent data, and decompose the biases into two categories. One part of the bias is introduced by the advection technique itself, for which possible corrections are proposed. The other part of the bias originates from the Coriolis fields, and the method reveals the improvements in the quality of these weekly Argo-float derived products, as the data distribution evolves over the period 2002-2007.

21


21

Observations of Eddy-induced Ekman Pumping Sustaining Phytoplankton Blooms in Indian Ocean AnticyclonesGaube, Peter

Peter Gaube, Oregon State UniversityDudley Chelton, Oregon State University

Mesoscale eddies are known to modulate near-surface ocean biogeochemical cycles. Numerous past studies have discussed the biological importance of upwelling of nutrients into the interiors of nonlinear eddies. Such upwelling can occur during the transient stages of formation of cyclones from shoaling of the thermocline. In their mature stages, upwelling can occur from Ekman pumping driven by eddy-induced wind stress curl. Eddy surface currents act to drive a local wind stress curl over the rotating interior of an eddy even in a spatially uniform wind field. This eddy-Ekman pumping results in average upwelling velocities at the cores of anticyclones on the order of 10 cm day-1, often exceeding 1 m day-1 and has been cited as sustaining anomalous phytoplankton blooms in the cores of highly nonlinear anticyclones in the North Atlantic. We show that globally, regions exist where a clear relationship between eddy-Ekman pumping and phytoplankton variability can be observed. In this presentation we highlight the Eastern Indian Ocean where large, highly nonlinear eddies form along the west coast of Australia and propagate into the oligotrophic central Indian Ocean.

The Leeuwin Current is an anomalous eastern boundary current flowing poleward along the coast of Western Australia. Seasonally, large anticyclonic eddies pinch off of the meandering Leeuwin Current entraining the relatively eutrophic shelf waters of Western Australia. Using an automated eddy tracking procedure we identify and track nearly 500 individual long-lived anticyclonic mesoscale eddies (spatial scales of ~ 100km with lifetimes ≥ 12 weeks) to investigate the nature of the large phytoplankton blooms observed in their cores. Chlorophyll-a concentration derived from observations made by the SeaWiFS radiometer along with vector wind estimates made by the QuikSCAT sensor are collocated with the interiors of eddies identified by the automated eddy tracking procedure to create composite averages of eddy-induced chlorophyll anomalies in a frame of reference translating with the eddies.

The observed blooms in the cores of anticyclonic eddies in the Eastern Indian Ocean are seasonal in nature, with a maximum chlorophyll anomaly during the Austral fall and winter. During this time of year, wind mixing and a decrease in incident solar radiation act to deepen the mixed layer. The seasonal variability of mixed layer depth within eddies is investigated by collocating individual ARGO float profiles with the interiors of eddies. Our observations suggest that the seasonal nature of the biological response to eddy-induced Ekman pumping is linked to changes in mixed layer depth

2222


Biogeochemical impact of mesoscale disturbance in the subtropical North PacificSuga, Toshio

Toshio Suga, Tohoku University / JAMSTECKanako Sato, JAMSTECShigeki Hosoda, JAMSTECTaiyo Kobayashi, JAMSTECFumiaki Kobashi, Tokyo University of Marine Science and Technology / JAMSTECNoriyuki Matsuo, Marine Works JapanHiroyuki Nakajima, Marine Works JapanKatsuya ToyamaTohoku UniversityToshiro Saino, JAMSTEC

Mechanisms to supply the nutrients required to sustain primary productivity of the subtropical ocean are not fully understood yet. Intermittent upwelling induced by mesoscale disturbances has been considered as one of the possible mechanisms, while observation of such a process has been very limited. Combination of profiling float observation and satellite altimetery would be an effective tool to evaluate the impact of mesoscale disturbance on the primary productivity. As a preliminary experimental study using this tool, we deployed five profiling floats equipped with oxygen sensors in the vicinity of the biogeochemical time series station S1 (30N, 145E) south of the Kuroshio Extension in September-October 2010. The profiling cycle of those floats were 2 or 3 days. Shallow oxygen maximum (SOM) in the seasonal pycnocline at 50-100 m is a typical feature in this area, which reflects subsurface primary productivity. One of the floats drifting across a cyclonic eddy detected clear oxygen increase in SOM associated with isopycnal heaving, which possibly indicated nutrient supply due to upwelling induced by the passage of the cyclonic eddy. Since the seasonal pycnocline prevents oxygen in the SOM from escaping to the atmosphere, the primary production due to the passage of the cyclonic eddy can be estimated from the oxygen increase in the SOM. Such estimation is underway as a step toward characterization of the primary productivity associated with mesoscale disturbances of various types and sizes in the subtropical North Pacific, which could be extended to the global subtropical ocean.

23


23

Mean dynamic height in boundary currents: an evaluation of observations from the East Auckland CurrentBowen, Melissa

Melissa Bowen, GNS Science, New ZealandPhilip Sutton, National Institute of Water and Atmospheric Research, New Zealand

Surface dynamic height, or a level of no motion, is required to accurately estimate geostrophic transport from hydrographic sections or surface geostrophic velocities from altimetry. Satellite altimeters provide an estimate of height anomalies, but a mean dynamic topography is also required to estimate absolute flow. Mean dynamic height is usually estimated from climatologies of accumulated hydrographic data and an assumption of a level of no motion. Recently mean dynamic topographies have become available that incorporate satellite gravity and drifting buoy observations. Boundary currents are one of the most challenging places to determine the mean topography: their narrow spatial and rapid time scales are generally undersampled by satellite and in situ observations.

Here we assess how accurately we can measure mean dynamic height in the East Auckland Current off northeastern New Zealand. We subtract altimeter height anomalies from dynamic height estimated from ten hydrographic surveys collected between 1994 and 2004 along a Jason ground track. The mean and variance of the dynamic height are calculated incrementally by adding one survey at a time. We find the variance in the mean is reduced until the seventh survey and then plateaus, suggesting a limitation with the assumption of a fixed level of no motion. Variance is also reduced by subtracting the spatial mean of the altimeter across the current at the time of each hydrographic survey. This spatial mean corresponds to barotropic fluctuations and has amplitudes between +/- 5 cm. We compare several mean dynamic topographies across the section (the mean we have calculated, the mean from climatology, the CLS_09 mean, and the Maximenko et al. (2009) mean) by comparing velocities calculated using each mean with direct velocity measurements in the East Auckland Current.

2424


Analysis of the interannual variability of the Southern Ocean’s mixed-layer temperature and salinity from hydrographic profilesVivier, Frederic

Celine Heuze, LOCEAN-IPSLFrederic Vivier, LOCEAN-IPSL

The Southern Ocean is an important exchange window between the atmosphere and the deep ocean, where large amounts of mode and intermediate waters are formed. Properties of these waters are set in the mixed layer, hence the great interest of examining non-seasonal anomalies of mixed-layer temperature and salinity, dominant processes that generate them, and how they relate to climate modes of the Southern Ocean.

The density of hydrographic profiles has remarkably improved in the Southern Ocean in recent years thanks to the Argo program. We examine to what extent this improved coverage makes it possible to map and analyze the large-scale interannual variability of the mixed-layer properties.

As of 2004/2005, the data coverage appears dense enough to derive monthly maps of the mixed layer at a relatively coarse spatial resolution (5°) with reasonable accuracy.

We examine the consistency of the mapping of hydrographic observations by comparing the halo- and thermo-steric variations of the upper 700m of the ocean with sea surface height anomalies derived from altimetry, with and without correction from the barotropic variability estimated from GRACE data.

Interannual variations of the mixed-layer temperature and salinity are thus analyzed based on the last seven years of data. We examine the impact of dominant climate modes of the Southern Ocean (SAM, ENSO) on these variations. The influence of different forcing mechanisms in accounting for the observed large-scale variability is also investigated. In particular, regarding salinity changes, we assess the impact of non-seasonal changes in precipitation/evaporation and Ekman transport, derived from the ERA-Interim reanalysis.

25


25

Reliability estimates of decadal sea-level trend hindcasts: a case of predictions with uncertain verification dataRogel, Philippe

Philippe Rogel, URA1875/CERFACS/CNRSDavid Salas Y Melia, CNRM/Meteo-FranceBenoit Meyssignac, LEGOS, OMPMelanie Becker, LEGOS, OMPEmilia Sanchez, URA1875/CERFACS/CNRSChristophe Cassou, URA1875/CERFACS/CNRSElodie Fernandez, URA1875/CERFACS/CNRS

Sea level future evolution is of primary importance in several sectors of our societies. For example, the work presented here is part of a French project aimed at evaluating impacts on coastal erosion in several French regions, including overseas (see http://www.anr-cecile.fr/). Reliability of projections for the next decades (also called “near term” in the CMIP5 exercice) must therefore be evaluated carefully prior to evaluating such impacts. Our focus here is on the reliability of basin scale regional patterns of sea level projections, beyond the global lont-term trend over the last 50 years.

The primary data used is the set of retrospective decadal hindcast ensembles produced in the framework of CMIP5 with the CNRM-CM5 model (Sanchez et al., 2011), initialised with ocean reanalyses obtained from ECMWF. One particularity of this data set, along with the accompanying 20th Century historical simulation ensemble including all forcings, is that they all have been performed with the same ocean model component. The problem arises with observational datasets for forecast verification. Here, most sources of sea level are used to assess hindcasts, either direct or indirect observations: accurate continuous quality controlled tide gauge records, several reconstructions based on these, ocean reanalyses of temperature and salinity, precise altimetry since 1992.

In this poster, we show different estimates of reliability of decadal trend hindcasts, and the gain of initialised over uninitialized historical forced hindcasts. This is assessed with indirect observations, which are available continuously and globally, although they bear uncertainties in some regions. Using direct observations, we show that calibration using tide gauges improves hindcast reliability over the altimetry era. Finally, we investigate how such reliability diagrams can be used for projections up to three decades ahead.

2626


Temporal Variability of Eddy Kinetic Energy and Eddy StatisticsBiri, Stavroula

Stavroula Biri, Institute fuer Meereskunde (IFM)Detlef Stammer, Center fuer Erdsystemforschung und Nachhaltigkeit (CEN)

We study temporal changes of the ocean’s eddy variability and underlying dynamical causes. The goal is specifically to investigate to which extent the eddy field is reacting to changes in the wind forcing. To examine this effect we used a continuous time series (1993-2011) of Sea Surface Height anomalies (TOPEX/Poseidon, Jason-1, Jason-2), analyzed eddy kinetic energy fields. Decadal changes in eddy variability are identified in terms of normalized (with respect to the mean variance) temporal trends for SSH and EKE and are compared with changes in the NCEP reanalysis daily averages of 10m wind speed components. Results are compared also with 30 year-long SSH and EKE timeseries obtained from the eddy-resolving global circulation model STORM, driven by NCEP forcing and by climatological winds, respectively.

Our results suggest that EKE is increasing over the last 19 years in most parts of the world ocean. However, the distribution of temporal trends in SSH variability is quiet different over extended regions of the equatorial Pacific, the North Atlantic and the southern ocean where negative trends are acquainted. The time series of wind stress and EKE suggest that stronger winds cause an increase in eddy activity with a lag of 2-3 years on a global scale. However, time series of several regions suggest that the ocean responce to an increase in the wind field is not uniform for all regions. STORM results are in very good agreement with altimeter results. The lag of maximum cross correlation of the wind speed and 30 year EKE derived from a eddy-resolution global model STORM are mostly positive but quite variable.

Further investigation is needed to quantify how much of the energy derived from the wind is responsible for the generation of eddies, through which processes, and how this energy is dissipated again subsequently.

27


27

Vertical motion estimated from combined ARGO and altimetry observationsPascual, Ananda

Ananda Pascual, IMEDEA(CSIC-UIB)Simón Ruiz, IMEDEA(CSIC-UIB)Bruno Buongiorno Nardelli, CNRStéphanie Guinehut, CLSYann Drillet, Mercator-OceanSandrine Mulet, CLSMarie-Hélene Rio, CLSFrancesco BignamiCNRGilles Larnicol, CLSJoaquín Tintoré, IMEDEA(CSIC-UIB); SOCIB

Vertical motion associated with mesoscale and sub-mesoscale oceanic features is of fundamental importance for the exchanges of heat, fresh water and biogeochemical tracers between the surface and the ocean interior. Unfortunately, direct measurements of the vertical velocity are difficult to obtain for usual values (order 10’s m/day). This is due, on one hand, to the few current measurements available, and, on the other hand, to the high error that would result from the computation of the divergence from measured horizontal velocities, that may include significant instrumental errors. It is impossible to use the continuity equation to estimate the vertical velocities from dynamic heights, as the geostrophic velocities are non-divergent by definition. Various indirect methodologies have thus been proposed to estimate vertical velocity from observed density and geostrophic velocity fields. The most used technique is based on the solution of the quasi-geostrophic (QG) Omega equation.

In the frame of the MESCLA project, a R&D proposal of the MyOcean european project focused on the estimation and analysis of the vertical exchanges associated with mesoscale, a new dataset of QG vertical velocities is generated. QG omega is integrated from the 3D fields of temperature and salinity derived from the ARMOR3D reanalysis, an observational-based product that combines satellite (SST and altimetry) and in-situ (Argo profiling floats, XBT, CTD and moorings) data. The new QG omega time series have been produced in the Gulf Stream area covering the 1993-2009 period at a monthly temporal resolution on a 1/3 regular grid and from the surface down to 1000-meter depth on 100 vertical levels.

The variability of the vertical velocity field is analysed at different depths through standard analyses including identification of seasonal and interannual signals, trends and Empirical Orthogonal Function decomposition. We extent our work through the comparison with other relevant parameters such as kinetic energy and by investigating the possible relations of the observed signals with ocean productivity. The hypothesis behind these relationships relies on the fact that significant variations in the vertical exchange associated to mesoscale dynamics could impact both the ocean capacity to absorb CO2 and the nutrient availability for phytoplankton growth.

This study is a contribution towards improving our understanding of the net effect of mesoscale variability on water mass formation and transport at global scale, as well as on its impact on the biochemical tracer redistribution and consequent marine ecosystem response.

2828


Spatial patterns and variability of near-surface vertical gradients of salinity from historical CTD and Argo float dataMelnichenko, Oleg

Oleg Melnichenko, International Pacific Research Center, School of Ocean and Earth Science and Technology, University of Hawaii, Honolulu, HawaiiNikolai Maximenko, International Pacific Research Center, School of Ocean and Earth Science and Technology, University of Hawaii, Honolulu, HawaiiJames Potemra, Hawaii Institute of Geophysics and Planetology, School of Ocean and Earth Science and Technology, University of Hawaii, HonoluluPeter Hacker, Hawaii Institute of Geophysics and Planetology, School of Ocean and Earth Science and Technology, University of Hawaii, Honolulu

Sea surface salinity (SSS) is an important variable that characterizes the intensity of the marine hydrological cycle [US CLIVAR Salinity Working group, 2008]. The Aquarius and SMOS satellite missions are providing, for the first time, global repeat observations of SSS with space resolution and frequency not accessible by other components of the ocean observing system. Among these components, the Argo float array is the most compatible due to its continuous global coverage. Yet, Argo float measurements are limited to layers at and below 5 m depth, thus leaving the most active near-surface ocean layer unobserved. As a step towards a synergy between the satellite and sea-based observations, we analyze near-surface vertical gradients of salinity in historical CTD and Argo float data.

This way, to characterize salinity differences in the uppermost ocean layer and their relation to subsurface stratification, we analyze open ocean data of high-resolution CTD profiles collected in the World Ocean Database 2009. Globally, the mean value and standard deviation of the difference between salinity at 5 m depth and SSS do not exceed 0.03 psu and 0.2 psu, respectively. At the same time, the probability distribution of this difference is strongly skewed towards positive values due to events of anomalously low SSS. Using the statistics, gained from the analysis of historical CTD casts, the Argo float data are then utilized to reconstruct seasonal maps of probability of appearance of a complex vertical structure of salinity in the near-surface layer. The areas of high probability indicate the areas where the Aquarius and SMOS satellite missions are expected to add fundamentally new information for climate and ocean research. Alternatively, the areas of low probability indicate the areas most suitable for calibration and validation of the satellite data. A struggle between precipitation and vertical mixing, which appears to be responsible for the observed evolution of the complexity of the near-surface salinity structure, is also discussed on a seasonal basis.

29


29

Impact of Argo Data in a Data Assimilation System for the North Pacific AreaOgawa, Koji

Koji Ogawa, Meteorological Research Institute/Japan Meteorological AgencyNorihisa Usui, Meteorological Research Institute/ Japan Meteorological AgencyYosuke Fujii, Meteorological Research Institute/ Japan Meteorological AgencyTakahiro Toyoda, Meteorological Research Institute/ Japan Meteorological AgencyTakanori Iwao, Meteorological Research Institute/ Japan Meteorological AgencyMasafumi Kamachi, Meteorological Research Institute/ Japan Meteorological Agency

In the Meteorological Research Institute (MRI), we have been developing an ocean data assimilation system, MRI Multivariate Ocean Variational Estimation System (MOVE system). The MOVE system adopts the three-dimensional variational (3DVAR) method with coupled T-S EOF modal decomposion and Incremental Analysis Update (IAU). The MOVE system assimilates SSH, Temperature, and Salinity observation data into the model. The Ocean General Circulation Model (OGCM) adopted in the assimilation system is the MRI Community Ocean Model (MRI.COM). The system includes three versions, i.e., the global version (MOVE-G), the North Pacific version (MOVE-NP), and the Western North Pacific version (MOVE-WNP).

In this study, we examined an impact of the Argo profile data on the North Pacific version of the MOVE system, MOVE-NP. The domain of MOVE-NP is 15S-65N, 100E-75W. The horizontal resolution is 0.5 degree. We performed six assimilation runs during the period of 2000-2009 using MOVE-NP with the 5-day analysis cycle. The number of the Argo floats assimilated into the model is changed in different assimilation runs: the 100%, 80%, 60%, 40%, 20%, and 0% of Argo floats are used in those assimilation runs, respectively. For each runs, we evaluated the impact of Argo data by estimating the reduction of the Root Mean Square Errors (RMSEs) compared to the runs where all Argo floats are withheld. Here, we use the 20% of Argo floats that are withheld in all assimilation runs except for the run assimilating all floats (100%), as the independent reference data for the purpose of estimating RMSEs. We confirmed that the RMSEs are generally reduced monotonically as the number of the assimilated Argo floats increases, which implies that the information of the floats is adequately reflected in the data assimilation system. We also found that the impact of the Argo float data on the salinity field is larger than that on the temperature field in MOVE-NP. This result is reasonable, because Argo data relatively includes rich information of salinity rather than the other observation data, e.g., altimetry data, CTD data, and so on. For each runs, since we always used the other observation data, which include rich information of temperature rather than that of salinity, we estimated that an impact of temperature is less effective than that of salinity by increasing Argo data.

3030


Quantifying the thermosteric sea level contribution for different oceanic layers at global and basin scales over 1950-2010Llovel, William

William Llovel, JPLIchiro Fukumori, JPLBenoit Meyssignac, LEGOS/CNES

For the last two decades, steric sea level change has been identified as one of the major contributors to the observed sea level rise (Bindoff et al, 2007). This contribution is not steady and varies with time both globally and regionally. Different estimates, based on different analyses and computing strategies of in situ hydrographic data, have found a substantial global ocean warming signal located principally in the upper 500m to 700m depth of the ocean (Lyman et al., 2010), although a deeper ocean signal may also be non-negligible (Purkey and Johnson, 2010). In this study, we analyze the contribution of different oceanic layers to thermosteric sea level changes over the last decades using existing temperature databases. We study the depth and time evolution of thermal expansion and its contribution to sea level. Comparisons between the different databases and outputs of Ocean General Circulation Models are also investigated and discussed.

program & abstracts - university of washington · 4:00 guillaume dencausse, rosemary morrow,...

Documents