Group Leaders: Ralph Keeling and Chris Sabine

Objective: To encourage the completion of data quality, synthesis and archiving activities on large-scale ocean data sets (e.g., global CO₂ survey) relevant to the goals of SMP and to ensure uniform data access to those data sets and derived products.

Background: The working group identified a variety of activities that are only marginally (if at all) represented within JGOFS-SMP funded investigations, but which are needed to fulfill the overall goals of SMP.
 

1. A coordinated analysis of the WOCE/JGOFS carbon-system data to place the relevant variables (DIC, ALK, pCO₂, and pH) on a "consistent basis". Not all of these were measured on all cruises, but in principle they can be computed for all cruises where at least two were measured. A uniform dataset for these variables is needed for a wide range of potential JGOFS-related research activities.
2. A coordinated analysis/syntheses of other WOCE/JGOFS hydrographic data including but not restricted to isotopic, nutrient, and transient tracer data. Included in such activities are generation of generally useful derivative fields, such as "anthropogenic CO₂" or pre-bomb ¹⁴C, etc.
3. Studies which compute the fluxes of various chemical quantities across hydrographic sections. Such studies might profitably be linked with studies of fluxes (e.g., of CO₂) across the air-sea interface and with inventory studies. A global joint synthesis of transports, air-sea fluxes, and inventory changes should be viewed as one of the tangible long-term goals of JGOFS SMP but with strong ties to the WOCE program.
4. Studies which can address how best to correct upper ocean data sets for seasonal variations in order to assess long-term trends.
5. The need to tie in all of the process study and time-series data into the large-scale observational program.

The working group believes the SMP should encourage a unification of the process studies, time-series, and global survey data to specifically address point #5.  The PI meeting was a great step forward in unifying the modelers, but a similar approach should be taken to unify the individual measurement programs sponsored through JGOFS (including international cooperation).

Tasks: Open discussions with other data facilities including the WOCE hydrographic office and CDIAC to develop a seamless data system. This system will be needed for many research activities both within and outside of JGOFS-SMP. The generation of this data system would eventually help to avoid duplicated effort. We envision that such a system would provide consistent carbon-system data, discussed above, and other hydrographic data, such as self-consistent nutrients, O₂, ¹³C, ¹⁴C, transient tracers, and "product" fields, such as "anthropogenic CO₂," or pre bomb ¹⁴C, etc.

We suggest that a positive development would be securing some core support from within JGOFS SMP to establish and manage such a data system. Whether this data system would be housed at WHOI, through CDIAC, or elsewhere, we suggest that some "core" support would be beneficial in providing critical management infrastructure, and in communicating to a larger community (including funding agencies), JGOFS-SMP intent to serve as a central program for facilitating and coordinating research on large-scale chemical datasets.

Membership:
Bob Bacastow
Chris Sabine
Ralph Keeling
Scott Doney

Group Leaders:  Christopher Sabine and Ralph Keeling

Objectives:  Encourage the completion of data quality, synthesis and archiving activities on large-scale ocean data sets (e.g. global CO₂ survey) relevant to the goals of SMP and to ensure ready access to those data sets and derived products. We also wish to promote a close interaction among individual projects and principal investigators (PIs) working with these data sets.
 

Data Reporting and Archival

The working group felt it was important that the large-scale data products being generated by the various SMP projects be reported and housed at the JGOFS office in a timely manner.  These products should be packaged in a manner that makes them easy to access and useful to investigators not directly related to that project.  These data packages may be produced as part of individual projects or by groups of projects with overlapping interests.  Although we do not wish to impose a rigid format, the data should be made available as flat ASCII format (if possible) and have an associated "metadata" file containing full documentation of the data (e.g. units) and how the various products were derived.  Exactly what is included in the metadata will initially be up to the relevant PIs, but as data packages are submitted to JGOFS, they will be examined for completeness by the JGOFS data management office.  The working group expressed an interest in having the data management office try to bring related products into a common format (e.g. global pCO₂, satellite data and global Trichodesmium distributions).  These merges will potentially involve adjustments to put all data onto a common time and space scale.  At a minimum, the JGOFS office should organize all of the products in a logical manner to help promote easy access to all of the data.  We expect the JGOFS office to continue to build and organize the archive of products for the lifetime of the JGOFS program.  Once the program is completed (3-5 years) the entire JGOFS package should be transferred to NODC for long-term archival.  We would like to see NODC maintain the JGOFS archive separate from the bulk data archive and to retain the organization developed over the lifetime of the JGOFS program.

Summary of Relevant Proposed Products - (future link)

Examples of Potential Cross-project work - (future link)

Major Hurdles

The working group identified two potential hurdles to the cross project interaction we are trying to promote.

• The first is lack of access to relevant data sets that are not being examined directly as part of JGOFS SMP.  We need to identify these data sets and investigate ways of making these data available to the general community (e.g. Takahashi's pCO₂ data, D. Keeling's carbon data).
• Second, we need to investigate whether resources are available at the JGOFS Data Management Office to unite the different products discussed in the Data Reporting and Archival section above.  It would be helpful to bring related products into a common format (e.g. global pCO₂, satellite data and global Trichodesmium distributions). These mergers will potentially involve adjustments to put all data into a common time and space scale.

Future Sampling Needs

The working group felt it was important to maintain an ocean-going measurement program, but at the same time strongly support technology development for putting carbon sensors on moorings and drifters. It is becoming clear that higher resolution data sets are necessary to address the issues of spatial and temporal variability in the carbon system.  An effort should be made to tie in carbon measurements with existing efforts to deploy buoys and PALACE floats.  We believe that this effort should emphasize the Southern Ocean where the issues of carbon uptake are thought to be important and ship work is difficult.  There are still a number of technical problems with including carbon measurements on floats and moorings, but we believe it is important to continue that development. In addition, we should begin to emphasize ties between inorganic carbon and remote sensing, the possibility of profiling buoys with carbon systems and expendable sensors that include T, S, CO₂ and O₂. Although the technology development described above is important for the future of ocean carbon research, we need to address the issues of global change right now. Given the current state of the technology, the best and most reliable way to do that is with shipboard measurement programs.  These measurement programs will also be needed to test the technology development and provide ground truthing for remote systems.  The group felt that any future carbon sampling program should also include total organic carbon and iron measurement to help resolve carbon issues that cannot be addressed with the current data set.  There are four main areas that need to be addressed with future measurement programs:
 

• We need to fill in holes in the current global carbon survey.  Missing regions/lines include a zonal line in the North Pacific, full I5 line, I7S, southwest Indian Ocean and southwest Pacific.  These areas were not included in the WOCE survey because sufficient hydrographic data were available, but carbon data are not available on these lines. The most important lines are a zonal North Pacific line (i.e. 24N) and a complete I5 line.  If we are going to fill these holes, the cruises need to run as soon as possible to maintain some continuity with previous WOCE/JGOFS lines and while the expertise in WOCE and JGOFS is still available.

• We also feel that a continuation of the time series stations is very important.  There needs to be a core set of measurements that are maintained over decades in a manner similar to the atmospheric monitoring network.  This work will help us to evaluate and understand temporal variability in the oceans and can be used to monitor future climate changes at a few key locations. In addition to carbon parameters, these core measurements should include other anthropogenic trace gases (e.g. CFCs).  The suite of observations must regularly be evaluated and kept to economically and logistically practical levels. The times series stations will also provide important test beds for technology development discussed below.

• A program designed to monitor future changes in anthropogenic trace gases needs to be supported.  Repeat sections in key areas (e.g. areas where the largest changes are expected) would be sampled on roughly a 10-year time-scale.  These cruises need to run to at least WOCE/JGOFS standards.  The group felt that, in general, the physical oceanographic community may not support the idea of repeat lines, but we feel that the anthropogenic issues are sufficient to drive this effort.

• We need to continue to develop our understanding of variability in the surface ocean carbon system.  By outfitting ships of opportunity with autonomous sampling packages, we can get high resolution temporal data sets at relatively low cost.  These sampling packages should sample a suite of parameters (e.g. temperature, salinity, two carbon parameters, chlorophyll, nutrients) to help address the mechanisms responsible for the observed variability.

Working Group Tasks

1. Investigate NODC's willingness to maintain JGOFS products as a package separate from the bulk data. This package would be maintained in the form generated by the JGOFS Data Management Office. -- P. Murphy
2. Work through the US and International JGOFS offices to compile a list of carbon data sets relevant to JGOFS and identify those data which are not currently available. Talk with the chairs of the US process studies to see what other international data may be available from those areas. -- C. Sabine, A. Dickson, P. Murphy
3. Talk with J. Swift and L. Talley to find out what WOCE type cruises we may not be aware of which had carbon measured on them. -- P Robbins
4. Compare different algorithms for gridding large-scale data sets.  These comparisons would include an evaluation of various objective mapping techniques as well as the utility of interpolations based on multi-parameter fits to common hydrographic parameters.  Comparisons will be made for sections along cruise tracks, basin-scale mapping and generation of 3-D gridded data. -- R. Key and P. Quay
5. Consider writing a review paper on science issues that require additional technology development. -- ???

C. Sabine
P. Robbins
R. Wanninkhof
K. Lee
R. Feely
A. Dickson
P. Quay
R. Key
L. Gordon
H. Garcia
B. Bacastow
P. Murphy

Group Leaders:  Chris Sabine

Objectives:  Encourage the completion of daa quality assessment, synthesis and archiving activities on large-scale ocean and satellite data sets relevant to the goals of SMP, and to ensure ready access to those data sets and derived products.  We also wish to promote a close interaction among individual projects and principal investigators working with these products.

Major Hurdles:

• Lack of access to relevant data sets that are not being examined directly as part of JGOFS-SMP.
• Various data products need to be pulled together to a common location.

Priorities for JGOFS:

1.  Measured Data

• Get all JGOFS data onto the JGOFS data server
• Merge the data together as much as possible by data type
• Develop a search engine for data
• Develop visualization tools for data

2.  Synthesized Products

• Post the list of anticipated synthesized products together with the PI and proposed release date
• Get synthesized products into LAS

3.  Archiving

• Start producing CDs of JGOFS data
• Start discussions with NODC for long-term archival of the JGOFS products