The U.S. Joint Global Ocean Flux Study Global Survey Implementation Working Group (F.J. Millero, O.B. Brown, J.K. Cochran, G.C. Feldman, J.W. Murray, R.J. Toggweiler and J.A. Yoder) was charged with the task of reviewing the rationale for large scale sampling in JGOFS and discussing the past, present, and future implementation of the Global Survey component of the program. This report discusses the current status of the global survey and recommends ways to improve the program in the future.

3.1 Review of Status of Science Plan

The U.S. JGOFS Long Range Plan (U.S. JGOFS Steering Committee, 1990) gives a detailed rationale for the global survey component of the program. The plan suggests the need for a new global survey with the same spatial coverage as the GEOSECS program in which a new suite of biogeochemical variables are measured. The global survey is an important component of the JGOFS program. It will provide biogeochemical data on the carbon system needed for modeling and provide the link between the surface of the ocean viewed from space and the sedimentary record that views the past. The specific objectives of the large scale program are:

1. Provide seasonal characterization of biogeochemical parameters on at least basin scales for the computation of a flux budget validation.

2. Provide a consistent global description of surface pigment, primary production, CO2 and export fluxes and transformations.

3. The task group agrees with these scientific objectives and the scientific rationale behind them.

The implementation of this plan was to be achieved by:

1. Implement key U.S. JGOFS sections to acquire data essential to describing U.S. JGOFS oceanic zones. These sections will transect the oceanic provinces within which process studies are sited, thereby providing far field data beyond the process study area.

2. Coordinate the U.S. JGOFS programs with survey components of other JGOFS nations to achieve a balanced, comprehensive coverage of the world ocean.

3. Promote a variety of intercomparison and intercalibration activities between various national groups to enhance complementarity of various national JGOFS efforts.

4. Cooperate with other global programs to produce large scale seasonal descriptions of mixed layer productivity and resultant vertical flux (WOCE, TOGA, GLOBEC, IGAC, IGBP) by augmenting their observing systems as required.

1. An agreed suite of well-defined intercalibrated, biogeochemical core measurements will be made throughout the water column at regular spacing along a world-wide series of JGOFS transects. Such a suite will include pigments, nutrients, biomass components, gases, dissolved organic species, particulate organic carbon and nitrogen, and radionuclides.

2. The same suite will be measured, whenever possible, on ship-of-opportunity transects, either when on passage to JGOFS transects or process studies, or in cooperation with other national or international programs.

3. A global array of sediment traps will be deployed to estimate the vertical particle flux. Ideally these traps would be in close proximity to the JGOFS transects.

4. A global set of benthic measurements will be made to provide boundary constraints.

5. Along track observations of surface pigments, nutrients, CO2 and O2 will be made on all JGOFS cruises.

3.2 Review of Current Activities

It was recognized early in the program of the need for a major global survey of the oceanic carbon chemistry (CCCO, 1988). This led to the agreement by JGOFS and WOCE to coordinate the measurement of CO2 on the WOCE WHP cruises. Two berths were made available to make carbonate measurements. The measurements of TCO2 (total carbonate), TA (total alkalinity), and pCO2 (partial pressure of carbon dioxide) were to be made, respectively to 1 mol/kg, 1 eq/kg, and 1 atm. This accuracy is presently achievable. Intercalibration and intercomparison exercises have been initiated for the components of the carbonate system by the DOE science group under the direction of Andrew Dickson (Scripps). This group has also written detailed protocols for the measurement of the CO2 parameters. As part of this work, TCO2 and TA standards are now available. This will insure that the measurements made by various groups will satisfy the requirements of the JGOFS program. This global carbon dioxide survey is the primary goal of the global survey and should be strongly pushed.

An agreement was reached that when possible a third berth would be available to JGOFS scientists who would make measurements of the underwater optical field and phytoplankton pigments distribution. The implementation of the optics and pigment part of the global survey has been considered by the U.S. JGOFS Optics Task Team (Marra, 1991). The goal of this work is to be able to improve the algorithms used for computing chlorophyll from satellite ocean color. It is important to get as much global coverage as possible to develop a world-wide pigment data base.

Much of the above satisfies the JGOFS contribution to the global survey. Notably lacking in the present program is the global array of sediment traps and benthic measurements. This is due to the lower priority of these studies and the large scale costs of carrying out these activities on a global scale. The activities have been carried out as part of the process studies and are not completely lacking in the JGOFS program. More effort should be made in making continuous pCO2 measurements on all the JGOFS and WOCE cruises and ships of opportunity.

3.3 Detailed Implementation Strategy

3.3.1 Present Global Surveys (Near Term)

To satisfy the important goals of the present JGOFS global survey program we recommend:

1. Every effort should be made to man all WOCE WHP cruises with carbonate chemists. Minimum measurements should include continuous pCO2 measurements and vertical profiles of TCO2 and pCO2. If TA and TCO2 measurements are made on the profiles, measurements of pH should also be made to eliminate the possible problems with TA due to organic acids. As the techniques become available to make reliable dissolved organic carbon measurements, the routine collection of frozen samples should be carried out on all WOCE WHP cruises.

2. Every effort should be made to equip all JGOFS ships and ships of opportunity with continuous pCO2 systems to get the maximum global coverage of this important parameter. The two time-series stations should also be equipped to measure pCO2 at least in a batch mode. These measurements should be made part of the routine measurements made at the sites.

3. Every attempt should be made to fill the third berth (when available) with an optical/pigment person. JGOFS scientists should make measurements of the underwater optical field and phytoplankton pigment distribution on as many cruises as possible. This can best be implemented by the optics task group.

4. One of the major weaknesses of the JGOFS carbonate global survey program has been in the coordination of the WOCE WHP legs on an international level. We strongly endorse the existence of a cruise coordinator for the global survey and process study components of the JGOFS sampling program. This provides the coordination of the equipment and personnel needed to make sure the carbonate parameters are measured on all WOCE legs and insures that the above recommendations are carried out.

3.3.2 Future Survey Programs (Medium Term)

The implementation task group also considered the possibility of carrying out more limited surveys in future years. It was realized that the funds and scientific interest in carrying out another large scale global survey was not in the cards. Further measurements on a few of the long survey lines may be of interest every five years, but more limited surveys may be needed to solve some key questions that are presently important or may come about as a result of process studies and the initial results of the global survey and the process studies. Some of the science questions that can be addressed by such studies considered by the task team are given below.

Seasonal Survey of pCO2 in the North Atlantic: There presently is a growing contro- versy over the size of the oceanic sink for CO2 (Tans et al., 1990). Ocean models suggest that the sink must be 2 Gt/y with 1.2 Gt/y in the southern and 0.8 Gt/y in the northern hemispheres. These models are based on the known geographical distribution of fossil fuel burning and the time constant for gas exchange between the hemispheres. They show that the southern hemisphere ocean has no net uptake or loss to the atmosphere. Thus, the northern hemisphere must be absorbing the entire 2 Gt/y. Present estimates, however, are only 1 Gt/y. It is important to assess the accuracy of the lower sink estimates based on oceanic data. The recent JGOFS process studies in the North Atlantic have shown that pCO2 is extremely patchy in the spring and summer. Strong correlations were found between chlorophyll and pCO2 as well as latitudinal gradients (Turner et al., 1989). There is thus an urgent requirement to document more accurately the seasonal cycle of pCO2 in the North Atlantic and Pacific and, if possible, the Southern Ocean. At present seasonal surveys of the North Atlantic would be the easiest to achieve. The minimum requirements would be underway measurements of pCO2 and TCO2 along the cruise track, backed by measurements of chlorophyll. Measurements of primary productivity, optical properties and plant pigments would also be useful. These studies could also supply the data needed to ground truth the SeaWiFS satellite color data. These measurements were regarded as a high priority for JGOFS (SCOR, 1990) and can be carried out as a series of survey cruises. We recommend that every effort be made to develop a U.S. JGOFS component for this program as soon as possible.

Thermocline Ventilation in Subtropical Gyres: A significant problem exists with regard to the carbon balance in the North Atlantic subtropical gyre. Several different approaches have been used to estimate new production. Oxygen-based remineralization rates utilizing helium/tritium and 228Ra lead to values ranging from 2.5 to 8.5 mol C m-1 y-1 (Riley, 1951; Jenkins, 1980; 1987; Sarmiento et al., 1990). Estimates based on mixed layer oxygen production range from 3.0 to 5.6 mol C m-1 y-1 (Spitzer and Jenkins, 1988; Musgrave et al., 1988). All of these estimates are higher than expected based on annual values of primary productivity and conventional ideas about food web structure and nutrient recycling. Estimates of new production based on vertical nitrate supply (Jenkins, 1988) and particulate flux measurements (Altabet, 1989) are about an order of magnitude lower (0.33 to 0.6 mol C m-1 y-1).

One hypothesis that may resolve this discrepancy is that oxygen consumption and nitrogen remineralization rates are greater than the vertical flux of local new production because of respiration of dissolved organic matter of remote origin transported into the thermocline by ventilation. There is a need to study the mechanism and processes controlling the horizontal and vertical transport of DOC/DON and chemical tracers of respiration from the surface region of density outcrops into the thermocline. There have been few studies of DOC in the thermocline region. In all cases a good correlation exists between DOC and AOU, although the slope of the relationships vary and their origin is unclear.

Another important reason for studying this region is that ventilation of the thermocline is one of the main short time scale pathways for the removal of fossil fuel CO2 from the atmosphere (e.g., Bradshaw and Brewer, 1988; Brewer et al., 1989). As atmospheric CO2 increases, so does the CO2 content of the water subducted. A detailed study of the source region is required to understand how the CO2 signature of the subducted water is determined.

Thermocline ventilation occurs primarily as a pulse produced during the winter cooling. The composition of the source waters supplying the ventilation needs to be determined to set the initial boundary values. This study will help define how subsurface ocean water acquires its preformed nutrient signatures.

This component of the Global Survey should consist of seasonal sections from the high latitude isopycnal outcrop regions to the center of the subtropical gyre. Sampling should cover the water column through the deepest density surface that outcrops (about = 27.4). A complete set of chemical tracers of respiration and ventilation rates should be measured and chemical and biological studies of the outcrop region should be conducted. It could be conducted in either the North Atlantic or Pacific, however, the larger historical data base in the North Atlantic makes it a more logical choice.

Measurements of Pigments and Optics: One of the key JGOFS strategies is to use satellite ocean color measurements to help determine the mean end fluctuating components of ocean basin primary production. To help accomplish this objective, global surveys can make two im- portant contributions. First, surveys will provide a large data base of near surface chlorophyll a (and other pigments) concentrations to compare with estimates derived from satellite sensors, either from concurrent observations or in a climatological sense. These observations will be particularly valuable if spectral reflectance measurements are made at the same time as the pigment measurements. Secondly, surveys will determine how well euphotic zone Chl a (depth-integrated) can be estimated from near surface measurements. The latter is particularly important, since Chl a is not always uniformly distributed in the euphotic zone.

The global survey of Chl a and optical measurements will be useful only if the mea- surements are made using the proper techniques and if the results are expeditiously compared with satellite observations. These requirements mean that:

1. Concentrations of Chl a and other pigments will be determined using HPLC techniques, with the results traceable to JGOFS standards.

2. Reflectance measurements are made using in-water optical sensors that meet specifications approved by NASA's SeaWiFS Project Office.

3. Results are collated and archived by NASA's SeaWiFS Project Office.

4. Pigment and optical results will be most useful if they can be easily related to upper water column hydrography (T, S, and nutrients).

No requirements have been set for horizontal station spacing. With respect to the measurements described here, the role of the survey is to define large scale gradients in Chl a and its relation to spectral reflectance. Thus, the sampling plan presently calls for samples to be collected once per day during WOCE WHP cruises, and this may provide an adequate data base. However, there are major parts of the global ocean and times of the year where we anticipate problems in using satellite ocean color data to estimate phytoplankton biomass, productivity and other in-water constituents and processes. These areas may need to be revisited during specific seasons and sampled more extensively than possible by WHP.

Measurements of Sediment and Benthic Parameters: The ultimate fate of particulate organic carbon which escapes the surface waters and sinks to depth lies in processes happening near the sediment-water interface or in bottom sediments. Studies of these benthic processes are required to determine the fluxes of carbon and nutrients to the sediments and back to the bottom water. Moreover, studies of the controls on organic carbon preservation, as well as opal and calcium carbonate, in the marine sedimentary record are necessary for the reconstruction of paleoproductivity. The magnitude of benthic fluxes varies from place to place, and in margin areas (shelf and slope) the intensity of carbon cycling is often greater than in the deep sea.

In the context of JGOFS, benthic studies have both process and survey components. The aim of the survey studies is to map the spatial and temporal variability of carbon fluxes in the ocean. Although a large scale benthic survey is now an unlikely component of JGOFS, significant information on the spatial variability in benthic carbon fluxes can be gained through incorporation of a survey component in process studies. The strategy favored by the JGOFS Benthic Process Task Team is to make flux measurements along transects through areas such as those of equatorial upwelling, eastern and western boundary currents, continental margins and shelves, polar seas and central gyres. Many of these areas have been targeted for process studies, and a benthic survey component can be added by making sure that each process study site has a transect of benthic stations through it. The priority for such transects should be:

High priority:

Equatorial regimes

Eastern and western boundaries (oceanic margins)

Medium priority:

Monsoonal regime (Arabian Sea)

Areas of strong seasonal variations (North Atlantic)

Polar seas

Low priority:

Central gyres

Larger scale mapping of benthic carbon (and perhaps nutrient) fluxes may be facilitated by measurements of proxy indicators of flux. For example, Pb-210 removal from the ocean water column has been correlated with carbon flux in sediment traps (Moore and Dymond, 1988), and measurements of excess Pb-210 inventories in sediment cores may reflect long term average carbon fluxes to the site (Cochran et al., 1990). Thus collections of box cores on JGOFS and related cruises could expand the benthic survey data base.

In order to constrain the near-bottom cycling of carbon, both flux measurements and rates of processes must be measured. Flux measurements include: 1) the sinking particle flux (measured by particle traps, filtration, optical or isotopic methods), 2) the permanent burial flux due to long term sediment accumulation, and 3) solute fluxes across the sediment water interface. Important rates of processes which must also be measured include: 1) the rate of organic carbon oxidation, 2) the rate of biological mixing of sediments (by particle mixing and burrow irrigation) and 3) the rate of physical transport processes which affect exchange between sediment and bottom water. A JGOFS Benthic Survey must then include components dealing with benthic biology, microbial ecology, pore water geochemistry, benthic flux measurements, sediment coring, and particle trap deployment and recovery.

3.4 Resources Necessary for the Future Limited Surveys

The resources necessary to carry out the more limited surveys or process studies are estimated to be in the $1 to $2M range. This estimate is based on the involvement of 3 to 5 scientists and estimated shiptime of 30 to 60 days. Since most of the surveys listed above can also be considered as mini-process studies, we recommend that they be funded as outlined in the prospectus approved by the JGOFS Steering Committee.