A. The Project
 

I. Introduction

POMME is a research project aimed at understanding the subduction mechanisms of 11-12Cmode water in the northeast Atlantic, how this affects the biological production and the carbon budget of the northeast Atlantic, and at describing the fate of organic matter after subduction. The study area is known to be a sink of carbon (Takahashi et al., 1995) and to present a high export of primary production (Jickells et al., 1996). The coupling of mesoscale dynamical and biological processes is one of the major objectives of the project.

The project includes both a field experiment and a data assimilation/analysis phase. There are also preliminary studies to help define the field experiment. The domain we focus on is a 500 km longitude by 750 km latitude area centered on 41.5N/19W(16 W-22 W/38 N-45 N, see figure 1), and the field experiment is scheduled for a one-year period starting in the autumn 2000 and lasting until the autumn 2001. There will be a more intensive phase in the winter and early spring of 2001. Efforts will be made to transmit a large amount of physical oceanographic data in near-real time in order to feed assimilation schemes and provide preliminary analyses and forecasts. The experiment is therefore also envisioned as a test of real-time oceanography (for GODAE, the "Global Ocean Data Assimilation Experiment"). It will be linked with high resolution modelling efforts.

The area investigated is a transition zone between quite deep late winter mixed layers in the north (reaching about 500 m) and relatively shallow mixed layers (100-150m) in the south. The water to the north is advected to the area by the southern branches of the North Atlantic current, where it experiences a net cooling and buoyancy loss, and has very high nutrient levels in winter. Part of this water then flows southward, at least intermittently. This has been diagnosed from hydrography (McCartney, 1982; Paillet and Arhan,1996b), as well as by inverse modelling (Paillet and Mercier, 1997). There is
also recent evidence from floats of this southward flow (Brügge, 1995; Kevin Speer, personal communication : ARCANE MARVOR floats at 450 meters depth), as well as by surface (15 meters) WOCE drifters. During the southward drift, the late winter mode water gets subducted into the thermocline, where it influences the vertical stratification (low potential vorticity) equatorward to 28N(Paillet and Arhan,1996b).

A water mass formation budget based on the air-sea buoyancy flux climatology suggests that no more than 3-4Sv of water in the density range can be formed, which puts a constraint on the magnitude of its subduction (Speer et al., 1995). The current theory is that the subduction is associated with a change along the flow in the annually-averaged buoyancy budget of the mixed layer, and that subduction happens when there is net buoyancy gain (Marshall et al.,1993; Marshall and Marshall, 1995; Paillet and Arhan, 1996a). The budget includes not only the net buoyancy flux across the air-sea interface, but also a contribution from Ekman convergence of buoyancy transport. Neither the theory nor the observations have actually included or resolved the effects of variability at high frequencies (small spatial scales) or low frequencies.

The winter mixed layer depth discontinuity zone divides the northeast Atlantic basin into two regions; a rather productive one, with a strong spring bloom to the north, associated with a deep winter mixed layer, and a more oligotrophic one to the south, with characteristics similar to those of the subtropical gyre (NABE, Wroblewski et al., 1988). Moreover, the spring bloom and the subduction of mode waters occur during the same period. The exact timing of these processes is therefore crucial to understanding the characteristics of the water masses before they are isolated on a decadal timescale from the influence of the atmosphere. The export production associated with dissolved organic matter is also driven by subduction. Finally, data clearly have shown that this area can be a strong sink of atmospheric CO₂ , mostly during the spring (Rios et al., 1991). Both NABE and other insitu data (Lochte and Pfannkuche, 1987 ; Ducklow and Harris,1993 ; Yoder et al., 1993 ; Karrasch et al., 1996) or modelling studies (McGillicuddy et al., 1995 ; Spall, 1998 ; Lévy et al., 1998 ; Oschlies and Garçon,
1998) show that the highly non-linear coupling between dynamics and biological activity at the mesoscale can be responsible for large variations in tracer distributions and estimates of primary and export production which cannot be neglected, even at basin scale.

On individual meridional sections, the change in the stratification is often quite sharp with the presence of mesoscale fronts and eddies (Pollard et al., 1996). From the eddy energy statistics estimated from altimetry and from drifters, this region is a boundary between a region of high energy levels to the north and one with much lower energy levels to the south. This transition probably has a dynamic origin, but this has not been thoroughly investigated. Despite this strong difference, large (100 km) coherent eddies are observed in both areas, although more frequently to the north than to the south. There is no suggestion of systematic differences between the cyclonic and anticylonic eddies in the domain (Paillet, 1999). The role of these eddies in mode water formation and subduction needs to be investigated. It is not clear whether they contribute to a substantial part of the watermass and property transport. Similar questions are also relevant for estimating biological/production and export fluxes. Besides regional surveys for estimating budgets of carbon and associated tracers (nutrients, oxygen) in the ocean and understanding the main processes at play, there will be efforts to follow one or two coherent eddies (by coherent eddy, we mean an eddy structure that can last 3 to 6 months). Budgets are easier to perform in this situation, as has been shown in previous JGOFS studies, and with the limited resources, it will be better to concentrate on a few structures particularly well-sampled dynamically and biologically (at least, one anticylonic eddy).

The net buoyancy flux is expected to have an interannual variability related to the air-sea fluxes. In the northeast Atlantic, this is to some extent related to the North Atlantic Oscillation (NAO), associated with the strength of the westerlies along 50°N in the central North Atlantic in winter. In recent years, it has witnessed dramatic changes, from a high index situation (large westerlies) through 1995, to a situation of weaker westerlies since then. The net air-sea heat and freshwater fluxes covary with the NAO (Hurrell,1995, 1996). While understanding and monitoring the effect of this variability on the subduction is not the object of the present proposal, we will design the experiment in a domain that has a fairly large meridional extent (at least 700 km) in order to accomodate possible interannual variability in the location of the transition area.
 

II Scientific objectives

The prime objectives are to understand the process of Mode Water subduction at the mesoscale, to understand and quantify the processes responsible for the evolution of its bio-geochemical prop-erties of the modal water and for the export of biogenic matter over a seasonal cycle.

More specifically, the following objectives will be addressed:

. The late winter mixed layer. To map its distribution, and, in particular, the area of largest gradient between the usually deeper layers to the north and the shallower ones to the south; to understand how that distribution is shaped by the air-sea buoyancy fluxes and the mesoscale circulation, and to determine the pre-bloom distribution of bio-geochemical tracers.

. Meso-scales: To understand what controls the physical fluxes (buoyancy, potential vorticity, mass, heat and salt) and the contribution of the meso-scale circulation to those fluxes. To identify and quantify the physical and biological processes controling the bio-geochemical fluxes within eddies.

. Subduction and bloom. To understand the coupling between restratification at the end of winter, Mode Water subduction and the spring bloom.

. Export from the upper ocean : To describe the dynamical and biological cycles responsible for the export of organic matter and of tracers to the thermocline (initial conditions and outcome), and to the deep ocean.

. Transformation of organic matter. To quantify the dynamical and biological processes involved in the bacterial and hetoerotrophic loops, which regulate the evolution of tracers in the subducted mode waters.

Fullfilling those objectives will, furthermore, contribute to the quantification of physical (buoyancy, potential vorticity, mass, heat and salt), as well as biogeochemical fluxes over the annual cycle. This will also contribute to the understanding of the differences in the coupled biological-physical system between the northern and southern parts of the domain.
 

III. Field experiment

The POMME field experiment will require specific measurements for the following topics:

1 The formation of the mode water: Air-sea heat and freshwater fluxes; the temperature and salinity evolution of the mixed layer; deepening of the mixed layer; nutrient distributions and biological activity during the winter.

2 The subduction of the mode water: Two or three dimensional surveys of the circulation, the water masses, and the vertical stratification; tracer studies to investigate age and dispersion of the water.

3 The net production of the system: This will require investigating the seasonal cycle of the carbon budget, which involves production (new and regenerated) and respiration, the particle flux, the dis-solved organic matter pool and its biodegradation by bacteria, the air-sea CO₂ fluxes, DMSp forma-tion and DMS exchanges.

This will require investigating both an extended domain on both sides of the transition region, and particular eddy structures, at least an anticyclonic quasi-coherent structure, but if possible also a cyclonic structure. Statistics on the distribution of eddies and their size (roughly 100 km), and on the interannual variability of the position of the transition zone, set the domain size to be at least 500 km longitude by 750 km latitude, which is the size retained for the experiment (figure 1). The measurements will be obtained with a combination of fixed Eulerian and Lagrangian instruments, complemented by ship surveys, with the more extensive of these taking place in the late winter - early spring. The different observations shall be adequately combined during the analysis and data assimilation phase (there will be both near-real time data assimilation, and a reanalysis later on, when all the data will be available) to provide a continuous and realistic mapping at the eddy scale during the course of the seasonal cycle, and in particular in the late winter and spring season.

Air-sea fluxes are needed, both at the large scale and on the meso scales characteristic of the largest oceanic eddystructures. Their absolute accuracy should be near to WOCE standarts, but it is more important for the objectives to have the spatial structure with good accuracy, and to track the time evolution at high frequencies. This requirement can be fullfilled partially with in situ observations of bulk parameters (wind, air temperature, humidity), complemented by the operational forecasts, and by air-sea flux measurements during cruises at least during the winter and spring cruises, as well as by satellite products for incoming radiative fluxes, SST, and sea surface wind stress.

For physical oceanography, the instrumentation requirements might be different for the current and temperature-salinity fields with the objective of sampling the domain on the meso-scales. Roughly, monitoring the currents requires on the order of 100 simultaneous velocity measurements in the domain to constrain the eddy circulation (but not the fronts). This can rely mostly on floats, but mooring data will be interesting to have a complementary Eulerian perspective in a few spots and current profiles during cruises will provide snapshots of the spatial variability at finer scales, in particular near fronts. The temperature and density of the upper ocean shall be sampled at the eddy scale during the cruises, which can be used to initialize numerical simulations. Finer surveys of the upper ocean, both of the density, T, S, current and fluorescence fields shall be carried to understand the 3-D dynamics at the finer scales, in particular in eddies or in the transition region. Time series of the stratification at a few locations are required to provide a time continuity, although it is not necesary to have them at the eddy scale, except in one or two well sampled structures. We expect that the combination of these data with the altimetric measurements will improve the sampling of the eddy-scale variability of the heat content. Part of the approach will be tested a year earlier with POMMIER (experiment taking place in the POMME area in late-1999- 2000; see appendix 1). The evolution of spatially integrated temperature, for example, from a combination of altimetric and tomographic measurements, is required to provide an important comparison with the heat and buoyancy surface forcing.

Most bio-geochemical parameters will be sampled during cruises at different times of the seasonal cycle (at least 3 cruises are required). They shall provide both a meso-scale mapping of instant variables (nutrients, O₂ , DIC, DOC, fluorescence, pigments, bio-optical parameters, zooplankton -OPCT, pCO₂ , particle loading) and an estimate of some major fluxes and kinetics, either at larger scale, or during 1D process study stations (primary production - net, total, regenerated, new; microbial activity, remineralization of organic matter; particle flux). For this, 3 to 4 cruises are required to sample the situation at the most characteristic seasons. At other times, sediment traps are required to provide a time continuity at four sites of the exported matter, whereas a few time series of the evolution of surface pCO₂ , chlorophyll, and if possible other components of the carbon cycle are required to illustrate the evolution in the surface layer where production takes place. This will be complemented by analysis of chlorophyll concentration from satellite ocean color mapping.
 

IV. Analysis phase

This has not yet been thoroughly discussed, and should be finalised in early 2000. However, some specific studies will need to be done, which are briefly outlined here.

Air-sea fluxes
A special effort shall be made to check the accuracy of operational atmospheric models to which most meteorological data will have been transmitted. In particular, the heat and freshwater fluxes of these analyses will be compared with those measured during the winter and spring cruises, and with estimates using bulk aerodynamic formula at the meteorological mooring site. Although this has not yet been decided, one could consider doing special reanalyses in this region in a delayed mode fashion into which carefully checked wind data and other data will be given a high priority (for example, the locally
reanalysed SST field could be used to provide a more accurate local boundary condition to the atmospheric model). It is also possible to envision directly producing hourly wind and wind stress fields from the in situ data.

Circulation
Specific studies of the floats shall be done to investigate the Lagrangian variability at 400 m. This shall be compared with the evolution of the three dimensional distribution of injected tracers, and with age distributions from 3 H/3 He distribution.

Specific studies of the moored Eulerian velocity measurements shall be done to compare the variability in the different parts of the domain, investigate wave propagation, and the evolution of the shear in the upper ocean in relation to the vertical stratification.

All the velocity data (both Eulerian and Lagrangian) shall be combined to map the meso-scale current variability at 400 meters. It will probably be possible to do that at a fairly short time scale on the order of a couple of weeks, at least in the best sampled parts of the domain.

Water mass and stratification
Mapping the temperature, salinity, and density fields shall be done for the better sampled periods before and during the surveys. Higher frequency mapping of mixed layer depth and near surface stratification shall be done based on the available data (in particular from floats). The results of these analyses shall be compared with the temperature retrievals of the acoustic tomographic array integrated along the ray paths. It is not clear whether the data available between the surveys will be sufficient to map the temperature and salinity fields between the surveys, but this shall also be attempted.

Specific studies of the vertical velocities measured by some floats shall be done to understand the regimes of vertical velocity, in particular during the late winter phase of mixed layer deepening and restratification.

The temperature and salinity profiles shall be assimilated at the same time as the satellite altimetric data in near real time in a regional GCM of the north-east Atlantic, as well as in a regional quasi-geostrophic model to provide analyses of the currents and of the eddy field. In a remote mode, a reanalysis of the observations shall be done with a more carefully controlled forcing field, and the velocity data shall also be included to better constrain the model evolution. Both the direct analyses and these later analyses based on an assimilation in an OGCM shall be used to estimate the importance of advection, and of water mass transformation after the subduction. At a later stage, the information brought by the tracers (at least the helium-3/tritium measurements) should be analysed and compared with the analyses.

Specific studies of the finely sampled eddies and fronts will be done based on the cruise surveys to better understand the dynamics and the budget of these structures, in particular, for temperature, salinity, oxygen, and stratification (potential vorticity). This will also involve diagnosing the vertical circulation in these structures, at least during the time of the high resolution surveys.

Bio-geochemical analyses
Mapping of the biogeochemical tracer distributions (oxygen, nutrients, DIC, DOC, pCO2, chlorophyll, pigments, zooplankton - OPCT) at mesoscale shall be done for the different cruises. This mapping will be performed using the data obtained with high horizontal resolution CTD profiling and Seasor/OPCT surveys, as well as with continuous surface measurements from drifters, and this will be compared/correlated with the T, S and density fields. The mesoscale/large scale signals of the tracer distributions will be described. At large scale, mapping of primary and bacterial production will be done based on data obtained during the first legs. Emphasis will be put on describing the links of these distributions with the dynamical environment.

Sediment trap data from moorings shall be used to provide estimates of the mesoscale coupling between surface and sub-surface fluxes, of the meridional gradient and of the regional particulate export production.

Specific analyses shall be performed for each time series station to provide a synoptic view of the behaviour of the trophic system inside and outside different eddies. Emphasis will be put on the different pathways of the organic matter in the food web (regenerated production via the microbial loop vs. new and export production via particles or dissolved organic matter). The ageing of the trophic system in one (or two) eddy and in the surrounding waters shall be more specifically analysed from the different cruises, taking into account the coupling with the dynamical environment.

Large scale gradients of tracers and fluxes on either side of the winter deep mixed layer discontinuity zone shall be estimated, which will provide a description of the whole area, and of the differences between the productive northern region and the more oligotrophic southern one. Emphasis will be put on the timing of the spring bloom and of the subduction of mode waters.

Assimilation into eddy-resolving models
The evolution of the eddy structure during each cruise shall be analysed and simulated using data assimilation in models (altimetry and in situ data), in order to provide a realistic view of the large scale and mesoscale circulation. Initialised with the fields measured and mapped during the first leg, these simulations shall be coupled with different biological models to study the biogeochemical tracer distributions, using estimates of parameters and/or fluxes obtained during the second leg with the time series stations. These simulations will be validated using data not considered in the implementation of the model (like surface data obtained during transit), including satellite data (sea colour). Budgets and major fluxes (primary production, export production, air-sea gas exchanges) will then be estimated and the role of the coupling between dynamics and biological activity will be studied at meso-scale and regional scale. It is planned to use eddy-resolving basin scale models as well, which will integrate the information obtained during the cruises on an annual time scale for the north Atlantic ocean : in that respect, sea colour data assimilation techniques will be developed.
 
 

B. Commitments and implementation

I. Committees and funding

The POMME project has been designed by a French group composed of physical and biogeochemical oceanographers. This group, chaired by Gilles Reverdin (LEGOS, gilles.reverdin@cnes.fr) and by Laurent Memery (LODYC, laurent.memery@ipsl.jussieu.fr) has submitted the project to French scientific funding agencies. Review has been made by the PATOM and PROOF committees (for the physical oceanography component, and for the carbon-biological component, respectively). The project has been recommended by both committees, and a French inter-agency meeting, which took place in May 1998, has discussed funding issues. Funding has started to be appropriated by the agencies, and cruise requests are under review. Both a project committee, and a liaison committee have been established, and have met. Further planning and implementation of the experiment POMME will be coordinated by the project committee (currently composed of French scientists), which reports to PATOM and PROOF, and which will be supported by the liaison committee (including two representatives of the project committee) to link to the funding agencies.

In France, the physical oceanography component of the project will be mostly carried by scientists at LPO/IFREMER (Brest), LODYC (Paris), LEGOS (Toulouse), French Met Office (Toulouse), LPCM (Villefranche), EpSHOM (Brest) and CELENV (Toulouse). The major laboratories involved in the biogeochemical component are LPCM and LOBE (Villefranche), COM and LMM (Marseille), IUEM (Brest), SHOM (Brest), LEGOS (Toulouse) and LODYC (Paris). The project will establish links with MERCATOR, the real-time operational oceanography project, and with CLIPPER, a French WOCE program aimed at simulating the global circulation of the Atlantic Ocean in an eddy resolving (1/6°) model.

In the UK, two projects have being proposed and are at different stages of planning. The Southampton Oceanographic Center plans to contribute with a profiling mooring (first in late 1999-early 2000; then in late 2000-autumn 2001) and with tritium/helium surveys. An other project is being proposed by the University of East Anglia and collaborative institutions, which will include release of a tracer SF 6 and the investigation of its later distribution from the French POMME cruises and from a dedicated process study cruise.

There is a number of preliminary studies which are required to show the feasibility of the project, some of which will take place during the preliminary experiment POMMIER occurring a year earlier (autumn 1999 -September 2000). This preliminary experiment has been presented as a prototype of an operational oceanography approach to MERCATOR (the French component of GODAE, the Global Ocean Data Assimilation Experiment), and has been approved by the committee. A brief description of the experiment is given in the appendix.
 

II. Committed instrumentation

A brief summary of the committed in situ effort is given below, and some of it is schematised on figure 1.

Eulerian instrumentation
(to be deployed mostly in the autumn 2000 for recovery a year later, with the exception of the sediment trap moorings which will be installed in January 2001 for 6 months and reinstalled for recovery in summer 2002)

- 1 meteorological mooring (wind, air and ocean temperature and humidity), somewhere south of 43°N. Currently, the instrumentation is sought from the French Met Office. Other instruments or other moorings would be beneficial to the experiment (this will be the only reliable air temperature and humidity measurement in this part of the Atlantic; other moorings are located further north at 48°N and higher latitudes).
 

Figure 1: In situ measurements during POMME: (a) Long-term moorings (one year or more). The rectangle is the POMME area, and the dashed line is the climatological transition zone between deep winter mixed layers in the north, and shallower ones in the south. The RAFOS 260 Hz sound sources array is maintained by SHOM. (b) CTD array to be carried out during the first leg of POMME 1 and POMME 2 cruises. The black stars represent measurements of the IFREMER/GENAVIR ship, the red ones measurements of the French Navy ship. During POMME 3, only an array equivalent to the black one (but maybe shifted) will be done. (c) measurements during leg 2 of POMME 1-2-3: the circles represent two coherent structures tracked from January 2001, sampled with long JGOFS-type stations (in red) and Tow-yo or Seasoar-OPCT lines (dot-ted). Long Seasoar or tow-yo lines are also wanted outside of these structures. Meteorological mooring RAFOS sound source B

 
 

- 3 subsurface ADCP moorings (near 300m), in a roughly north-south line. The moorings will also have small (100m) thermistance chains on top, one or two seacats, and subsurface current meters at at least three levels.

- 5 tomography moorings (400Hz) in a diamond array (fig. 1a) with 2 receivers and one source-receiver in a north-south direction from 39°N to 45°N (these moorings are the same as the ADCP moorings just mentioned) and two sources in an east-west direction (from 15.5°W to 22.5°W). There will also be a few current meters on the moorings, as well as sources for tracking the VCM floats.

- 4 sediment trap moorings. The sampling will be every 10 days, and the moorings will have to be replaced after 6 months. Two of the moorings will be located in the northern part of the domain, and the other two in the south (fig. 1a), within 200 km of each other in order to get the mesoscale signature of the particulate export fluxes in terms of primary production in the upper ocean and to study the spatial and temporal correlation between the moorings (Jickells et al, 1996). Two sediment traps will be placed on each mooring at 400m and 1000m to provide an estimate of the divergence of fluxes across the mode water layer.

- 1 profiling CTD mooring (proposed by S. Cunningham, SOC, UK). Although not part of the French proposal, this is mentioned here for completeness.
 

Lagrangian instrumentation

- 24 mixed layer drifters with 150m or 200m long thermistor chains and wind measurements (MARI-SONDE). The drifters will be deployed probably in early 2001, the ones having shorter thermistor chains in the southern part of the domain, and the ones having longer chains to the north. These drifters are not very good current followers, and will be recovered either in early May or in September 2001.

- 5 SVP surface drifters, either with salinity and air pressure (depending on the results of the tests on the salinity sensors) or with acoustic wind measurements and air pressure (possible contribution of Meteo France). These drifters follow closely the current at 15m, and their drift includes a component related to the vertical turbulent flux of momentum (on the average 70 to 90 degrees to the right of the wind).

- 5 CARIOCA drifters. These drifters are also drogued at 15 m, and measure the wind modulus, air temperature, as well as surface water CO₂ partial pressure and fluorescence. A recent experiment in the northeast Atlantic (GASEX) showed that these instruments provided reliable data for at least 15 days to one month (communication of Liliane Merlivat). Two of the drifters will be deployed in the coherent eddies, and three will provide contrasted situations between the northern and the southern parts of the domain.

- 10 SURDRIFT surface floats drogued at 300 or 400m. The current-following capability of the floats is probably quite good, but maybe not at the 1 cm/s level wished for. The deep drogues have been known to last for 3 to 6 months.

- 15 PROVOR floats at 400m (or at a deeper level). PROVOR is the French equivalent of the PALACE floats with profiles taken every 10 days (autonomy unknown, but they should last more than one year and up to three years at this sampling frequency). Most of these should be equipped with temperature and salinity, although a yet undefined number of floats will not have salinity. The profiles will be from 2000m to the surface. Nine floats will be deployed in an array across the domain in the autumn 2000, and the other six will be deployed in two coherent structures which will be more specifically followed
(January 2001). In addition, we hope that a few of the profiling floats deployed for POMMIER will still be in the area, or will have been recovered and redeployed to increase the coverage. The floats are currently being tested at sea, and the first long deployment is scheduled in September 1999.

- 20 MARVOR floats at 400 m. The floats are followed acoustically, and report their location when they surface. The surfacing time will be scheduled so that it precedes the next cruise (so this could be on he order of 3 months). These floats are known to live up to 5 years, and will mostly be deployed in September 2000 with some deloyed in January 2001

- 10 Rafos floats at 400 m (surfacing after one year in the water) deployed in September 2000.

- 35 VCM RAFOS floats at two depths in the upper 500 meters for a 9 months to one year mission. These floats record the vertical velocity of the flow past the float, as well as temperature and pressure. They have been recovered after previous use in the Alboran and Greenland Seas and will be refurbished. Deployment will be mostly in September 2000 with some deployed later on in January 2001.

Altogether, about three fourth of the floats and buoys will be deployed on regular grids in order to monitor the general mesoscale circulation in the area, the remaining fourth being devoted to the survey of one or two coherent structures.

Analysis of satellite data
In addition, space-related data and analyses will be provided to the experiment (CLS for ERS/TP/JASON/ENVISAT altimetric products; DOS at IFREMER for diffusiometer winds, if available, and surface temperature; Lannion (Met Office) for sea surface temperature and radiative fluxes; other groups for ocean color).
 

III. Oceanographic cruises

A preliminary reduced survey will be carried in the fall 2000 (POMME0 : pre-POMME cruise) with 50 hydrographic stations providing a reduced array. This cruise will be mostly devoted to logistics (mooring and float deployments, release of SF 6 ) and basic bio-geochemical measurements (nutrients, DOC, DIC, chlorophyll). Three cruises are proposed in 2001 (final decision not known until April 2000), each with two ships: in January-February 2001 (POMME1 : winter conditions, pre-subduction, pre-bloom), in April 2001 (POMME2 : subduction-bloom) and at the end of summer - fall 2001 (POMME3: pre-conditioning of the winter export of dissolved organic matter). POMME1 will include the deployment of the sediment trap moorings and of some of the floats, whereas POMME3 will include the recovery of the moorings and of some of the floats. The sediment trap moorings will be redeployed during POMME3 for a final recovery in spring or summer 2002 during a last cruise (POMME4).

During the first leg of the three cruises in 2001 (roughly 20 days), a meso-scale hydrographic survey will be carried out (resolution 55 km) (see fig. 1b). The stations will usually reach the depth of 2500 meters with samples collected from 23 bottles, mostly in the upper 500 meters of the water column. During POMME1 and POMME2, the survey will be done with two vessels (RV.1, being a yet unde-fined vessel from the GENAVIR fleet, R.V. 2 being provided by SHOM), half the stations on the outer rim of the domain being possibly replaced by XCTDs (no samples collected). During POMME3, only a smaller portion of the domain will be sampled. This last hydrographic survey might be shifted a little with respect to the previous surveys based on the displacement indicated by the floats.

The second part of the cruises (on the order of 20 days) will include repeated Towyo profiles (CTD/fluorescence profiles to 800 meters) together with more specific biological - production surveys by R.V. 1 (fig. 1c). This will include JGOFS time series lasting around 48 hours with drifting sediment traps, pumping devices and measurements (using bottles, camera, nets) aimed at quantifying primary, regenerated and export productions of N, C, Si and S, and at getting some basic information concerning zooplankton and secondary production (this last point being presently a weakness of the proposal). These biological - production surveys will in particular be done in and around some specific coherent features identified in January 2001 (in particular, an anti-cyclonic feature in the northern part of the domain, if one is identified). Seasor-OPCT surveys will be done during POMME2 and POMME3 during this second leg by R.V. 2 (R.V. d'Entrecastaux during POMME2). During POMME1 and POMME2, air-sea flux measurements will be made on R.V. 1 using an instrumented mast (CRPE-16 French Met Office) and dropsondes will be launched. Both R.V. 1 and R.V. 2 are equipped with an ADCP current profiler and a thermosalinograph. Altogether, this implies in 2000-2001 roughly 175 days of the R.V. 1, 60 days of the R.V. d'Entrecasteaux and 30 days of a French Navy BSHM.
 

IV. Data policy

The French committees are in the process of defining the best strategy in terms of sharing and validating the data among the different teams. As a first step, the biogeochemical data will be gathered at the French JGOFS data centre at LPCM (Villefranche/mer). Some of the physical data will be transmitted in real time, and will be freely available, as well as some of the near-real time model simulations. The remaining data and the remote time data will be available to the participants after due processing and preliminary validation, inclusion of these data in publications submitted to authorization by the data originator. The data will become public in their final version two years after their collection (for most, public release in September 2003). These data should be assembled in a data bank with WEB/ftp access.
 

C. Comments on international cooperation

We should mention certain weaknesses of the current French commitments to the experiment, which could be corrected by other research organisations. No turbulence measurements are planned in the ocean, and few surface layer drifters will be deployed. No rainfall measurements is currently planned, and only one meteorological mooring (with bulk variables) is proposed, whereas two would have had provided a more complete coverage of the domain. The number of cruises requested is a compromise taking into account funding constraints. Initially, both a summer and an autumn cruise were planned in 2001. Some of the biological processes will be missed by the merging of these two cruises (kinetics of accumulation of dissolved organic matter and of regeneration during the oligotrophic period), and additional cruises, in particular during the summer, would be welcomed. Also, because of the different objectives of the biological cruises and of the second legs of the January and April cruises, the ships are already fully occupied which limits what can be done and in particular the amount of higher resolution TowYo, Seasor-OPCT transects.

The number of floats deployed in the area is currently marginal to map the flow at the meso-scales. We also expect the flow closer to the coasts of Portugal and Spain to be different, with a possible seasonal variability. This will not be sampled. The experimental design has not focused on bottom topography. It is known that deeper in the ocean, the circulation is influenced by the Acores-Biscaye Ridge, which crossects the domain. What effects this interaction with the topography might have on the meso-scales above and of their role in the subduction process is not been investigated by the French research community.

Some of the floats are reusable, and attempts will be made to recover them (Marisonde, CARIOCAs, SVPs). However, it is likely that some of the floats will have drifted too far away from the core of the domain, and therefore will not be easy to recover. Contributions to help recover some of the floats in the autumn 2001 will therefore be welcomed. Initially, it had been proposed to close the section between the Acores and Cape Finistere during each cruise. However, constraints on ship-time make this unlikely. More generally, we rely mostly on satellite altimetry and modeling for providing information on the large scale features of the circulation at the time of the experiment. Although we anticipate an average drift to the south in this area, it is possible that the average flow will be quite different during the year investigated, possibly depending on NAO. It will be important to take into account interannual variability of the subduction processes in order to figure the relevance of the results on the role of the eddies in the subduction. This might require an effort on a reduced scale over a longer time frame; cooperation on that (possibly with floats) will be welcomed.

The data set will provide ample opportunity to test assimilation approaches at meso scales or larger scales. POMME might provide a test-bed for assimilation activities. The large number of data, in particular Lagrangian floats, could make this site an interesting place to test new instruments or compare the performance of different floats, profiling or not. This might therefore be an important test-ground for ARGO, the international project of profiling floats. Some of the floats (PROVOR, MARVOR) will live much longer than the experiment span, and will therefore contribute to the later float sampling of the North Atlantic. The PROVOR floats deployed during POMME will be a contribution to ARGO. We will attempt to make as much data available in real time as is possible, and we welcome their use by modelers interested in near-real time activities within EuroGOOS.

Currently, tritium/helium and SF 6 are the only ventilation tracers planned, and both are proposed by UK teams (tritium/helium, W. Jenkins (SOC); SF 6 , A. Watson (Univ. East Anglia, UK)). SF 6 will be released during POMME0 and surveyed during POMME3, and an additional cruise during the spring 2001 is necessary for the survey of SF 6 and to the investigation of fronts and other high resolution fea-tures. It will be reauested in the UK (NERC).

The bio-geochemical contribution of the program is designed to achieve accurate tracer budgets in the upper ocean and in the mode waters. This implies that besides tracer distribution at mesoscale, emphasis is mostly put on the major fluxes during the exchanges between the inorganic and the organic pools (primary, total, regenerated, new production; bacterial activity; particle fluxes; remineralization). Transformation of the organic matter within the trophic system is only addressed by measurements during the long stations. Partial information will be obtained mostly from OPCT data, several studies associated with the microbial loop, and stocks of zooplankton sampled with nets during the process study stations. Therefore, secondary production and coupling between phytoplankton and zooplankton are presently rather poorly studied in POMME.

Finally, as meso-scale surveys, as well as annual monitoring are sought providing an intensive data coverage, this programm is certainly a relevant frame to undertake tests or measurements with new automatized devices.
 
 

Figure caption

Top : the POMME region. The large square determines the area of intensive coverage. The oblique line represents the approximate position of the zone of discontinuity of the winter mixed layer depth, where the mode water subduction should occur and which delimits a relatively productive region in the north from a poorer region in the south.
    x : currentmeter moorings ;
    + tomography moorings ;
    diamonds : sediment trap
    (+ currentmeter) moorings.

Middle : first leg. x = CTD stations for the R.V. D'entrecastaux (SHOM) ; diamonds = CTD stations for the R. V. Atalante or Thalassa. Several southern stations will be sampled by both vessels for intercomparison purposes. D (Debut) localizes the first station of the leg and F (Fin) the last one. The call is supposed to be at the Azores.

Bottom : second leg. The two circles schematize two coherent eddies localized during the first leg of POMME1. The triangles show the position of the JGOFS type long stations (two days). In the north, the eddy structure will be sampled inside, outside and at the border, whereas a visit inside and outside are anticipated in the south. The dashed lines represent Tow-Yo sections. During the first part of the leg, a meridional Two-Yo section will be made from the south west corner (the vessel leaving from the Azores). The two eddies will be sampled with the Tow Yo as well (two orthogonal lines). After the five long stations are completed, a least Tow Yo section will be made during transit towards La Coroña or Vigo.
 

Appendix: French preliminary studies and POMMIER
 

Preliminary studies will contribute in 1999 and 2000 to better define the experimental design. Some of those will consist of the analysis of high resolution model experiments and of existing data sets (thermal sampling, hydrographic sections, float trajectories (EUROFLOAT, WOCE SVP...), heat and freshwater fluxes...). The development of assimilation methods in regional models will be continued. We should also mention the experimental design studies that will be done for the international profiling float array ARGO, and that are recommended by the international committee of the Global Ocean Data Assimilation Experiment (GODAE). These studies will consist in sampling floats with a high spatial horizontal resolution at different depths of different high-resolution simulations of the North Atlantic, and at sampling the temperature and salinity profiles at regular intervals. The issue that can be investigated with this study is whether there will be an array bias, either for the velocity or for the temperature and salinity sampling (the simulations that shall be considered will be the 1/3 and 1/6th of a degree CLIPPER simulations of the Atlantic and the 1/12th of a degree (MERCATOR) 18 month simulation of the North Atlantic). We will also investigate the respective interest of having floats drifting at a fixed depth or on temperature (or density) surfaces, the two options being a priori possible with the floats available (P-ALACE and PROVOR). It will also be important to address the issue of the sampling required to measure the circulation, and on whether it is possible to share the resources between two levels (two surfaces).
 

POMMIER
POMMIER has been designed to test the quasi-real time operational approach with data assimilation of satellite altimetric and in situ data between September 1999 and September 2000. It will take place in the same area as POMME, and will include the deployment of 15 PROVOR profiling floats (11 of which will measure T, and four T and S,using either a SeaBird or a FSI sensor), as well as one high resolution CTD-XBT survey, probably in March-April 2000 in the central part of the POMME domain where a large spatial variability of mixed layer depth is expected. There will also be other XBT launches in September 1999. A prototype of an EMMA autonomous station will also be deployed (the station shall ultimately be made of a pack of expandable CTDs deployed at the ocean bottom and released on a pre-set schedule, which profile to the surface, where they transmit the profile data to ARGOS). The opportunity of collecting water samples for biogeochemical tracer measurements (nutrients, DIC, alkalinity, DOC,chlorophyll) is also discussed.

It will be probably the first time that the PROVOR floats will be deployed for a long period of time (two prototypes with temperature only are currently tested at sea). The technology, however, strongly relies on the one used for the MARVOR floats (the salinity sensor will be one version of the FSI sensor used by the PALACE floats). This project will be an opportunity to test the processing and the accuracy of the PROVOR data, and to test in a regional model the data assimilation of the temperature and salinity profiles, as well as of altimetric sea level in near real time. The experiment will provide interesting estimates of the circulation and of the mixed layer structure before POMME. It is possible that some of the floats will be recovered (in particular, in case of failure), and that some will still be in the area during POMME. The group in charge of the project is the CELENV in Toulouse, with a contribution of EpSHOM (Brest) and of DITI/IFREMER (Brest).
 

References :
        Brügge, B., Near surface mean circulation and kinetic energy in the central North Atlantic from drifter data, J. Geophys. Res., C10, 20543 - 20554, 1995.
        Ducklow, H. W., and R. P. Harris, Introduction to the JGOFS North Atlantic Bloom Experiment, Deep-Sea Res., 40, 1-8, 1993
        Jickells, T. D., Newton, P. P., King, P., Lampitt, R. S., and C. A.Boutle, A comparison of sediment trap records of particle fluxes from 19 to 48N in the northeast Atlantic and their relation to surface water productivity, Deep-Sea Res., 43, 971-986, 1996
        Hurrell, J. W., Decadal trends in the North Atlantic Oscillation: regional temperatures and precipitation, Science, 269,676-679, 1995.
        Hurrell, J. W., Influence of variations in extratropical winter time teleconnections on Northern Hemisphere temperature. Geophys. Res.Lettrs., 23 (6), 665-668, 1996.
        Karrasch, B., Hoppe, H. G., Ullrich, S., and S. Podewski, The role of mesoscale hydrography on microbial dynamics in the northeast Atlantic :Results of a spring bloom experiment, J. Mar. Res., 54,99- 122, 1996
        Lévy, M., Mémery, L., and G. Madec, The onset of a bloom after deep winter convection in the North-western Mediterranean sea : mesoscale process study with a primitive equation model, in press in J. Mar. Syst., 1998
        Lochte, K., and O. Pfannkuche, Cyclonic cold - core eddy in the eastern North Atlantic. II. Nutrients, phytoplankton and bacterioplankton, Mar. Ecol. Prog. Ser., 39, 153-164, 1987
        Mc Cartney, M.S, 1982: The subtropical recirculation of Mode Waters. J. Mar. Res., 40 (suppl.), 427-464.
        Marshall, D., and J.C. Marshall, On the thermodynamics of subduction, J. Phys. Oceanogr., 25,138-151, 1995.
        Marshall, Nurser, Williams, "Inferring the subduction rate and period over the North Atlantic", J.Phys.0ceanogr., 23, 1315-1329, 1993.
        McGillicuddy Jr, D. J., Robinson, A. R., and J. J. McCarthy, Coupled physical and biological modelling of the spring bloom in the NorthAtlantic (II) : three dimensional bloom and post-bloom pro-cesses, Deep-Sea Res., 42, 1359-1398, 1995
        Oschlies, A., and V. Garçon, Eddy - induced enhancement of primary production in a coupled ecosystem - circulation model of the North Atlantic Ocean, submitted to Nature,1998a.
        Paillet, J., Central water vortices of the eastern North Atlantic. J. Phys. Oceanogr, Sous presse, 1999.
        Paillet, J., and M. Arhan, Shallow pycnoclines and mode water subduction in the Eastern North Atlantic J. Phys. Oceanogr., 26, 96-114, 1996a.
        Paillet, J., and M. Arhan, "Oceanic ventilation in the eastern North Atlantic, J. Phys. Oceanogr., 26, 2036-2052, 1996b.Brügge, B., Near surface mean circulation and kinetic energy in the central North Atlantic from drifter data, J. Geophys. Res., C10, 20543 - 20554, 1995.
        Paillet, J., and H. Mercier, "An inverse model of the eastern North Atlantic general circulation and thermocline ventilation, Deep-SeaRes., 44, 1293-1328, 1997.
        Pollard, R. T., Griffiths, M. J., Cunningham, S. A., Read, J. F., Perez, F. F., and A. F. Rios, , Vivaldi 1991 - A study of the formation, circulation and ventilation of Eastern North Atlantic Central Water, Prog. Oceanogr., 37, 167-192, 1996.
        Rios, A. F., Anderson, T. R., and F. F. Perez, The carbonic system distribution and fluxes in the NE Atlantic during Spring 1991, Prog. Oceanogr., 35, 295-314, 1995
        Spall, S. A., The impact of mesoscale jet activity on plankton heterogeneity and Primary Production ; a numerical modelling study, Thesis, University of Southampton, 1998.
        Speer, K., H.J. Isemer, and A. Biastoch: Water mass formation from revised COADS data, J. Phys. Oceanogr., 1995.
        Takahashi, T., Takahashi, T. T., and S. C. Sutherland, An assessment of the role of the North Atlantic as a CO₂ sink, Phil. Trans. R. Soc. Lond. B, 348, 143-152, 1995.
        Wroblewski, J. S., Sarmiento, J. L., and G. R. Flierl, An ocean basinscale model of plankton dynamics in the North Atlantic : 1. Solutions for the climatological oceanographic conditions in May, Global Biogeochem. Cycles, 2, 199-218, 1988
        Yoder, J. A., Aiken, J., Swift, R. N., Hoge, F. E., and P. M. Stegman, Spatial variability in near-surface chlorophyll a fluorescence measured bythe Airborne Oceanographic Lidar (AOL), Deep-Sea Res., 40,37-53, 1993