Proposed agenda

GTSPP Washington, 26 Mar, 2001

The meeting was opened at 0900 at the U.S. NODC in Silver Spring, Maryland by Bob Keeley. The agenda is shown in Annex 1. Attendees are listed in Annex 2. Because of the number of issues raised by Rick Bailey, and because the DPC meeting was to take place the next few days, it was decided to concentrate the discussions on matters related to V2 and V3 of the UOT CDs as well as the relationship of GTSPP to Argo. As a consequence, a number of the other issues recorded in the agenda were not covered. The report of the meeting includes written reports of these matters.

Bob was in Australia the following week and had discussions with Rick Bailey and others at JAFOOS and CSIRO. Results from those discussions are included as well.

Actions resulting from the meeting are compiled in Annex 6.

1. Status Reports

Bob reported that MEDS continues to support GTSPP through their regular funding. This is still an important program for MEDS and will continue to get resources as needed. MEDS has had a number of retirements in the last year, but has been able to hire some new staff. These are learning what a data centre does and the specific programs in which MEDS is engaged. Bob also reported that MEDS has recently started to provide real-time BATHY and TESAC data to Brest in support of developing Argo activities.

Kurt reported that NODC has the same resources as in past years to support GTSPP. They are in the process of re-examining their operations and in particular the communications between the GTSPP database and the Ocean Products Data Base, OPDB. They are also looking at the software used to support GTSPP and intend to improve its flexibility and function. He passed out some material relating to NODC plans and invited comments.

Norm reported on behalf of the Pacific Ocean Science Centre at Scripps. Warren White is still a strong supporter of GTSPP and has secure funding for the next 2-3 years. The person carrying out the quality control function is still available (after 50 years of work!) to spend the required 3 weeks needed to carry out the QC for a year of data collected in the Pacific Ocean. They have recently received the 1997 data and will soon have completed the QC.

Loic reported that Brest was also modernizing its processing systems to bring the GTSPP functions within those of their Coriolis system being developed in support of Argo. This is very well developed at the present. He also remarked that they were improving their QC system, in particular to permit automatic procedures as will be required to support Argo.

No representative from Miami was present, but they reported through email that they, too, are still supporting GTSPP. They were also building software to support Argo work, and this was going to be used in modernizing how GTSPP is handled. In particular, their QC functions used in support of Argo were going to be tested against BATHY and TESAC data and, if results are acceptable, the system will be used by both programs.

No representative for JAFOOS was present, but a report for the DPC meeting indicated continued support. They are also completing their new QC capabilities and these will be routinely applied.

2. Review of scientific QC and data status.

a)�� Status of scientific QC.

The 1997 data have been sent to all Science Centres. Norm reported that they had nearly completed their work. He noted that they have some work to do to correct information in the GTSPP data base for those data collected in marginal seas. These data have not been scrutinized since Warren does not feel confident in doing this function for these areas. However, the data are returned to NODC with a history record indicating they have passed through Scripps QC. The fact is that they have passed through the software but have not been scrutinized. Norm will work with Melanie to use the SCP$ parameter in the surface codes group to correctly indicate the actions performed on the data from marginal seas.

Melanie reported that the number of stations in the 1996 data returned from AOML was less than what was sent. She was not sure why this was so, but was going to investigate and would report after the meeting. Subsequently, she reported that AOML only views data from BATHYs and from XBTs, but other types are not QC'ed. Both Melanie and Yeun-ho are currently working to ensure that profiles that have not passed through AOML QC are suitably identified in the GTSPP archive at NODC.

b)�� Speed of data receipt

Statistics compiled by MEDS were not discussed at the meeting. Two figures are provided for information (in Annex 3). As can be noted, real-time reporting of TESAC data has improved. TESAC messages are dominated now by profiling floats and these data are inserted onto the GTS quite rapidly.

NODC provided some statistics about the state of the GTSPP archives. These are displayed in Annex 4. The charts of annex 4a indicate a rough (over estimate) count of the number of stations from the years 1990-1998 that will be sent back to Science centres for QC in order to provide the highest quality QC data for the upcoming WOCE UOT CD.

c)�� Status of Brest - NODC� and MEDS - NODC data exchange.

Both Loic and Melanie reported on this item. Melanie noted that the files received recently from Brest included only the first half of the year and also included some real-time data. As a consequence, Brest was rebuilding the files to include the full year. Loic raised the question that some of the real-time data they received from the French meteorological service and so they had included these with the submission. Bob noted that although MEDS thought they were reliably receiving GTS data, having files provided by JMA, USA and BSH, he agreed to look at the real-time data to see if there were any data that MEDS had not received by other channels. He would do this for a couple of years and examine the results, before looking at everything. Melanie agreed to split the real-time from the delayed mode in the Brest submissions and to send them to MEDS.

Melanie raised a question about some of the profiles originating at NODC and being returned from Brest having fewer depth value pairs than sent. This needs investigation and she and Loic will pursue this. They will report on this to Bob as soon as the issue has been resolved.

The comparison of Brest to NODC data holdings is targeted for completion by September, 2001. In preparation for the production of the V3 CD for UOT, there may be an opportunity for another comparison of databases before the approximate April "freezing" of input to the CD (see the tentative timetable in item 5 below).

Melanie reported that a few months ago some of the real-time data received from MEDS contained profiles of current speed and direction. Bob confirmed that these had come from Japanese vessels and were being reported in the TESAC code form. Melanie also noted that in delayed mode data received from MEDS in MEDS ASCII format, other profile types besides temperature and salinity were present. The software that she has to process data in this format restricts profiles to T and S and rejects stations with other types. This has caused delays in processing the data and is one of the reasons NODC is reworking the software used by GTSPP. NODC subsequently reported that it will split the incoming files so that profiles types of temperature and salinity will be processed immediately into the GTSPP data base, and other profiles will be set aside for inclusion in the new GTSPP archive when it is built.

d,e,f)�� Status on receipt of data from navies and other sources

Melanie reported that the SEAS data collected up to 2000 had been received at NODC. The processing of these data is routine and therefore much of it had already entered the GTSPP data base.

Melanie also reported that the 1998 delayed mode data from CSIRO/BMRC had been received by NODC and processing would begin shortly. Also, she reported that a submission of 1998 data had been received from AODC.

Melanie reported that some delayed mode data had been received from the Japanese. She had not looked at these data as yet and so could not provide details about its content. She remarked that some of the submissions from the Japanese in the past had been very low resolution data rather than the high resolution expected. She will look into this and report on it to Bob.

Kurt reported that he does not think that any of the data received from NAVOCEANO contains declassified data collected by the navy. The submissions appeared to duplicate much of the data the navy had received from NODC and from MEDS. He will have someone check into this, by contacting NAVOCEANO to determine what is happening. He will report on this to Bob

Loic reported that generally no data were forth coming from the French navy.

Bob noted that MEDS received a number of XBTs from the Canadian navy each year and that these were processed and sent in the regular exchanges to NODC.

3.� Semi-automated QC procedure development.

This was not discussed at the meeting. Both AOML and JAFOOS had provided reports to GTSPP and DPC concerning their activities with respect to automated QC procedures. These were discussed above in the section on scientific QC status.

4. Interaction with other programs

The only item discussed at the meeting from this section concerned the interaction of GTSPP and Argo. All agreed that the inclusion of profiling float data in the GTSPP data stream was logical. As well, users would be interested in the union of the float data with XBT data collected by other programs and this was exactly the issue that GTSPP was built to address.

At an Argo data management meeting held in Brest in October, 2000, NODC offered to maintain the long term archive of these data. The carrying out of such a function is completely compatible with the concept of GTSPP for a continuously managed archive. At a meeting of the Argo Science Team the previous week, a data management committee was formed so there was now an official body to direct the development of managing the data for the program. Activities were already underway, and it is important to exploit the momentum that data management built at the Brest meeting. Kurt agreed that NODC would act quickly to write a proposal that could be reviewed by the Argo data management committee and participants of GTSPP and that would meet both Argo and GTSPP objectives.

b)�� SOOP Line mapping products (MEDS)

This was not discussed at the meeting. MEDS has developed software that uses digitized SOOP lines to determine if a cruise has stations along these lines. This product is being distributed on a trial basis to SOOP members. If acceptable, this could also be used more generally. A sample of the product is available on request from Bob.

c)�� First J-COMM meeting (MEDS)

This was not discussed at the meeting. The first meeting of JCOMM will take place this June. The GTSPP is expected to be adopted as a JCOMM program and so will have reporting responsibilities to this committee. Likewise, experiences gained by members of GTSPP will be valuable in shaping the work program of JCOMM. Bob will be attending the meeting as part of the Canadian delegation and so can report on this afterwards.

d)�� GOOS plans and GTSPP participants

This was not discussed at the meeting. GOOS has accepted GTSPP as an Initial Operational System of GOOS and so will have reporting responsibilities to GOOS.

Agenda items a, b, c, d show how the GTSPP provides infrastructure and support to a number of programs. Because of this, and because requirements of these programs have additional or changing demands on GTSPP it was suggested that GTSPP should organize a full meeting of partiicpants. Bob agreed to work with all participants to organize such a meeting. Since many GTSPP participants are also involved in the Argo programme, and an Argo Data Management meeting is scheduled for Sep in Ottawa, this is also the proposed time and location for the GTSPP meeting. Dates are Sep 20-21 (22 if required).

e)�� Others (SEAS, JAFOOS)

Gary Soneira representing the SEAS program reported on new developments. SEAS 2002 software is under development at present. The major change will be in the way data are sent ashore. Instead of low resolution BATHYs or TESACs being sent, the full resolution profile will come ashore. On shore, BATHYs or TESACs will be created but the data can also be sent immediately to users. Gary informed the meeting that the ship board software will be completed by the end of this year. However, other software was needed at the stations receiving the data sent from the ship and this would take additional time to complete. No date for this was specified.

An additional point was made that SOOP had requested that the data when they come ashore be assigned a unique tag for every station. Gary said that the current plan was to use the Lloyds hull number plus a time stamp. Kurt noted that while this was suitable for SEAS ships, in a broader context, there were platforms such as moored buoys, and aircraft that did not have a Lloyds number. Since the idea of a unique tag is of general benefit and would likely be promoted at JCOMM, it would be better if a more general solution was used. He suggested using the platform's WMO identifier and time stamp and this was supported by Bob. Gary agreed to take this back to the software developers.

Gary also noted that the SEAS software was expected to be general enough to handle other kinds of data not currently transportable on the GTS but becoming of increasing interest. Bob noted that these additional types of data, the unique station tag and the rapidity at which full resolution data from SEAS will be passed to GTSPP requires consideration and will likely require some changes to operations. Gary agreed to provide Bob with more details, and Bob would write a proposal about how GTSPP should handle this. The draft should be circulated by September, with finalization as soon as possible after to permit time for software changes if required.

In the UOT report to DPC submitted by JAFOOS, it was reported that they were implementing a DODS server for the Indian Ocean data. There was limited discussion of what the general action of GTSPP should be. (At the subsequent DPC, all DACs were encouraged to implement a DODS server for the data they handle). NODC at one time had hosted a DODS server but presently it was not running.� Bob agreed to approach JAFOOS to see if they would be willing to host the entire GTSPP data on their server. (See also item 6a below on the NODC Climate Data Portal).

5. UOT CD

a)�� Reported problems with the latest CD

JAFOOS had been contacted by a Chinese scientist noting some peculiarities with some profiles from the Indian Ocean.� In the course of responding to this question, Lisa and Ann examined the V2 CD and generated a list of a number of issues that needed to be resolved. Some of these issues were of concern to all DACS and so were raised at the DPC meeting.

The specific issues and responses are as follows.

i)� CSIRO codes are different from WOCE codes.

JAFOOS reported that there were a number of profiles for which some or all CSIRO action codes (in the history structure of the GTSPP record) were missing. Melanie Hamilton explained that this appeared to be a misunderstanding about what information should be included on the V2 CD. She noted that when data are first received from CSIRO they contain the CSIRO information and that this is stored in the GTSPP data base. Later, when yearly files are sent to JAFOOS and then received after QC, the actions are recorded using the Science Centre agreed codes for WOCE. Melanie also suggested that some codes returned in the yearly files are marked with a CSIRO identifier rather than WOCE. In any case, the WOCE codes are preserved in a different location from the CSIRO codes and only the WOCE information was included on the CD. NODC agreed to include the complete history information for V3. After subsequent discussions with Rick and others in Australia, it was agreed that NODC should retrieve the data for one or two of the stations noted with the problem as they will for the V3 CD. The results should be sent to Ann to check that all of the relevant information is present.

ii) Duplicates on CD

CSIRO noticed that there were some delayed mode profiles marked as duplicates of other profiles but that these appeared on the V2 CD. Melanie confirmed that this was true. She explained that when delayed mode data were marked as duplicates, software did not change another field in the NODC archive (the ACTIVE field) to mark the station not to be retrieved. As a consequence, the duplicate stations were extracted and appeared on the CD. This will be corrected.

iii) Faulty FORTRAN code created

JAFOOS noted that the FORTRAN code generated by the code generator on the CD did not produce usable code. Charles Sun agreed to look into this. This was also raised at DPC and Nathan Bindoff said that he had had subsequent discussions with Rick to clarify the matter. Nathan was tasked by DPC to rewrite the netCDF primer for the V3 CD.

In a follow up conversation with Rick, it seemed that we would need to be sure that the FORTRAN code generated did meet FORTRAN rules. It may be that some of the variable names in the netCDF structure will need to be changed. Charles should work with Rick to sort this out.

iv) Mixture of data on V2 with WOCE QC and without

This is true but was not a mistake. The intention for V3 is to pass all data in the WOCE period (1990-1998) through scientific QC for the final V3 CD. A timetable for CD production was proposed and is included in Annex 5. This was also brought up at DPC for the advice of the larger panel. They endorsed that intention of carrying out QC on all data from the WOCE period. However, they also noted that data from before and after WOCE provided a wider context for the interpretation of the WOCE data and so expressed a desire that such data also be included. They endorsed the suggestion, made by Rick, that data not having passed through scientific QC should be placed in a separate directory so that there could be no confusion about the level of QC carried out. This alternate directory tree will include data from outside the WOCE period and not scientifically QC'ed as well as data from marginal seas (see item 1 above) and any other data that could not meet the cutoff for scientific QC as indicated in the timetable.

To determine the relative amounts of such data, NODC agreed to produce some statistics by year of the number of WOCE QC'ed and non-WOCE QC'ed data on V2. This would also be used on the V3 CD with updated figures (see Annex 4).

v) Fall rate correction

Rick noted that the data on V2 was a mixture of data with and without the XBT fall rate correction applied. This is true and was discussed and agreed to at the last DPC meeting. It was also agreed that UOT would explore what could be done to correct the data so that as much as possible XBT data on the V3 CD would be reported with corrected depths.

Bob used Canadian XBT data in MEDS archive as a test bed for a study of this problem. In summary, and as expected, he found that for a substantial fraction of historical data, there was insufficient information to make an unequivocal determination of what, if any, correction should be applied. He also noted that the conversion from old to new fall rates was completely reversible so that if a mistake was made, the original reported depths could be restored. He also recommended that when measures were taken to correct fall rates, that every station receive a new "tag" indicating what action was taken. He argued that this was necessary to ensure that there could be no doubt in global archives if a station had been corrected or not.

Bob also raised this issue at DPC and it was noted that though it may not be possible to be sure about certain corrections needed, if a reasonable guess on the probe type could be made, and this used as the basis of a correction, the resulting profile may still be improved in terms of the uncertainty of the depth. The DPC offered no advice on this, suggesting instead that UOT science centres discuss this and come to some agreement. Rick Bailey was of the opinion that this may, in fact, be a reasonable strategy. Bob will contact other GTSPP science centres for their opinion.

Bob had noted at the GTSPP meeting that most of the work involved is in tracking down the missing information for stations. He agreed to co-ordinate getting this work done using the resources of both science centres and data centres of GTSPP.

The question was also raised about the validity of the QC flags set before the profile had depths corrected. Again the DPC requested UOT/GTSP to discuss this and come to the best solution. In a subsequent conversation with JAFOOS and CSIRO they did not think reQC of the data would be necessary. Bob will contact other science centres of GTSPP about this.

As an aside, Bob was requested to look at the incidence of the use of old or new fall rate equations in the real-time data stream since the code change in November 1995 which permitted these to be reported for BATHYs. He agreed to do this.

After subsequent discussions with Bailey, Bob agreed to write a document that lays out the strategy for correcting the fall rates. It was suggested that a two pronged approach may be necessary with an automated procedure using existing information to correct what can be managed for the V3 CD and then a follow up action to try to acquire additional information from originator countries. Bob will circulate this to others of GTSPP for comment.

vi) Some records on V2 had non-WOCE records generated after WOCE QC was carried out.

Melanie explained that indeed this was true, and stemmed from how CSIRO handled data from the first 3.7m of an XBT profile, and the way NODC software works. In brief, CSIRO substitutes 99.999 for all temperatures in the top 3.7m, writes the original values in the history structure and puts a flag of "5" (changed value) against the upper values in the profile. NODC software actually recontructs some of the history information and has rigid rules about what sort of history records must be present and in which order. Consequently, NODC process codes were added in the history structure to record the changed flag (from 1 to 5) for the upper 3.7 m values. This will be corrected for V3. Discussions with Ann suggested that she would be willing to modify her software to write the necessary record so that the NODC code need not be altered but this problem fixed.

vii) Some profiles showed a zero as the deepest value.

Melanie reported that this was true, but was inconsistent in its application. Subsequent to the meeting it was found that nulll values were present at the last depth for several data set s received from CSIRO. Melanie and Ann will resolve this problem.

viii) Need an articulation of rules for handling data returned from Science centres

�Melanie explained that when data are returned, the database id is used to match the incoming to an existing data base record. After the match is made, NODC software also examines the location, dates and profiles to ensure there are matches. In some cases, this has detected problems. In any case, NODC will draft rules for deletes/replacements for circulation

ix) We need a timetable that sets aside 2 months for SC review

The timetable was discussed and is included in Annex 5. Subsequent discussions with Rick, Ann and Lisa agreed to the schedule.

x) Acknowledgements

Rick noted that it was very important that JAFOOS and CSIRO receive proper acknowledgement for their contributions to the UOT CDs. Bob brought this up at DPC and it was agreed that the WOCE IPO would provide recommended text to go on each CD urging users to cite the specific CD when they used the data or information in subsequent analyses.

b)�� Review V2 CD content

This item was not discussed at the meeting, apart from the problems documented above. As noted above, data that have not undergone WOCE QC will be segregated from the data that have. This will alter the layout of the CD. In addition, Bob would like to expand the suite of products as well to include any more developed since the last CD was issued. He will contact science centres directly seeking contributions and comments on the layout of the CD. Note that DPC also discussed the general layout and proposed a standard template for a welcome page. Bob will assemble a new layout for the CD and circulate this to participants in the time noted in Annex 5.

6. New Directions

a)�� NODC and Ocean Data Portal

Charles Sun gave a brief report on the development of the Climate Data Portal. He noted that they were still waiting to hear if the latest proposal had been funded. In the interim, he could make available to URL from which people could download the java application and the URL to show the web browser version. He cautioned that since this was still under development, there were likely still problems. He also requested that the URL not be circulated beyond the members of GTSPP.

b)�� NODC GTSPP CMD System Enhancement Plan

Kurt and Mike Simmons provided some material that described NODC's plans for enhancing the CMD. Participants were invited to read the material and offer comments at a later date.

c)�� Web pages review

This item was not discussed at the meeting. However, Charles Sun showed how NODC intended to improve access to GTSPP data through the Climate Data Portal. Bob will circulate his ideas to all for comments and suggestions.

Other Business

a)�� Review of items from last year's meeting.

This was not discussed at the meeting. Participants are asked to review items and report these for inclusion in the report.

b)�� Other variables in MEDS ASCII format

See item 2 above.

c)�� OOPC

Bob alerted participants to a plan being formulated by Neville Smith of OOPC for a conference on data management. This would be structured along the same lines at the Ocean Observations Conference 1999. Bert Thompson reported that Neville's ideas had been endorsed by GOOS SSG and provided a copy of Neville's presentation. Bob remarked that he hoped and expected that members of GTSPP would contribute to the planned workshop.

Annex 1: Agenda

1. Status

�� Review of continuing support at each site.

2. Review of scientific QC and data status.

a)�� Status of scientific QC.

b)�� Speed of data receipt (MEDS, NODC)

c)�� Status of Brest - NODC data exchange.

d)�� Status on receipt of data from navies (NODC)

e)�� Status of Japanese data. (NODC)

f)�� Other data receipts? (NODC, Brest)

3.� Semi-automated QC procedure development.

�� Status: How is this going at AOML, JFOOS/CSIRO/BMRC?

4. Interaction with other programs

a)�� Developments for Argo and implications for GTSPP.

b)�� SOOP Line mapping products (MEDS)

c)�� First J-COMM meeting (MEDS)

d)�� GOOS plans and GTSPP participants

e)�� Others

SEAS

JAFOOS

5. UOT CD

a)�� Reported problems with the latest CD.

b)�� Given DPC directions for the 3rd WOCE edition, what are implications for us?

c)�� Review CD content to ensure we are happy with the layout and products.

6. New Directions

a)�� NODC and Ocean Data Portal

b)�� NODC GTSPP CMD System Enhancement Plan

c)�� Web pages review

Other Business

a)�� Review of items from last year's meeting.

b)�� Other variables in MEDS ASCII format

c) OOPC

Annex 2: Attendees

Melanie Hamlton - NODC (melanie@nodc.noaa.gov)

Tim Boyer - NODC (boyer@nodc.noaa.gov)

Mike Simmons - NODC (michael.simmons@noaa.gov)

Charles Sun - NODC (charles.sun@noaa.gov)

Kurt Schnebele - NODC (kschnebele@nodc.noaa.gov)

Loic Petit de la Villeon - IFREMER (Loic.Petit.De.La.Villeon@ifremer.fr)

Nathan Bindoff - Aust Antartic Research Centre (n.bindoff@utas.edu.au)

David Legler - US CLIVAR Office (legler@usclivar.org)

Bert Thompson - GOSIC U.Delaware (bthom2807@aol.com)

Bob Keeley - MEDS (keeley@meds-sdmm.dfo-mpo.gc.ca)

Norm Hall - NODC/Scripps (nhall@nodc.noaa.gov)

Gary Soneira - OAR/Miami (goar.soneira@noaa.gov)

Discussions were subsequently held with Rick Bailey and Lisa Cowen of JAFOOS and Ann Gronell of CSIRO.

Annex 3:

Statistics on data receipts from MEDS

Annex 4a: NODC statistics on QC carried out on GTSPP data.

�

These graphs show the number of unique stations by year and ocean. Those marked DQC have undergone Data Centre QC only, while those marked SQC have undergone Science Centre QC. It should be noted that QC of data from 1997 and subsequent years is not completed. Also, I the Atlantic, AOML carries out QC on BATHY and XBT profiles only, though other types are sent to them.

Annex 4b: Proportion of unique real-time compared to delayed mode stations in the GTSPP archives.

�

These graphs show the number of unique real-time (all types of data) and delayed mode stations by year and ocean. Those marked tRT are the total real-time stations, while those marked tDM are the total delayed mode stations. Typically the proportion of real-time to delayed mode stations is higher in recent years. Also, the large number of real-time stations in the Pacific (and in the last couple of years in the Atlantic) reflect the data coming from the TAO array. In recent years, a large number of Argo floats dominate the real-time counts especially in the Atlantic.

Annex 4c: This shows information about GTSPP data processing at NODC. The "Best Copy" data are the unique stations (either real-time or delayed mode) in the archive. The "successful" and "No updates" show how much data returned from Science centres were moved back into the GTSPP archive, or had some problem on update. Finally, the group labeled "Deletes" are records that were removed from the archives because of duplicates identified or other reasons.

Annex 5: Timetable for V3 UOT production

Action�� Start Date

Complete V3�� Sep, 2002

Start building V3�� Aug, 2002

Resolved any problems identified�� Jul, 2002

Sent draft CD to SCs for scrutiny�� May, 2002

Corrected fall rates�� Apr, 2002

Received comments on V3 design�� Jan, 2002

Received 1990-98 data from SCs�� Dec, 2001

Sent draft V3 design to all�� Nov, 2001

Sent 1990-98 unQC'ed data to SCs�� Oct, 2001

Received 1998 data from SCs�� Sep, 2001

Received ideas for new design�� Sep, 2001

Addressed IO DAC problems�� Sep, 2001

Reconciled 1990-1998 differences with Brest�� Sep, 2001

Sent 1998 data to SCs�� Jul, 2001

Received 1997 data from SCs�� Jun, 2001

Sent 1997 data to SCs�� Apr, 2001

Annex 6: Extract of actions resulting from the meeting.

1. From item 2a. Review of scientific QC and data status.

Norm will work with Melanie to use the SCP$ parameter in the surface codes group to correctly indicate the actions performed on the data from marginal seas returned from Scripps.

2. From item 2a. Review of scientific QC and data status.

Melanie and Yeun-ho will ensure that profiles that have not passed through AOML QC are suitably identified in the GTSPP archive at NODC.

3. From item 2c. Review of scientific QC and data status.

Melanie agreed to split the real-time from the delayed mode in the Brest submissions and to send them to MEDS. Bob agreed to look at the real-time data to see if there were any that MEDS had not received by other channels.

4. From item 2c. Review of scientific QC and data status.

Melanie and Loic will investigate profiles originating at NODC and being returned from Brest having fewer depth value pairs than sent.

5. From item 2c. Review of scientific QC and data status.

NODC will split the incoming files so that profiles types of temperature and salinity will be processed immediately into the GTSPP data base, and other profiles will be set aside for inclusion in the new GTSPP archive when it is built.

6. From item 2d. Status on receipt of data from navies and other sources.

Melanie will look into submissions from the Japanese to determine the vertical resolution of the data. In the past some had been very low resolution data rather than the high resolution expected.

7. From item 2d. Status on receipt of data from navies and other sources.

Kurt will have someone check into data submissions from NAVOCEANO to see how much duplicates data already held at NODC.

8. From item 4: Interaction with other programs

Kurt agreed that NODC would act quickly to write a proposal on how NODC would act as the long term archive for Argo, so that it could be reviewed by the Argo data management committee and participants of GTSPP and that would meet both Argo and GTSPP objectives.

9. From item 4e: Others (SEAS, JAFOOS)

It was suggested that using the platform's WMO identifier and time stamp was a more widely applicable unique tag in SEAS software Gary agreed to take this suggestion back to the software developers.

10. From item 4e: Others (SEAS, JAFOOS)

Gary agreed to provide Bob with more details about other kinds of data that could be handled in SEAS software, and Bob would write a proposal about how GTSPP should handle this. The draft should be circulated by September, with finalization as soon as possible after to permit time for software changes if required.

11. From item 4e: Others (SEAS, JAFOOS)

Bob agreed to approach JAFOOS to see if they would be willing to host the entire GTSPP data on their server.

12. From item 5ai: CSIRO codes are different from WOCE codes

It was agreed that NODC should retrieve the data for one or two of the stations noted with the problem as they will for the V3 CD. The results should be sent to Ann to check that all of the relevant information is present.

13. From item 5aii: Duplicates on CD

Melanie confirmed that this was true. NODC will correct this for V3.

14. From item 5aiii: Faulty FORTRAN code created

We need to be sure that the FORTRAN code generated did meet FORTRAN rules. It may be that some of the variable names in the netCDF structure will need to be changed. Charles will work with Rick to sort this out.

15. From item 5aiv: Mixture of data on V2 with WOCE QC and without

This is true but was not a mistake. NODC will build V3 containing all available data from both WOCE years and after. The data not having passed through scientific QC will be placed in a separate directory so that there could be no confusion about the level of QC carried out. This alternate directory tree will include data from outside the WOCE period and not scientifically QC'ed as well as data from marginal seas (see item 1 above) and any other data that could not meet the cutoff for scientific QC as indicated in the timetable.

16. From item 5aiv: Mixture of data on V2 with WOCE QC and without

NODC agreed to produce some statistics by year of the number of WOCE QC'ed and non-WOCE QC'ed data on V2. This would also be used on the V3 CD with updated figures.

17. From item 5av) Fall rate correction

DPC was of the opinion that though it may not be possible to be sure about certain corrections needed, if a reasonable guess on the probe type could be made, and this used as the basis of a correction, the resulting profile may still be improved in terms of the uncertainty of the depth. Bob will contact GTSPP science centres for their opinion.

18. From item 5av) Fall rate correction

Bob agreed to co-ordinate tracking down probe information using the resources of both science centres and data centres of GTSPP.

19. From item 5av) Fall rate correction

JAFOOS and CSIRO did not think reQC of the data would be necessary after the depth correction was made. Bob will contact other science centres of GTSPP about this.

20. From item 5av) Fall rate correction

Bob agreed to look at the incidence of the use of old or new fall rate equations in the real-time data stream since the code change in November 1995.

21. From item 5av) Fall rate correction

Bob agreed to write a document that lays out the strategy for correcting the fall rates. He will circulate this to GTSPP members for comment.

22. From item 5avi) Some records on V2 had non-WOCE records generated after WOCE QC was carried out.

This will be corrected for V3. Discussions with Ann suggested that she would be willing to modify her software to write the necessary record so that the NODC code need not be altered but this problem fixed.

23. From item 5avii) Some profiles showed a zero as the deepest value.

Melanie reported that this was true. It was found that nulll values were present at the last depth for several data set s received from CSIRO. Melanie and Ann will resolve this problem.

24. From item 5aviii) Need an articulation of rules for handling data returned from Science centres

NODC will draft rules for deletes/replacements in 2 weeks for circulation

25. From item 5b: Review V2 CD content

Bob will contact science centres directly seeking contributions and comments on the layout of the CD. Bob will assemble a new layout for the CD and circulate this to participants.