APPLICATION OF OCEAN REANALYSIS

TO THE DIAGNOSIS OF TWO PELAGIC SQUID STOCKS

H. Igarashi1, T. Awaji2,3, T. Toyoda2, S. Masuda2, N. Sugiura2, Y. Sasaki2, T. Ichii4, H. Akiyama5,

K. Hatakeyama1, and S. Saitoh6

1Marine-Earth Data and Information Department, JAMSTEC, Yokohama, Japan

2 Frontier Research Center for Global Change, JAMSTEC, Yokohama, Japan

3 Department of Geophysics, Kyoto University, Kyoto, Japan

4 National Research Institute of Far Seas Fisheries, FRA, Yokohama, Japan

5 Seikai National Fisheries Research Institute, FRA, Nagasaki, Japan

6 Laboratory of Marine Bioresearch and Environment Sensing, Hokkaido University, Hakodate, Japan

 

Abstract

Toward the creation of a variety of social benefits in the ocean and climate fields, we have been currently developing a leading-edge 4-dimensional variational (4D-VAR) data assimilation system and enhancing the ability to work as an interactive platform capable of linking ocean/climate studies with biogeochemistry and fishery science up to the level of full descriptions of essential processes. The 4D-VAR approach solves the minimization problem of the model-data misfit while satisfying the model equations using the Lagrange multiplier and thereby it can provide the best possible time-trajectory fit to the observations and create a dynamically self-consistent dataset capable of offering more information and forecast potential on the dynamical state than can be derived from models or data alone (Masuda et al., GRL,2006).

An accurate ocean state estimation is one of the most important factors for fishery stock assessment. In case of neon flying squid (Ommastrephes bartramii) in the North Pacific, the close relationship is found between the interannual variation of the catch per unit effort (CPUE) of the autumn cohort and that in temperature and salinity fields around the 160W thermocline depth. These results suggest that the survival of young squids could be strongly affected by the variation of the time-varying subtropical upper ocean structure. Moreover, the CPUE change in the jumbo flying squid (Dosidicus gigas) in Peruvian waters shows highly correlated with that of Peru current and the coastal upwelling around 20S. We attempt further analysis to gain more knowledge about such relationships by collaboration with fishery scientists on the 4D-VAR DA research platform.

 

---

Number 90 - Session 5

 

Teiji In1, Y. Ishikawa2, S. Shima1, T. Nakayama1, T. Awaji2, T. Kobayashi3, H. Kawamura3, O. Togawa3

1Japan Marine Science Foundation, Mutu, Japan

2Kyoto University, Kyoto, Japan

3Japan Atomic Energy Agency, Tohkai Viledge, Japan

Abstract

 

We have successfully developed a high-resolution coastal forecasting/nowcasting system focusing on the region near Rokkasho village, which is located in the northernmost part of Honshu Island of Japan. In this area, an important national project toward the implementation of the next-generation nuclear power station in Japan is now underway, so that the good monitoring and forecasting system of coastal circulation is urgently required in the aspect of social benefits and security. Actually, the oceanographic condition in the vicinity of this area is very complicated in association with the geographic features. In fact, the circulation in the target area is controlled primarily by the seasonally-varying Tsugaru current flowing from the Japan Sea and the Oyashio cold current which is the western boundary current in the subarctic North Pacific. Previous observations show that the seasonal variation in this region is characterized by the two different modes: the coastal mode in a cold season when the near-shore circulation takes a straight-path flow along the Japan coast, and the gyre mode in a warm season when swirl-like circulation appears (e.g., Conlon, 1982). When attempting the accurate nowacast/forecast of such complicated regional circulations, a downscaling approach is an effective means since a 4-dimensional variational data assimilation product that covers the larger model domain with a coarse resolution is used for the open boundary conditions suitable for the high-resolution coastal model with horizontal grid sizes of about 1.5 km in both directions. Consequently, our forecasting result demonstrates the good reproduction of both modes in a hindcast sense.

 

---

Number 100 - Session 4

RECONSTRUCTION OF SURFACE MESOSCALE VELOCITIES FROM A

SINGLE SEA SURFACE TEMPERATURE SNAPSHOT

J. IsernFontanet1, B. Chapron1 , P. Klein2 ,G. Lapeyre3, F. Collard4 , J. Johanessen5, V.K. Kudryavtsev5

1Ifremer, Laboratoire d'Océanographie Spatiale, Plouzané, France

2Ifremer, Laboratoire de Physique des Océans, Plouzané, France

3Laboratoire de Météorologie Dynamique/IPSL/ENS/CNRS, Paris, France

4CLS, Radar division, Plouzané, France

5Nansen Environmental and Remote Sensing Center, Bergen, Norway

 

Abstract

 

Recent advances in the understanding of the ocean dynamics in the upper layers have allowed to develop a methodology for the estimation of mesoscale velocities from remotely sensed Sea Surface Temperatures (SST). The proposed methodology does not follow standard sequential temporal analysis but follows the application of an effective version of the Surface QuasiGeostrophic (SQG) theory. Accordingly, under favourable environmental conditions, the implementation for this methodology is simple and robust, and most importantly, solely requires a single SST image. First, surface low-resolution (between 100 and 350 km) velocities have been derived from microwave SST images and, then, compared to geostrophic velocities derived from altimetric maps. Results have shown a remarkable coincidence between both velocity fields. Second, surface high-resolution (smaller than 100 km) surface velocities and vorticity have been derived from infrared Brightness Temperatures (BT) snapshots. Although a direct validation of these velocities have not been possible yet, indirect comparisons with SAR images show remarkable coincidences. These results are very encouraging and strongly invite to consider the systematic use of this approach to derive surface velocities from remotely sensed SST and to complement existing altimetric observations.

 

 

 

 

---

Number 101 - Session 4

 

3D RECONSTRUCTION OF OCEANIC MESOSCALE CURRENTS FROM

SURFACE INFORMATION

J. IsernFontanet1, G. Lapeyre2, P. Klein3, B. Chapron1, M.W. Hecht4

1Ifremer, Laboratoire d'Océanographie Spatiale, Plouzané, France

2Laboratoire de Météorologie Dynamique/IPSL/ENS/CNRS, Paris, France

3Ifremer, Laboratoire de Physique des Océans, Plouzané, France

4Los Alamos National Laboratory, New Mexico, USA

Abstract

The ability to reconstruct the 3D dynamics of the ocean by an effective version of Surface QuasiGeostrophy (eSQG) is examined. Using the fact that surface density plays an analogous role as interior Potential Vorticity (PV), the eSQG method consists in inverting the QG PV generated by sea surface density only. We also make the extra assumption that Sea Surface Temperature (SST) anomalies fully represent surface density anomalies. This approach requires a single snapshot of SST and the setup of two parameters: the mean Brunt-Väisälä frequency and a parameter that determines the energy level at the ocean surface. The validity of this approach is tested using an OGCM simulation representing the North Atlantic in winter. It is shown that the method is quite successful in reconstructing the velocity field at the ocean surface for mesoscales (between 30 km and 300 km). The eSQG framework can also be applied to reconstruct subsurface fields using surface information. Results show that the reconstruction of velocities and vorticity from surface fields are reasonably good for the upper 500 meters and that the success of the method mainly depends on the quality of the SST as a proxy of the density anomaly at the base of the Mixed Layer (ML). This situation happens after a ML deepening period. Therefore, the ideal situation for the application of this method would be after strong wind events.

 

 

 

 

---

Number 80 - Session 5

 

IMPACT OF 4D-VAR ASSIMILATION PRODUCTS ON BIO-GEOCHEMICAL SIMULATION

Y. Ishikawa1, T. Awaji1,2, Hiromichi Igarashi2, Shuhei Masuda2,

Nozomi Sugiura2, Takahiro Toyoda2, Yuji Sasaki2

1Graduate School of Science, Kyoto University, Kyoto, Japan

2 Frontier Research Center for Global Change, JMSTEC, Yokohama, Japan

 

Abstract

 

Accurate descrtipitions of bio-geochemical fields and their variabilities are quite important for the better monitoring of marine environment and its prediction. Toward this goal, the marine ecosystem modeling has gathered inreasing interest in the past decade and has been energetically developed to the level capable of representing the basic bio-geochemical processes in actual oceans (e.g. Kishi et al., 2007). One interesting result of recent modeling studies is that the behaviors of ecosystem models are strongly dependent on the background physical state of temperature, mixed layer depth and current fields and so on. Hence realistic representation of the oceanic physical field is considered as one of the critical issues in simulating realistic bio-geochemical processes. Recent reanalysis datasets obtained by a 4-dimensional data assimilation system are expected to provide useful background data for this challenge. In particular, the optimal synthesis of observational data available and the current state-of-the art ocean general circulation models (OGCMs) by an adjoint method has the advantage of creating dynamically consistent reanalysis datasets. Such products are more suitable for the accurate identification and characterization of transport processes of bio-geochemical tracers than those obtained from other methods.

In this study, as a first step toward an innovative assimilation system that can integrate both physical and bio-geochemical data using the adjoint approach, we take up the challenge of coupling the lower trophic level ecological model, "NEMURO" (Kishi et al., 2007), with an OGCM to estimate 3-dimensional structures of biological and geochemical fields. In doing so, we used the reanalysis dataset derived from a 4-dimensional variational ocean data assimilation experiment (Masuda et al. 2006) as the background physical data. The assimilated elements in their experiment are mainly temperature and salinity from the Fleet Numerical Meteorology and Oceanography Center (FNMOC) Dataset, OISST values, and sea-surface dynamic-height anomaly data derived from TOPEX/Poseidon altimetry, as well as the monthly mean temperature and salinity field of the World Ocean Database 2001 (WOD2001). The OGCM used employs several sophisticated parameterization schemes to better reproduce the ocean mixed layer process and the associated advection processes. The OGCM covers a global ocean with a horizontal resolution of 1 degree in both latitude and longitude, with 36 vertical levels.

Using the ecological parameters in the northern North Pacific (Kishi et al., 2007, Yoshie et al., 2007), the coupling experiments were performed for two cases in which the reanalysis dataset is used or not as the background physical data, respectively. The comparison between these two cases was made focusing on plankton biomasses and those variabilities. The result showed the significant difference in the biological and geochemical field and suggested the efficiency of the reanalysis dataset for the ecological state estimation.

---

Number 36 - Session 2

 

Operational use of regional ocean data assimilation system

Shiro Ishizaki1, N. Usui2, H. Tsujino2, T. Umeda1, and M. Kamachi2
1Japan Meteorological Agency, Tokyo, Japan
2 Meteorological Research Institute, Tsukuba, Japan

 

Abstract

JMA started operational use of new ocean analysis / forecasting system for western north Pacific in March 008. In this system, we use an ocean data assimilation and prediction system (MOVE/MRI.COM-WNP), developed by JMA/MRI. It consists of a dynamical model and data assimilation system.

We use the Meteorological Research Institute Community Ocean Model (MRI.COM) for the dynamical model. The model domain spans from 117E to 160W zonally and from 15N to 65N meridionally. The horizontal resolution is variable: it is 1/10O around Japan. There are 54 layers in vertical. The model is driven by wind stress and heat flux from the JMA's operational Climate Data Assimilation System (JCDAS) in analysis run and forced by the result of the climate forecasting model in forecasting run.

For assimilation system, we use the western North Pacific version of the Meteorological Research Institute Multivariate Ocean Variational Estimation (MOVE-WNP) system. The analysis scheme adopted in the MOVE system is a multivariate three-dimensional variational (3DVAR) analysis scheme with vertical coupled temperature-salinity Empirical Orthogonal Function (EOF) modal decomposition. The MOVE system assimilates in situ temperature and salinity profiles, and Sea Surface Height anomaly (SSHA) from satellite altimeter into the dynamical model. In the operational system, in situ temperature and salinity data are obtained from Global Telecommunication Network, and supplied directly by domestic organizations. The SSHA data is the along-track data from the Jason-1 and ENVISAT. In addition, daily global sea surface temperature analysis (MGDSST) is also used for assimilation.

The assimilation run is implemented every five days, and the forecasting period of prediction is one month. This system reproduces current ocean state well, and provides good forecast in the seas around Japan. Results of hindcast experiment using this system will be shown and discussed.

 

---

Number 110 - Session 5

SOAP-3: A NEW OPERATIONAL OCEANIC FORECAST SYSTEM FOR THE FRENCH NAVY

Didier JOURDAN

SHOM, 42 avenue G. Coriolis, 31057 Toulouse, France

 

Abstract

 

From the mid-90's, SHOM (www.shom.fr) has been conducting research and development in the framework of the SOAP program (System for Operational Analysis and Prediction) to support French Navy activities. This incremental program has been generating successive prototypes of increasing complexity and growing end user applications.

Primarily designed and developed to support anti-submarine warfare at operative level (description of the synoptic scale), capabilities of the SOAP System focus on forecasting oceanic mesoscale in deep waters to get relevant "primary information" for water mass analysis and acoustic products computation. With changing missions -overseas military operation- and growing involvement into sea policy regulation activities, French navy requirements are now extending to the costal domain and exceed acoustics applications.

In the meantime, operational ocean forecast systems have been growing in number through european and international initiatives such as GMES and GODAE, increasing and improving oceanic modeling offer.

In this context, the new version of the SOAP System (SOAP-3) is scheduled to enter service early 2009.

In the poster, we propose to quickly introduce SOAP prototypes as an illustration of the progress in operational forecasting systems throughout the GODAE period.

SOAP-3 will then be presented in more details and improvements with respect to the current operational version will be highlighted such as a larger and homogeneous support capacity and an extended offer of products. Examples on how military products are elaborated will be described with emphasis to the issue of bringing down the modeling results to products relevant for end-users. Samples of products sent out to the navy ships ranging from classical iso depth maps of sound channel interfaces to high added-value products will be shown.

Results on comparisons between forecasted and observed products will also be presented as an indicator of the system's skill.

SOAP-3 capacity to handle "multi-model" and "multi-scale" (cascade of nested high resolution models) production will be discussed as an answer to take advantage of the emergence of numerous and various ocean model outputs and to anticipate coming needs for regional/coastal products. Emphasis will be made to the links and the use of Mercator Ocean information as a "pathfinder" of GMES Marine Core Service.

 

---

Number 107 - Session 3

 

M. Juza1, T. Penduff2, B. Barnier2

1CNES-INSU-LEGI, Grenoble, France

2CNRS-LEGI, Grenoble, France

Abstract

The ARGO global array of profiling floats has been sampling the upper 2 kilometers of the global ocean for several years now. The interest of these global measurements for climate monitoring and research is certain, but their accuracy in depicting the climate-relevant spatial and temporal variability of mixed layer heat/salt contents remains unclear.

In this study, we make use of a DRAKKAR 1/4° global ocean/sea-ice 50-year simulation to assess the sampling error of the existing ARGO array at global scale within 30°x30° monthly bins in terms of mixed layer depth (MLD), temperature (MLT), salinity (MLS), heat and salt contents. Statistical regional estimates of these time-varying sampling errors are obtained by comparing full model fields with ARGO-like model-derived T/S profiles.

The sampling errors derived from this eddy-admitting solution depend on both the number and distribution of drifters in each bin. Monthly MLD/MLT/MLS sampling error maps are presented at global scale. They reveal non-zero annually-averaged values reaching +/- 10m/2°C/0.4 over certain regions, with maximum values reaching +/- 100m/5°C/1.0 at certain periods of the year. Implications for the design of future in-situ observing arrays are discussed.

 

---

Number 108 - Session 3

M. Juza1, T. Penduff2, B. Barnier2

1CNES-INSU-LEGI, Grenoble, France

2CNRS-LEGI, Grenoble, France

Abstract

The international DRAKKAR group is building a hierarchy of ocean/sea-ice models to simulate and study the dynamical processes involved in the oceanic variability and scale interactions over the last 50 years. Global simulations (2°, 1/2°, and 1/4° resolution) and Atlantic/Nordic Seas simulations (1/4°) have been performed, driven through bulk formulae by a hybrid (reanalysed and observed) surface forcing function built over this period. The 1/4° global configuration is currently being used for scientific research, and in collaboration with Mercator-Océan for ocean prediction and reanalyses.

In order to guide physical investigations, to characterise the structure of model in-situ biases, and to assess the impact of numerical and physical choices, DRAKKAR simulations are quantitatively evaluated against most temperature/salinity profiles available since 1956 (T,S ENACT/ENSEMBLES: CTD, TAO, ARGO, XBT). Model outputs are first sub-sampled like observed profiles to build model-derived counterparts. Real and synthetic observations and more integral quantities (mixed layer depth evolution, heat/salt contents, etc.) are then compared over various regions, periods, and depth ranges. These results guide model improvements, help identify their sensitivity to various parameters (atmospheric forcing, resolution, subgrid-scale parameterizations, etc), and provide the community of DRAKKAR users with quantitative skill assessment. This study presents our approach, our metrics, and the realism of DRAKKAR simulations.

 

---

Number 69 - Session 3

 

M. Kamachi1, N. Usui1, S. Ishizaki2, Y. Kanno2, and T. Kuragano2

1 Meteorological Research Institute, Tsukuba, Japan

2 Japan Meteorological Agency, Tokyo, Japan

Abstract

An ocean data assimilation system, MOVE/MRI.COM-WNP, has been operated in Japan Meteorological Agency (JMA) from March, 2008. The purposes of the system are understanding ocean variability in the western North pacific as a local response to a global climate change with assimilated four-dimensional data sets, nowcasting and forecasting of ocean states, and a contribution to the GODAE project.

Before starting the operation, JMA headquarters, four Marine Observatories and Meteorological Research Institute have cooperated to promote a project of validation of the products of assimilation and prediction of MOVE/MRI.COM-WNP for two years.

Fundamental idea about the internal metrics in the GODAE intercomparison project has been reported (Le Provost et al., 2002). We adopted the metrics in the North Pacific according to the paper. Comparisons with observation data (e.g., temperature, velocity, sea surface height, Kuroshio axis and water masses) for assimilation and prediction are shown. Future directions of improvement of the system suggested from the validation project are also reported.

 

---

Number 68 - Session 4

 

M. Kamachi1, Y. Fujii1, and S. Ishizaki2

1Meteorological Research Institute, Tsukuba, Japan

2Japan Meteorological Agency, Tokyo, Japan

Abstract

Variational data assimilation method can adopt nonlinear constraints for improving representation of target phenomena more realistically. Two kinds of nonlinear constraints, not previously studied in oceanography, have been adopted in a three-dimensional oceanic variational data assimilation system MOVE/MRI.COM. One is the constraint for the variational Quality Control (var-QC) procedure and the other is used to avoid density and temperature inversions.

Estimation of the large heat content anomaly in the upper ocean related to El Nino and La Nina phenomena is improved with the var-QC. For example, it prevents unusual but correct observation data on the thermocline deepening in the 1997/98 El Nino from being ignored. As a result, it improves the temperature field estimation in the eastern equatorial Pacific.

The constraint for avoiding inversions prevents the low salinity layer at the surface and the barrier layer in the eastern equatorial Pacific in the El Nino period from being destroyed by the convective adjustment procedure performed after minimizing the cost function. Incorporating nonlinear constraints in variational analyses is thus a strong candidate for increasing the accuracy of analysis.

 

---

Number 161 - Session 2

 

J. Karstensen1, U. Send2, R. Lampitt3, M. Villagarcia4, G. Meinecke5, D. Wallace1, J. Fischer1, M. Visbeck1, S. Østerhus6

1Leibniz-Institute for Marine Sciences, Kiel, Germany

2Scripps Institution for Oceanography, La Jolla, USA

3National Oceanographic Centre, Southampton, UK

4Intituto de Sciencas Marinas, Gran Canaria, Spain

5Marum, Bremen, Germany

6Bjerknes Centre for Climate Research and Geophysical Institute, Bergen, Norway

Abstract

Physical and biogeochemical time series data from the ocean interior is of importance in the context of data assimilation but moreover for the validation of ocean model products. Time series observatories provide data from critical regions and on all relevant time scales ? from small scale events to long term climate change. As such, time series observatories complement the other "two" major platforms used for the global ocean observing system: Satellites and Argo profiling floats. Satellites can provide only access to surface ocean physical and limited biogeochemical data (e.g. no carbon relevant observations); while Argo profiling floats data provide a global view of the temperature and salinity distribution but not necessarily in critical regions at critical times e.g. in the centre of a deep convection area in winter.

Here we present an overview about the status of a set of North Atlantic open ocean time series observatories which cover the whole North Atlantic from the tropics to the Arctic Ocean. The observatories have been established and organized under national and European projects (ANIMATE, MERSEA, THOR, EuroSITES). The observatory site locations have been carefully selected at locations representative of particular physical and biogeochemical regimes. All sites discussed here have been in operation for several years up to decades and are expected to remain as sustained observatories for the foreseeable future.

The time series data is most suitable to derive indicators to be used for ocean model validation, e.g. the seasonal evolution of the mixed layer. It is however rather surprising that the data has so far not been systematically used for model data assessment. One main reason for this, we suppose, is the difficulty in accessing the data. As heterogeneous as the initial motivations (and funding) of the observatories was the data dissemination policy. All time series sites discussed here are now part of recent European projects (MERSEA, EuroSITES, THOR and ESONET) and this has put forward the process in harmonization of, access to, and comparability of the data collected at the sites.

 

---

Number 66 - Session 4

 

T. Kobayashi, T. Nakamura, and N. Ogita

IORGC-JAMSTEC, Yokosuka, Japan

Abstract

To estimate global surface and subsurface velocities is another goal of the array of numerous profiling floats, Argo. However, until now there are only a few studies about the velocity field estimations from the Argo data. One of the reasons is that Argo trajectory data (including float positions fixed during sea surface drifting) are not as common as the profile data which consist of temperature and salinity profiles. However, it is mainly attributed to the fact that Argo trajectory data are far from completion. Now we do not have any protocol for quality-control (QC) of Argo position data, and then, many suspicious float positions make it more troublesome to use the trajectory data and to estimate the surface and subsurface velocities from them.

Here, we introduce an automatic QC method of Argo float positions. The method discards suspicious position data based on the float's speed estimated from the surface trajectory as follows. Considering a segment composed by two temporal-continuous positions, one position, at least, is identified suspicious if float speed along the segment is estimated at 3 m/sec or faster. The suspicious position is determined by the relation among the segment and the back/forth positions of the float trajectory. In case that the distance between the positions is less than the error length determined by Argos position errors of them, both positions are considered to be acceptable. The method gives us fairly reasonable QC results which are comparable with those by visual inspection of experts, and several percents of position data are identified "bad" in average. An execution program of it has been prepared from PARC-JAMSTEC web-site so as that this method will be tried widely (PARC-JAMSTEC: http://www.jamstec.go.jp/ARGORC/location_top.html).

This method has been suggested as (one of candidates for) the standard QC method to Trajectory Working Group in International Argo community. We expect that the method will work as a preliminary QC of Argo trajectory data and that it will be succeeded by a more sophisticated inter- and extrapolating scheme (i.e., "delayed-mode QC") to estimate actual float movements and locations where a float arrives at and departs from sea surface. Since these estimations can be distorted largely by suspicious positions in raw data. Another purpose of our suggestion of the method is to homogenize quality of Argo trajectory data by a single QC scheme. Since it will confirm homogeneity of velocity atlases and the trajectory data seem too complicated to be QCed only by a set of criteria.

 

---

Number 70 - Session 4

T. Kobayashi1, K. Mizuno1, and T. Suga3

1IORGC-JAMSTEC, Yokosuka, Japan

3IORGC-JAMSTEC, Yokosuka, Japan / Tohoku Univ., Sendai, Japan

Abstract

The ocean has an important role in climate variability and change. The ocean's heat capacity is about 100 times larger than that of the atmosphere, and ocean's net heat uptake since 1960 is around 20 times greater than that of the atmosphere. The large amount of heat, which has been mainly stored in the upper layers of ocean, plays a crucial role in climate change, in particular variations on seasonal and decadal time scale. In the subtropical regions of the south Indian Ocean, decadal changes of water-masses in the thermocline and intermediate layers, Subantarctic Mode Water (SAMW) and Antarctic Intermediate Water (AAIW), have been found by trans-Indian hydrographic observations (about 32°S). The details of variations have been closely discussed, but it is not identified whether these variations are oscillations or trends. Here, we will attempt to examine time series of water properties at 32°S, which are obtained with Optimal Interpolation (OI) method from a historical dataset (Indian Ocean HydroBase) and recent observations, Argo.

Variations of SAMW and AAIW are examined based on time series of properties on isopycnal surfaces for every 0.1s? between 26.2 and 27.5s?. First, "climatological" fields of pressure, potential temperature, salinity, and layer thickness (0.1s? interval) are estimated from all data with OI method. Anomalies from the climatologies at each location are estimated and then yearly anomaly fields along 32°S are estimated. Here, Gaussian function is used for the data covariance in OI calculations, its correlation radii are 5° for zonally and 3° for meridianally. The decadal analysis (1950-present) uses temporal scale of 3 years because of less dense data distribution, which will screens out variations with shorter scales than about 5 years. After 2000 while the Argo array provides abundant data, seasonal analysis is additionally done with time scale of 0.15 year.

The Argo data provides us with a remarkable result that the seasonal variations are detectable even at the great depth. The variations are due to perturbations in the depths of isopycnal surfaces there and their influences on the water properties seems limited to the layers above the core of SAMW (about 26.8s?). The features also provide a basis that we can discuss the decadal variations of SAMW/AAIW properties estimated from historical data.

On the isopycnals above 26.7s?, more saline waters are distributed around 1960 throughout the section. They are replaced with less saline thermocline waters in the later half of 1980s after 5-10 years oscillations with the amplitude exceeding 0.1 PSU. Then its low salinity retreats and saliner SAMW occupies again after 2000; no long term trend of SAMW water properties is identified from the analysis. It is worth describing that SAMW is much thicker around 1965 and then thinner to show the minimum around 1980. After then SAMW becomes as thick as it was, but the core density seems to become less dense slightly. On the other hands, the variations of AAIW are very different from those of SAMW. In the whole section, AAIW becomes less saline gradually with a smaller peak of saline AAIW in 1980. The salinity changes of AAIW during 40 years (from 1960s to 2000s) are about 0.1 PSU in 26.9s? and 0.04 PSU in 27.3s?, respectively. These features of SAMW seem consistent with recent studies with numerical models, which demonstrated that the properties of the both water-masses may be changed toward less salinities with very large variations in time due to natural variations and global warming effects.

 

---

Number 64 - Session 2

ARGO-JAMSTEC
- ACTIVITIES OF JAMSTEC FOR ARGO PROJECT -

T. Suga1, 2, M. Hirano1, S. Hosoda1, N. Iwasaka1, 3, T. Kobayashi1, T. Nakamura1, N. Ogita1, E. Oka1,4, K. Sato1, H. Ueno1, S. Asai5, T. Idai5, N. Matsuo5, H. Nakajima5, T. Ohira5, M. Yokota5, N. Shikama1, and K. Mizuno1

1IORGC-JAMSTEC, Yokosuka, Japan
2Tohoku Univ., Sendai, Japan
3Tokyo Univ. Marine Science and Technology, Tokyo, Japan
4Ocean Research Institute, Univ. Tokyo, Japan
5Marine Works Japan Co. Ltd., Yokohama, Japan

Abstract

The Argo Project, underway since 2000, aims to maintain a network of 3000 Argo floats that provide real time monitoring of temperature and salinity of the global upper ocean. In Japan, the Japan Agency for Marine-Earth Science and Technology (JAMSTEC), in cooperation with other agencies, is responsible for Argo. This is the summary of our activities.

JAMSTEC is one of the largest Principal Investigators in Argo: Since 2000 about 650 floats, which are about 90% of the Japanese Argo floats, have been deployed in the North and South Pacific, the Indian, and the Southern Oceans. Recently we deployed about 80 floats every year by more than 10 ships in close cooperation with other Japanese agencies, universities and a high school. We will make an effort to continue the float deployment in almost the same scale as now in the future.

We conduct several steps of preparatory work of float deployment at Mutsu Institute for Oceanography of JAMSTEC by technical experts. One of the major steps is to check accuracies of float sensors with a calibration bath and only the floats with sensors which pass our criteria are shipped to be deployed. Another important step is the weight adjustment of floats. Our reliable float measurements are mainly attributed to these steps of work. These technical experiences have helped us to find failure modes of float/sensor hardware, which is one of our important contributions to Argo.

Data management is one of the most important activities for Argo at JAMSTEC. We are operating Japanese Data Assembly Center (DAC) in cooperation with Japan Meteorological Agency and we are responsible for the delayed-mode quality control (DMQC) of all Japanese Argo floats. We have done lots of contributions to enhance its performance, especially to provide high-quality reference datasets, in the Argo Project and float data DMQCed by JAMSTEC are fairly consistent with measurements of nearby shipboard CTD. We also operate Pacific Argo Regional Center (PARC) in cooperation with IPRC (USA) and CSIRO (Australia). PARC has important functions to confirm the consistency of Argo data in the whole Pacific sector.

We have carried out lots of scientific researches mainly on water ventilations and air-sea interactions in the North Pacific and their variations. Recently, we found decadal changes of sea surface salinities probably due to reinforcement of atmospheric water-vapor circulation. Products associated with several scientific results (e.g., monthly atlases of temperature/salinity, geostrophic current, water properties in mixed layer) are provided via Internet as an activity of Japanese DAC and PARC.

It is worth explaining our efforts to educate younger peoples with lectures, TV broadcast, and actual float deployments. This is one of our purposes to spread understanding of Argo and global environment.

 

---

Number 9 - Session 5

 

V.H. Kourafalou1, H. Kang1, C. Paris1, C. Hu2, P.J. Hogan3 and O.M. Smedstad4

1Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami, FL, U.S.A.

3Institute for Marine Remote Sensing, University of South Florida

3Naval Research Lab, Stennis Space Center, MS, U.S.A.

4QinetiQ North America, Technology Solutions Group ? PSI, Stennis Space Center, MS, U.S.A.

 

Abstract

The coastal seas around the Florida Keys Reef Tract exhibit complex dynamics resulting from the interaction with offshore flows, namely the Loop Current/Florida Current system and the frontal eddies that interact with the complex reef topography. A nested modeling approach has been employed to ensure the proper representation of such interactions. A high resolution (~900m) application of the HYbrid Coordinate Ocean Model has been developed focusing on the Florida Keys (FKEYS-HYCOM model). In order to downscale the effects of these large scale flows to scales appropriate for the study of reef related processes, GODAE products (North Atlantic and global HYCOM models, the U.S. contribution to GODAE) have been employed to provide boundary conditions to a succession of regional (Gulf of Mexico, GoM-HYCOM) and local (South Florida, SoFLA-HYCOM) models. Eddies that travel along the Loop Current/Florida Current front are known to be an important mechanism for the interaction of nearshore and offshore flows. They enable upwelling in the vicinity of the Reef Tract and they influence transport and recruitment pathways for coral fish larvae, as they carry waters of different properties (such as river-borne low-salinity/nutrient-rich waters from as far as the Mississippi River) and waters containing larvae from upstream sources, or entrained from nearby spawning grounds. As such, they play an important role in the circulation around the Reef Tract and connectivity pathways with the Gulf of Mexico and the Caribbean at large. The FKEYS-HYCOM nested model is able to simulate both mescoscale and sub-mesoscale eddy passages during a targeted 2-year simulation period (2004-2005), forced with high resolution/high frequency atmospheric forcing. Coupling with the ecological population connectivity BOLTS model (BiOphysical Larval Tracking System) allows simulations of larval transport, taking into account not only the dispersion of active physical larvae, but also the interaction of factors influencing larval survival, habitat selection and condition at settlement. Coastal value added products to GODAE include the prediction of changes in the position of the Florida Current front and the evolution of eddies entering the model domain, as shown through comparison to high resolution SeaWiFS data.

 

---

Number 97 - Session 4

 

M. Krysta1, E. Blayo2, E. Cosme1, J. Verron1, A. Vidard3

1LEGI, BP53, 38041 Grenoble Cedex 9, France

2University of Grenoble, Grenoble, France

3INRIA, Grenoble, France

Abstract

Observations may be abundant in nowadays oceanography but they are restricted to some components of the state vector only. Consequently, a need for some prior information on the unobserved variables arises. In the standard 4D-Var it is accounted for via regularising properties of the background error covariance matrix. Secondly, feasibility requirements may result in reduction of the size of the control vector and confine data assimilation correction to a small-size subspace of the original state space. Hence the necessity of an appropriate definition of the control subspace which must be of a small dimension and, at the same time, preserve regularising properties of the background error covariance matrix.

The question addressed in this study goes even further and tackles the problem of adapting the definition of the control subspace to account for additional information gained in the data assimilation procedure. The problem has been studied in the framework of a hybrid approach to data assimilation. Hybrid in this study refers to a method merging an incremental 4D-Var with an equivalent Kalman smoother, both in a reduced rank approximation. The skeleton of the hybrid is thus formed by the 4D-Var enriched with an admixture of the smoother delivering a recipe for the evolution of the error covariance matrix. Its update is made at each transition from one assimilation window to another, at both analysis and forecast step. The analysis update modifies the reduced size error covariance matrix accordingly to the quality of the measurements assimilated into the system. Following the system's trajectory evolution in the forecast step, the basis spanning the control subspace is also adjusted.

A series of OSSEs implementing the hybrid method into a shallow water model in a wind driven double-gyre circulation has been performed. It has been opted for a definition of the reduced-size control subspace reflecting the directions of the largest variability of the system. These directions have been obtained via principal component analysis of several different samples of model trajectory. A number of tests have been performed in various configurations of twin experiments and the circumstances where the hybrid outperforms the standard 4D-Var have been identified. The general conclusion inferred from this study indicates that the propagation of the basis spanning the control subspace is capable of compensating for imperfect initialization of this subspace. Moreover, as the information contained in the observations is gradually being assimilated into the system's trajectory, the control subspace evolving accordingly.