INTERGOVERNMENTAL OCEANOGRAPHIC COMMISSION
THIRD SESSION OF IOC-WMO STEERING GROUP ON THE GTSPP
Ottawa, Canada, 15-19 November 1993
1. Opening of the Meeting
2. Administrative Arrangements
3. Real-time Systems Development
4. Delayed Mode and Historical Data Systems Development
5. Publications and Products
6. Co-operation With and Services to Other Programmes
7. Software and System Improvements
8. Review of the Action Plan
9. Other Business
10. Approval of the Summary Report
11. Dates and Place of Next Meeting
12. Closure of the Session
II. GTSPP Implementation Plan for 1994-1995 - Recommendations
III. List of Participants
IV. List of Documents
V. Progress by Participants in Implementing Aspects of the Delayed Mode System
VI. IGOSS and IODE Systems Data Management Objectives in Service for Goos
1. OPENING OF THE MEETING
Dr. J.R. Wilson, the Chairman of the IOC-WMO Steering Group on the GTSPP opened the Third Session of the Group at 9:30 a.m. on 15 November 1993 at the Headquarters of MEDS in Ottawa, Canada.
Dr. Wilson invited the First IOC Vice-Chairman, Mr. G. Holland, to address the Session.
On behalf of the Department of Fisheries and Oceans of Canada, Mr. Holland welcomed the participants to the Session and to Ottawa. He emphasized the importance of the Session at the time when oceanography in general, and ocean data management in particular, are facing new challenges in many time and space scales, from climatic to seasonal variations, from the atmosphere to the bottom of oceans on many fronts, from science to economical and political interests in the dramatic increase of volume of data from satellites and instruments with very high data rates, and in many geographical areas, from the Polar Oceans to the tropics. Mr. Holland referred to the decisions of UNCED and stressed that the climate change issues have impacted on the social and political climate and forced governments to look at longer-term agendas than is usual for politicians. The decisions to be made are expensive and far-reaching and governments are uncertain on how to react. Ocean data is an important element in reducing scientific uncertainty, in providing the basis for effective sustainable development.
Mr. Holland then focused attention of the participants on the need for global ocean observations and data management under GOOS and pointed out that GTSPP is leading the way to the development of a prototype data management system for the future in which institutes and scientists will respect the efficiency and effectiveness of placing collected data into the global community as soon as possible. He emphasized that the tradition of scientists retaining their data until after the publication of all their researches must be reviewed. The needs are so urgent that data sharing arrangements that protect the rights of the scientists must become the norm.
Finally, Mr. Holland wished the Session every success and a good stay in Ottawa.
Dr. Wilson welcomed the participants to the Session and summarized the objectives of the Session:
(i) be ready to cope with increased volume of data;
(ii) promote data services for customers;
(iii) consider ways of co-operation with international organizations which have or may have interest in GTSPP, e.g., ESA, PICES and others;
(iv) improve further data management technologies to make GTSPP procedures of quality control, exchange and monitoring smoother and more effective.
He informed the Session about his participation at the Seventeenth Session of the IOC Assembly and the most encouraging comments made by the Assembly participants on GTSPP, which considered the project as an outstanding achievement. Dr. Wilson emphasized that he fully agrees with the Secretary IOC, Dr. G. Kullenberg, who thinks that GTSPP is leading the way and is crucially important to the development of the model for data management that must be built to meet our needs in the next century.
Over the past few years, and particularly since the Rio Conference, the number of client programmes needing oceanographic data and data services has been increasing significantly. These programmes have a particular job to do. It is important that they can concentrate on that job, and do not have to re-invent and re-develop international data management.
If one looks at the number of programmes that IGOSS and IODE will have to serve in the future, it looks daunting. However, these clients are all looking for the same services; the provision of ocean data and products from well organized, standardized databases of known quality. Dr. Wilson stressed that if we concentrate on developing and implementing good data management practices that take advantage of the technologies available to us, we will go a long way towards servicing these programmes. To reach a satisfactory level of performance, many data centres will have to increase their throughput of data by one to two orders of magnitude. In fact, the technology exists to do this, and it is generally affordable.
Finally, Dr. Wilson expressed hope that the Session would consider the ways to attract new Member States not actively participating in GTSPP to assist GTSPP by accepting a variety of responsibilities in the development of delayed mode data monitoring and in the historical aspects of this project.
2. ADMINISTRATIVE ARRANGEMENTS2.1 ADOPTION OF THE AGENDA
The Agenda for the Session was adopted as given in Annex I.
2.2 DESIGNATION OF A RAPPORTEUR
The Meeting designated Ms. M. Hamilton (USA) and Mr. D. Hamilton (USA) to be the Rapporteurs for the Session.
2.3 CONDUCT OF THE SESSION
The Technical Secretary, Dr. I. Oliounine, reviewed the arrangements and documentation for the Session. The List of Documents is given in Annex IV and the List of Participants is Annex III of the Summary Report.
3. REAL-TIME SYSTEMS DEVELOPMENT3.1 REVIEW OF THE EXISTING OPERATIONAL DATA FLOW
Document IOC-WMO/IODE-IGOSS/GTSPP-III/6, summarizing the data flow within GTSPP at present was presented by Mr. J.R. Keeley of MEDS. He noted that besides the participants of GTSPP, there were more outside users being given direct access to data prepared either weekly or monthly. Amongst users were ocean modelers, meteorological agencies, fisheries agencies and users in the satellite community. He also noted that besides input to the real-time data flows from the GTS there were two other sources supplying data in BATHY format albeit too late for insertion of these data onto the GTS.
The meeting noted that there was some confusion in the use of the term delayed mode data between the various data flows. In the WOCE and TOGA contexts, delayed mode data are usually considered to be the full resolution data from which BATHY and TESAC reports are coded and which are processed within a few months and submitted to replace the BATHY and TESAC reports as a more accurate and useful version of the observation. For IGOSS delayed mode data refer to data which are received after the 30 day cutoff for transmission on the GTS. Historical data is a third term that is used. This term generally refers to datasets that are rather older than delayed mode data, and which may or may not already have been incorporated in the IODE and WDC system databases.
The meeting also noted that the terms 'low resolution' and 'high resolution data' were useful in describing datasets. When an XBT or a CTD observation is coded for transmission on the GTS in the BATHY or TESAC code forms, these reports can be said to be a low resolution version of the observation. When the fully processed data that has a value of temperature or salinity every meter or so is prepared, this can be described as high resolution data. It is important to be precise in using these terms in order that users can understand the characteristics of the data they are working with. The terms 'low and high resolution data' were considered to be more descriptive of the particular version of the data than 'real-time, delayed mode, and historical data' and GTSPP participants were encouraged to also use these terms in describing their datasets.
The Second Session of the IOC-WMO Steering Group on GTSPP considered the inclusion of surface variables collected by ships underway and reported on the GTS using the IGOSS TRACKOB code form. That meeting called for the inclusion of TRACKOB temperature and salinity data in the GTSPP databases to make these data widely available to the marine community. The GTSPP centres responsible for this action were unable to accomplish this goal before GTSPP-III.
The Meeting reviewed the needs in relation to TRACKOB data and decided that there was still a requirement to distribute these data. The method chosen for distribution was to treat an observation at a location as an oceanographic station with surface observations only. A stream identification is to be assigned to these stations that identifies them as TRACKOB data as opposed to BATHY or TESAC. The data will then be merged with the real-time data files provided 3-times weekly to the CMD in NODC and to all real-time users receiving the 3-times weekly files.
The Meeting identified a further requirement to include TRACKOB data in the monthly monitoring reports on the GTS data flow produced by MEDS. The purpose of doing so was not only to make sure that TRACKOB data are not going missing on the GTS, but also to advertise the vehicle of the TRACKOB code form for circulating along track surface observations in real-time. It is hoped that by doing this, other countries will be persuaded to report data in TRACKOB, therefore increasing the amount of surface data available in real-time. MEDS agreed to include TRACKOB data in the monthly monitoring reports for this purpose.
The Meeting also decided that participants should make any along track temperature and salinity data along with any other parameters observed simultaneously that are not reported via TRACKOB available to the GTSPP. This is to be accomplished by coding the data in the GTSPP format and providing it to MEDS in the manner that Japan and the US Navy make fisheries and declassified data available. France in particular indicated that there are data being collected that could be provided in this manner. Copies of all TRACKOB data and surface salinity data should be forwarded to the surface salinity DAC in ORSTOM, Brest.
The Meeting discussed the question of temperature and salinity data collected using undulating instruments such as sea soar and batfish. In particular the TOGA COARE project has collected such data and wish to include it in a comprehensive dataset that will be made available from that project. Data from undulating instruments has a different basic structure from traditional oceanographic data types. Traditionally, oceanographic data can be treated as profiles of variables vs depth at a point, time-series of variables at a point in 3-dimension (latitude, longitude and depth), or observations along a track. Data from undulating instruments have in some cases been treated as profiles rather than as a function of time, geographical position and depth. The difficulty in the latter case is that the data are then very difficult to merge with ocean station data and require separate data management software and database structures. It was noted that the GETADE already had a responsibility to formulate structures and methods for handling these complex data and it was therefore decided to refer the problem to this group along with recommendations as to short- and longer-term actions.
The short-term action is for the Chairman of the GETADE to establish contact with the person responsible for TOGA COARE data management at the International Project Office in Boulder, Colorado, USA. The purpose of the contact is to establish co-operation in addressing methods of handling, storing and disseminating the TOGA COARE sea soar data. The first question to be addressed is whether the requirements for these data could be met by encoding the data as profiles in the GTSPP format with additional variables that describe the characteristics of the instrument path being stored in the surface parameter table. For example, the starting and ending latitudes and longitudes could be assigned codes and stored in this table.
The second activity for the GETADE should be to assemble information on how present national activities process, store and disseminate data from undulating instruments. This information will be useful to the GETADE in its final deliberations on the subject of data from undulating instruments, and to suggest options to the TOGA COARE scientists, should the solution of using the GTSPP format not prove acceptable.
In addressing these questions, the GETADE should include AODC, BODC, US NODC,
the WOCE/TOGA Sub-surface Data Centre and the German BSH (Dr. G. Becker) in the
consultations as these centres already have experience to contribute.
3.3 NAVY DECLASSIFIED DATA
For the past few months MEDS has been receiving US Navy data that has been declassified by the Fleet Numerical Oceanography Center. The volume is of the order of 1,000 stations per month.
Data holdings are reviewed by FNOC every two weeks. The data that are declassified are converted to BUFR format and included in a file that is sent to MEDS via Internet. The file typically includes data that have been collected from 30 days to 6 years ago and may include complete cruises or single stations. MEDS has implemented the necessary systems and is in a position to forward the data to the CMD in Washington on a 3-times weekly basis. It is expected to start doing this almost immediately. There is presently a backlog of 12 to 15 bi-weekly files which MEDS will also process and forward to the CMD.
These observations are in fact low resolution data since the observations being declassified are those that were reported in BATHY code. The Meeting recommended that the US NODC contact the US Navy and clarify whether the high resolution version of these data might be available and, if so, to encourage its submission for research use. The US and Australian navies have been declassifying and submitting data to the international community for some time to assist with the Global Research Programmes. GTSPP encouraged the navies in other countries to review their data holdings and to submit their oceanographic data to support global research programmes. The Technical Secretary recommended that those countries in which navies have begun to declassify selected XBT data provide formal letters of announcement to this effect to the Secretary IOC. The USA and Australia volunteered to provide such letters. IOC can then use these letters to encourage the navies of other Member States to begin to declassify their data.
The representatives from the Russian Federation and Germany announced that their navies were also in the process of declassifying data. Mr. N. Mikhailov mentioned that the first contribution of Russian naval BATHY data was made to the GTSPP in 1993 for inclusion in the GTSPP database. The representatives of Germany pointed out a problem with the German data, in that navy ships of these countries do not have call signs. It was decided by the Meeting that in this case and whenever possible, a unique identification code should be assigned to ships when the call sign is not available rather than assigning the call sign SHIP. Such a procedure enables the GTSPP to perform the quality checks related to position and speed made good, and to review near neighbour profiles. The GETADE was requested to investigate this requirement further and recommend procedures for assigning such codes.
3.4 DATA FLOW MONITORING
The Fourteenth Session of the IOC Committee on IODE considered the subject of real-time and delayed mode data flow monitoring and decided that it was necessary to design specific data flow monitoring systems for each distinct type of data. One of the types chosen was the temperature and salinity data managed within GTSPP. The Meeting discussed at length the method proposed in Document IOC-WMO/IODE-IGOSS/GTSPP-III/7 and the report on TOGA-WOCE data rescue prepared by the WOCE-TOGA Sub-surface Data Centre.
It was decided that the scheme as presented was workable in general, but that some bounds should be established for an initial period to make the project more manageable. The bounds were to limit the project to the data collected by the present 7 active participants, and to separate the information used to monitor the real-time and delayed mode data into two distinct types in order to avoid confusion with the present IODE and IGOSS monitoring systems.
The goals of the monitoring and the monitoring products to be produced are as described in the working document referred to above namely; Document IOC-WMO/IODE-IGOSS/GTSPP-III/7, goals (1-5). To accomplish these goals the following actions and responsibilities were decided.
For the real-time data, MEDS will prepare an annual report along the lines of that presented in the TOGA/WOCE data rescue document. This report will be circulated to the NODCs in the 7 active GTSPP Member States. The information in the report will be used by these NODCs to pursue acquisition of the delayed mode versions of the real-time data. In addition, the adequacy of the information in the report for use in identifying the data in question and any conflicts of this information with the information to monitor data flows in other projects, such as GODAR, will be assessed and resolved.
For the delayed mode data, the 7 Member States will establish more active contacts with national agencies that collect oceanographic data to improve the submission of Cruise Summary Reports (CSR) as called for in IOC-ICSU Manuals and Guides No. 9 on IODE. These CSRs should be submitted to NODCs who will in turn submit copies to WDCs Oceanography. It is expected that one or more WDCs will enter these CSRs into electronic systems that can provide an annual database to the 7 NODCs.
It was recommended that IGOSS NOC in the German BSH will survey the IGOSS National Representatives in the 7 Member States to determine what surface observational data are being collected and reported in TRACKOB to ensure that the headers for all such data are known and that these data are reaching the GTSPP via GTS and MEDS.
The report prepared by MEDS above, the CSR database provided by the WDC system and the TRACKOB monitoring reports described under Agenda Item 3.2 will provide the inventory of data collected as described in goal one of the monitoring scheme in Document IOC-WMO/IODE-IGOSS/GTSPP-III/7. The NODC of the USA will complete the development of the software system now in its final testing that assigns the real-time data assembled by the GTSPP to the WOCE/TOGA lines. When operational, this product will be prepared on a quarterly basis and forwarded to WOCE, which will satisfy goal two. Goal three is already being satisfied by the monthly GTS data flow analyses now provided by MEDS. MEDS will continue to produce these analyses. Goal four calls for a report every 6 months that describes the flow of data through the WOCE UOT Centres. The US NODC as the operator of the CMD, which is a hub of this data flow, will work with the WOCE IPO by providing information for the report requested by WOCE for future updates to the WOCE Data Management Handbook. For goal five, MEDS will perform an analysis of the information produced for the 4 goals above, each year and publish the results in the GTSPP Quarterly Report Series to provide an overview of the flow of real-time and delayed mode data, and identify problems and priorities for solution.
The IGOSS Operations Co-ordinator has been actively working with MEDS and GTSPP to realize several goals in real-time data monitoring. Through these efforts, several significant problems have been solved and the corresponding data flow and consequently the database has been improved. Monthly IGOSS statistics have been elaborated and distributed over Omnet, particularly for the benefit of WOCE. This monthly report is used to identify data flow problems, update call signs and the ship information database maintained by the Co-ordinator. Data flow problems between France (LFPW) and FNOC, Monterey, have been solved, as well as the routing and correction of headers for salinity data from Argentina.
4. DELAYED MODE AND HISTORICAL DATA SYSTEMS DEVELOPMENT4.1 PROGRESS ON DEVELOPMENT OF THE CONTINUOUSLY MANAGED DATABASE (CMD)
The US NODC reviewed the development and implementation plans for the continuously managed database.
The NODC implemented the first version of the GTSPP CMD as a file management system on a VAX minicomputer. The datasets were loaded into the CMD and have been distributed to the WOCE UOT DACs on a monthly schedule since January 1990. In July 1992, NODC migrated the application to a relational database system using a SHAREBASE database machine. Since then, the CMD has undergone another migration to a Silicon Graphics Workstation, using the Sybase Database Management System. This last migration occurred in November 1993. It was carried out because the workstation solution is more cost-effective, and with greater computing power.
The following functions to support the CMD are operational at the time of the GTSPP-III meeting.
(i) Accept real-time GTSPP data from MEDS;
(ii) Distribute the real-time data to the WOCE UOT DACs on a monthly basis. (It was noted that Scripps and AOML will no longer QC the real-time datasets after December 1993. Instead, MEDS will carry out this work and alert both centres of problems. The AODC will carry out a comparative quality control on the Indian Ocean data for a short period in 1994 to validate MEDS procedures for that data.);
(iii) Accept delayed mode data after GTSPP QC;
(iv) Distribute complete real-time and delayed mode datasets to the WOCE UOT DACs annually as required by the WOCE UOT data management plan;
(v) Provide information on request about the CMD and the GTSPP/WOCE UOT dataset;
(vi) Provide data to users on request.
The following functions to support the CMD are not operational in the NODC at the time of the GTSPP-III meeting.
(i) Acceptance and input of flagging information for the real-time data returned from the WOCE UOT Science Centres;
(ii) Identification and flagging of real-time data in the database for which the delayed mode data has been received, has passed GTSPP QC, and has been loaded into the database;
(iii) Acceptance and input of the scientific QC flags for the annual datasets from the WOCE UOT Science Centres;
(iv) Acceptance and input of historical data after GTSPP QC.
The schedule for full implementation of the GTSPP CMD in NODC is as follows:
(i) 1 January 1994 - identification and flagging of data in the database for which the real-time data have been replaced by delayed mode data;
(ii) 1 March 1994, - acceptance and input of flagging information for the real-time data from the Science Centres;
(iii) 1 June 1994 - Complete reformat of the 1990-1992 delayed mode data;
(iv) 1 October 1994 - Complete GTSPP QC on the 1990-1992 delayed mode data;
(v) 1 November 1994 - Add the 1990-1992 delayed mode data that has been passed through GTSPP QC to the CMD;
(vi) 1 December 1994 - Distribute the annual datasets for 1991-1992 to the WOCE UOT Science Centres;
(vii)1 February 1995 - Distribute the annual dataset for 1993 to the WOCE UOT Science Centres.
Finally under this agenda item it was noted that meta data was being supplied with the delayed mode data that could not be accommodated in the GTSPP structure and format. An additional text-table may be required that will provide for 132 character records of free text information being stored in the database with the GTSPP data. This table would be at the "ocean station" level.
A small group was established under the chairmanship of Mr. M. Symmons of NODC and including also Mr. S. Diggs, Mr. B. Searle, Mr. Keeley, and Mr. J-P. Rebert. The Group is to recommend how best to incorporate all meta data fields into the GTSPP format with the option of adding the text-table discussed above. The Group is to complete its work by April 1994.
4.2 PROGRESS BY PARTICIPANTS IN IMPLEMENTING ASPECTS OF THE DELAYED MODE SYSTEM
Under this agenda item participants of the Meeting reported on progress in implementing the delayed mode aspect of the GTSPP. These reports are included in Annex V.
4.3 QUALITY CONTROL AND INTEGRATION OF HISTORICAL DATA FILES INTO THE GTSPP
The Chairman introduced this topic and reminded the Meeting of the definition accepted under Agenda Item 3.1.
Document IOC-WMO/IODE-IGOSS/GTSPP-III/10 was presented by Mr. Keeley. It reviewed the present handling of delayed mode data within GTSPP and some of the related projects and issues of ocean data management. Within GTSPP, progress has not been as swift in managing delayed mode as was originally planned. Outside of GTSPP, the GODAR project has been very successful in gathering substantial volumes of historical data. Other programmes such as WOCE and GOOS, are and will be making demands on the services and capabilities developed under GTSPP. The OceanPC project provides an opportunity for improving the acquisition of ocean data from a broader range of data collectors.
Discussions took place on the relationship between GODAR and GTSPP, particularly in the acquisition of historical data. It was decided that GTSPP would leave the acquisition and assembly of historical data to the GODAR project. GTSPP would concentrate efforts to pass all current archived data through the GTSPP QC procedures and duplicates management control to improve the state of existing digital archives. At the same time, participants were encouraged to take more active roles nationally in acquiring data and increasing the speed at which data enter their respective archives. Representatives of all NODCs present were committed to upgrading their archives in data quality and eliminating duplications, although some could carry out this task more quickly if software developed under GTSPP could be transferred to work at their centres. It was suggested that GTSPP data centres could undertake these tasks for particular ocean areas which coincided with their national interests. This would spread the projected work load of the US NODC, as well as speed up the incorporation into the digital archives of data gathered through the GODAR project.
5. PUBLICATIONS AND PRODUCTS5.1 QUARTERLY REPORTS
The Chairman of the Steering Committee reported on the status of the production of the GTSPP Quarterly Report Series. Due to a heavy work load and difficulties in soliciting written material for the Quarterly Reports, MEDS has fallen a year behind schedule. The Meeting discussed the reasons for producing these reports and decided that it was advisable to continue to produce such reports, although perhaps every 6 months instead of every quarter. The reports provided a means of keeping users aware that GTSPP was making progress and served to provide documentation of GTSPP technology for use by developers active in GTSPP, and as a means of transferring the GTSPP data management technology to IOC and WMO Member States in general.
Although there had been difficulty in finding material for the reports in the past year, it was noted that this Meeting of the Steering Committee had resulted in the development of a number of documents that should be included in future issues.
The Meeting also discussed whether the contents of the reports should be enhanced. It was noted that there are some analyses produced by the WOCE UOT Science Centres that are not included in the IGOSS Products Bulletin. It was agreed that it was appropriate to include such analyses in the GTSPP reports, but it was also agreed that these sorts of products should be offered to the IGOSS Products Bulletin first. The IGOSS and IODE Chairmen will approach contributors of products to the IGOSS Products Bulletin and urge them to identify, when appropriate, the sources of information on which products have been developed.
The Meeting decided that it would be appropriate to include in the reports, lists of available software and associated documentation, and information on regional and higher resolution climatologies. The NODC of Russia volunteered to prepare a description of some climatological and statistical products that might be included in the GTSPP reports. It was agreed that GTSPP should ask the scientific community what products they might like to see and that the request should include examples of what might be produced including the products to be described by the NODC of Russia. MEDS agreed to conduct the survey with the assistance of WOCE IPO and TOGA ITPO.
The final decision was that quarterly reports would be produced for the final two quarters of 1992 and semi-annual reports would be produced in 1993 and subsequent years. The reports would have additional content as discussed above and depending on the results of the response of the scientific community to the survey.
5.2 TIME-SERIES, CLIMATOLOGIES, DATASETS AND OTHER PRODUCTS
During early GTSPP meetings, the possibility of compiling long-term time-series datasets was discussed. It was agreed that a list of the best and most useful 100 locations and sections would be identified. In 1991, the USSR NODC proposed a list, compiled from information in their archives, with the hope of getting comments from other countries on the list. No comments have been received.
At the same time, the US NODC and WDC-A Oceanography, independently of GTSPP, gathered data from 10 Ocean Weather Stations and approximately 100 sections. These datasets, which were released on CD-ROM in July 1993, can be considered to be a start, but more work is needed. Meanwhile, data from other countries have been identified, and it was agreed that all data should be uniformly checked using GTSPP quality control. The Russian NODC and WDC-B Oceanography compiled data from OWS 'C', and was willing to check the data, provided they could use the GTSPP QC software on their computers.
It was agreed that during the next intersessional period, as a high priority, datasets from Ocean Weather Stations will be compiled as described below and GTSPP will work with the GODAR project to quality control and issue them as a CD-ROM.
It was also agreed that each GTSPP participant will identify repeat sections in its region of interest. The Russian NODC volunteered to co-ordinate the following tasks: obtain the WOCE list of time-series locations, contact the GODAR Project Leader for his comments on the most significant datasets, gather lists of sections from GTSPP participants, and the assembly and quality control of data by various countries.
While the GTSPP database is being developed, there will always be an opportunity to develop products for climate studies by managing combined (real-time, delayed, historical) datasets.
Two categories of climate products were proposed:
(i) 1 degree square climatic characteristics arranged by months and standard levels of observations - averages, standard deviations, distributions, maximum and minimum, etc;
(ii) Climatic characteristics in the grid of 1 degree square net arranged by months and standard levels - averages, standard deviations, etc.
The GTSPP database should have unique characteristics vis-à-vis quality and volume of data. Taking it into account it was suggested to consider GTSPP climate products as international standards for climatic characteristics and use them widely in national and international programmes of the World Ocean investigations.
It was recommended that during the intersessional period, a pilot sub-project be implemented on climate product development for an agreed-upon geographical area, e.g., for 2-3 10 degree squares in the North Atlantic. The NODC of Russia volunteered to take the responsibility for its implementation.
5.3 GTSPP CD-ROM
The Meeting reviewed the proposal for the publication of GTSPP CD-ROM series as proposed in Document IOC-WMO/IODE-IGOSS/GTSPP-III/12. This project is a joint effort between MEDS, US NODC and the IODE Group of Experts on Marine Information Management (GEMIM). The CD-ROM would contain 4 software tools to provide interaction with the information and data that would be included. It was proposed that the software tools to be included were the UNESCO CDS/ISIS package for bibliographic searches; a commercial document viewer to navigate and browse the documents and information; the PC ATLAST product for viewing and working with the temperature and salinity data; and some extraction scripts to enable users to download subsets of the data to the PC disk for processing by local applications software. Other software tools may be incorporated in co-operation with the OceanPC Project
The information collection on the disk would include all GTSPP meeting reports, the GTSPP project documents, and the GTSPP Monthly and Quarterly Report Series. There would be directories of the organizations and participants in the project including abstracts of responsibilities and expertise. There would also be data summaries, statistics and distribution maps. Finally, there would be bibliographies of data sources and of related publications.
It was agreed that the best copy of all data available in the CMD of US in NODC at the time of publication should be included on the CD-ROM. The GTSPP software that was available for distribution to users would also be included along with available documentation.
The Meeting noted that GTSPP had been designed and developed in close co-operation with the WOCE Upper Ocean Thermal DAC. It was therefore considered appropriate to include the reports from the WOCE UOT DAC meetings. The WOCE Representative agreed to this and offered to provide these reports in electronic form.
To use the CDS/ISIS product on the CD-ROM, it will be necessary to develop appropriate formats for the bibliographic database, and the directories of organizations and participants. The Meeting very much appreciated the support of the IODE Group of Experts on Marine Information Management in establishing a subgroup of technical experts to advise MEDS and US NODC on these matters. It was also noted that there is a requirement to develop a CDS/ISIS customized user interface to be used in conducting searches of these bibliographic and directory databases. Since there is no current expertise on building CDS/ISIS interfaces available in GTSPP, it was recommended that IOC find resources to engage a consultant to carry out this very important part of the project.
The Representative of the GEMIM strongly supported the GTSPP proposal stressing the fact that the integration of data and information was a high priority for GEMIM.
One other technical hurdle will be the development of the link between the data and the PC ATLAST product. The US NODC has experience in developing these links which have been included in the CD-ROM of ocean time-series data that they have recently issued. NODC volunteered to assist MEDS in developing these links for the GTSPP data.
The final points of discussion concerned the advertising and distribution of the CD-ROM, and the target date for its availability. It was agreed that MEDS and US NODC would provide copies to the GTSPP participants and to the IOC Secretariat for distribution to developing countries free of charge up to a total of 250 CD-ROMs. If the demand went beyond these numbers, there would be a charge not to exceed production of the disks and the royalties for the software on it. It was decided that the target date for making the CD-ROM available should be mid-summer 1994. Once it was clear that the CD-ROM would be available by a certain date, it should be announced in the IMS and International WOCE Newsletters.
6. CO-OPERATION WITH AND SERVICES TO OTHER PROGRAMMES6.1 WORLD OCEAN CIRCULATION EXPERIMENT (WOCE)/TROPICAL OCEANS AND GLOBAL ATMOSPHERE (TOGA)
GTSPP has provided a very useful service to WOCE with the major achievements being an increase in the data reported in real-time, identifying and fixing problems with the GTS and an improvement in the quality of the WOCE XBT dataset.
WOCE does not have a requirement for real-time data, but rather a need to monitor the network to ensure ships making consistent errors are contacted and that the required lines are being adequately sampled.
As part of the existing GTSPP monitoring mechanisms, MEDS provides monthly Ship Performance reports to the IGOSS Operations Co-ordinator, the WOCE IPO and the WOCE UOT DAC group. The IGOSS Operations Co-ordinator has had some success in using information in these reports to make ships aware of their problems which have subsequently been corrected.
The WOCE IPO requested GTSPP to extend the monitoring system in accordance with WOCE Report No. 106/93 (Report of the WOCE UOT/DAC Co-ordination Group, 5-6 April 1995) which asked that GTSPP provide monitoring of activity along TOGA-WOCE lines. The US-NODC agreed to provide quarterly summaries of the real-time data starting with the July-September 1993 set and yearly summaries of the delayed mode data sent to the Science Centres beginning with the 1990 set. These reports will include for each ship on each line, the number of sections, the number of observations and the start and end dates.
The WOCE IPO also requested GTSPP provide statistics for the annual datasets (from 1990 onwards) for incorporation in the WOCE Data Management Handbook every 6 months. The statistics should consist of the number of real-time and number of delayed mode profiles at US NODC which have undergone GTSPP quality control, and the number of real-time and number of delayed mode profiles which have undergone scientific quality control. The US NODC agreed to provide such statistics every 6 months starting in January 1994.
The WOCE IPO reviewed the proposed US NODC work schedule listed in Section 4.1 and compared it to the requirements for the delivery of the datasets. The dates for the 1990 to 1993 datasets are summarized in the table below. The WOCE IPO asked that the planned date for distributing the 1991 delayed mode data to the Science Centres be brought forward from November 1994 to April/May 1994.
Availability of GTSPP Datasets
Most Complete Set
- Time Period refers to the length of time after collection at which the data should be available to users.
- Op means presently in operation.
- Normal type indicates the required time period and dates for delivery of datasets according to WOCE Report No. 45/90 (Report of the WOCE UOT (Report of the WOCE UOT/DAC Co-ordination Group, 15-15 February 1990).
- ( ) indicates the dates according to the work scheduled proposed by US NODC in Section 4.1. The dates allow 6 months for the Science Centres to carry out the scientific QC and return the data and flags to US NODC and hence to Brest, and assumes both US NODC and Brest are ready to accept all meta data, flags, and feature and test codes with the data.
It was noted that considerable additional delayed mode data had arrived at US NODC since the 1990 dataset was passed to the Science Centres at the beginning of 1993. WOCE IPO will contact the Science Centres to inquire if they wish to receive the additional 1990 profiles with the 1991 dataset. The US NODC must write several routines to convert the originators data into the GTSPP format before the data are ready to be passed to the Science Centres. The scientific quality control of real-time data will cease at the Science Centres at the end of 1993 when MEDS will assume the whole responsibility for quality control of real-time data. Any real-time data which is not replaced by delayed mode data by the time the annual datasets are passed to the Science Centres will be subjected to scientific QC along with the delayed mode data. Meanwhile the flagged real-time data which has undergone scientific QC (January 1991 - December 1993 for AOML and SIO, and late 1993 - early 1994 for AODC/CSIRO) must be returned to US NODC and incorporated into the GTSPP CMD as soon as possible.
The Meeting considered the possibility that a user may request all the delayed mode data from specific programmes rather than the complete set. The Meeting agreed that it would be useful to identify the programme producing the delayed mode data by including its name in the surface codes group for each station. All originators of data are asked to supply the name of the programme under the auspices of which the data were collected and GTSPP will encode that information in the surface code group.
Within GTSPP it is important to avoid duplication of effort. There was a clear need expressed to improve the information between data centres regarding newly arrived datasets in order to avoid duplication of efforts. Mr. Rebert volunteered to propose to other centres simple electronic forms with information on datasets received and status of processing. These forms may be circulated amongst the GTSPP participants either by mail or put on a common directory accessible by ftp.
6.2 GLOBAL OCEAN OBSERVING SYSTEMS (GOOS)
The Technical Secretary referred the participants to the findings of the First Session of the IOC Committee for the Global Ocean Observing System (Paris, 16-19 February 1993) and to the Document IOC-XVII/8 Annex 2 rev. entitled "Approach to GOOS". He explained that GOOS is developing now as a global framework for systematic ocean observations to meet the needs of different users' groups from forecasting climate variability and change, and assessing the health of the marine environment and its resources, to monitoring of the coastal zone and providing a wide scope of meteorological and oceanographic operational services. He then presented the major elements of GOOS which are operational oceanographic observation and analysis, collection and distribution of data and products, data assimilation into numerical models leading to predictions and capacity building within participating Member States to develop analysis and application capability. The Technical Secretary emphasized that operational programmes including IGOSS, IODE and GLOSS will form a foundation of future GOOS and that these programmes should be enhanced and strongly supported and their accelerated development should be considered as a priority for GOOS development.
The Meeting agreed that IGOSS and IODE services and, in particular its GTSPP, can provide a sound contribution to GOOS in the area of operational ocean data management. The Meeting discussed and identified the on-going activities in GTSPP and IODE in general, relevant to GOOS. Using as basis a paper submitted by Mr. Searle on GOOS Data Management Goals, the Meeting recommended the following goals which need to be achieved to meet GOOS objectives in data management and noted that IGOSS and IODE have already undertaken activities to meet these goals. The List of Goals is presented in Annex VI to the Summary Report which also contain the status and further actions required for achieving these goals. The Meeting fully supported the GOOS objectives and its development and committed itself to undertake activities in order to achieve goals presented in Annex VI.
The Meeting also fully supported the concept that GOOS should be developed on a sound scientific basis using the findings of on-going research programmes, including WOCE, TOGA and JGOFS, and agreed that GOOS will have a lot to gain from TOGA and WOCE successes. In this regard, the Meeting noted with interest information provided by Dr. Wilson on the efforts made by MEDS in summarizing TOGA/WOCE achievements in ocean data management and in interpretation of these successes on the GOOS development in oceanographic data services. The Meeting encouraged MEDS to continue the efforts and keep the participants informed on the progress.
Mr. Keeley presented Document IOC-WMO/IGOSS-IODE/GTSPP-III/15 discussing the possible relationship between GTSPP and OceanPC. There was a number of possibilities presented and these provoked discussion in more general terms about the capabilities and goals of OceanPC.
The Meeting viewed the OceanPC project as a means to deliver to users both data and products not only of GTSPP but also of IGOSS and IODE. The Meeting expressed concern that a shortage of staff available at IOC to devote to this project would hinder a coherent development path for the future.
The Meeting agreed that the production of the GTSPP CD-ROM would be an effective way to help in the acceptance and promotion of OceanPC.
It was noted that OceanPC had fulfilled its goal of making software available that could be used in the management and analysis of ocean data on PCs. The Meeting felt that it was time to develop further goals for the project and that the issuance of the GTSPP CD-ROM with ATLAST software and perhaps data entry software included would provide the next step forward to OceanPC users. Since the CD-ROM would contain not only the data from GTSPP but also a suite of documents describing the project as well as software, it was felt that this would provide increased capabilities to users. At the same time, the presence of the related information describing GTSPP would sensitize users to the importance of including meta data with data.
The Meeting was informed that OceanPC was a prime candidate for the entry of data for the GODAR Project. Consequently, the output format for these data would become the defacto standard for the exchange of data amongst participants of GODAR. Concern was expressed that because of the current input and output capabilities of OceanPC, valuable information could be lost, and in particular meta data useful in the interpretation of the data. For this reason, it was considered important that more capable input and output software be included so that the meda data can be preserved and exchanged. It was suggested that improved data input and output software using the GTSPP format would enhance the integration of new data with the GTSPP data and also accept the meda data that has formed an important part of the GTSPP data and information management philosophy. It was suggested that such software be included on the GTSPP CD-ROM. Inputting data in the GTSPP format provides access to the suite of data already available in the ROM. Developers of OceanPC are encouraged to convert the internal format of OceanPC to that of larger collections of data such as, for example that of GTSPP, in order to increase the ease of inclusion of a larger pool of data into the project. As well it is important for OceanPC to accept software that is written in other computer languages so that advantage can be taken of the more sophisticated software available from within GTSPP and from others.
6.4 GLOBAL OCEANOGRAPHIC DATA ARCHAEOLOGY AND RESCUE (GODAR)
Since September 1990, when the idea of the GODAR Project was discussed and accepted and the implementation have been started by 3 World Data Centres for Oceanography and ICES, the data archaeology project has made substantial progress in acquiring data that had never reached archive centres. About 500,000 new temperature and salinity stations have been processed as well as almost 1 million BT stations.
The Technical Secretary highlighted the main achievements in the implementation of GODAR in 1993. The IOC Assembly at its Seventeenth Session decided to widen the scope of the project and made it international with the purpose to facilitate the creation of global oceanographic databases for use by the international research community for the study of the role of the world ocean as part of the earth's climate system. Dr. S. Levitus, Director of WDC-A, Oceanography, became the project leader. The First Regional Workshop for Member States of Eastern and Northern Europe (GODAR-I) was held in Obninsk, Russian Federation, under the auspices of the World Data Centre B, Oceanography (Obninsk, Russian Federation, 17-20 May, 1993). The Workshop succeeded in assessing the state of data holdings in the region, identified common goals and problems with data preservation and recommended implementation steps and approaches to solve these problems. It is envisaged to arrange the same type of regional workshops in the future for the regions where there is a good potential to find big data holdings which are not yet accessible for a wide international community, and there is a danger of losing these data.
Mr. Keeley presented Document IOC-WMO/IODE-IGOSS/GTSPP-III/10 with the proposals for the co-operation of GTSPP with the GODAR Project. He emphasized that this co-operation may be achieved in different ways. First, since participants of the GTSPP have pledged to improve the timeliness of data exchange it should provide more delayed mode data sooner to international exchange and therefore be available sooner to the GODAR Project. At the same time, if GTSPP managed this aspect well, the GODAR Project could concentrate on reaching co-operative agreements with other countries in acquiring data. It was important that GTSPP demonstrated improved data flows for delayed mode data.
On the other hand, the GODAR project was capturing large volumes of data. The Meeting noted that large influx of new data will challenge processing systems of WDCs Oceanography, to include the data in their archives and make them available for exchange in a relatively short period of time. The US NODC has estimated that it will be about 10 years of work for the US NODC to incorporate the data being collected under GODAR into the database. Of course, this estimation was made on the basis of past technology. The progress in computer technology development may considerably decrease this figure. The Meeting agreed that GODAR and GTSPP could co-operate by sharing processing responsibilities. There was a substantial amount of work to reformat and process new data. GTSPP could offer software already developed such as for the identification of duplicates and quality control procedures. Participants of GTSPP could also offer to help in reformatting new data so that it becomes available more quickly. This will allow GODAR to concentrate on identifying and acquiring more data and on improving ocean climatologies.
The Meeting accepted the principal that the task of finding historical data will be left for GODAR and that GTSPP will help GODAR implementing more effectively and efficiently data processing. QC software written in C will be useful and should be shared with data centres participating in GODAR data processing. The Meeting noted with thanks the offers made by some participants to help GODAR to accept QC and put it in GTSPP database (B. Searle - Indian Ocean and Western Pacific; MEDS - Areas of Interest to Canada; AOMC - Southern North Atlantic and South Atlantic, MEDS and Scripps - Eastern Pacific). It was decided that QC software should be carefully and thoroughly developed, tested and well documented in US NODC.
Taking into account that the GODAR project is very active in digitizing data not already entered on computer media, the Meeting recommended that it would be very useful if MEDS and the US NODC could provide a PC programme (as referred to under Agenda Item 6.3) for key entry of data that would result in the production of the data in the GTSPP format. It would also be very helpful if GTSPP were to provide GODAR with the description of the meta data that should be captured along with the data as determined by the ad hoc Group established under Agenda Item 4.1. The provision of the PC application described above would also simplify the capture of this meta data.
The Meeting reiterated also the importance of NODCs in the GODAR data collection and exchange component and advised that NODCs should be the main national contacts for search, rescue and retrieval of necessary data holdings.
The Session noted that OceanPC may become a useful tool for GODAR rescue operations as an exchange format and recommended that the OceanPC group would consider the steps to be taken in order to make OceanPC better able to respond to this task. GTSPP could help OceanPC in designing a forms entry programme.
7. SOFTWARE AND SYSTEM IMPROVEMENTS7.1 FORMATS AND META DATA
Mr. Keeley reviewed issues related to formats in Document IOC-WMO/IODE-IGOSS/GTSPP-III/20. It was agreed that the GTSPP format has performed very well and will continue to be used as GTSPP expands. The ASCII version of the format is now used for exchange among GTSPP participants, and makes possible, distribution of data to other users with a wide array of computers. Continued work on formats is needed, as GTSPP data will be combined with meteorological data and to ease the loading of data into spreadsheets and database software. The Meeting was informed of work by GETADE to test the loading of the GTSPP data into BUFR and the GETADE formats. GETADE was urged to continue these tests.
In response to the need to record textual information, it was agreed that MEDS and the US NODC will work with DACs to test adding 80-character free-text groups to the GTSPP format. These groups will be available for recording comments, documentation and meta data (supplied with delayed mode data) which cannot be recorded in the present GTSPP format.
A meta data strategy for GTSPP discussion was led by Mr. Hamilton, who introduced Document IOC-WMO/IODE-IGOSS/GTSPP-III/21. It was agreed that the regular publication of CD-ROMs meets the need to provide information and documentation on the project, participants and data management systems. Also, the agreement to add free-text groups to the GTSPP format meets most requirements for meta data which accompanies delayed mode data.
7.2 CONVERSION OF THE GTSPP QC SOFTWARE TO C AND UNIX
Mr. Hamilton reported on progress in converting the QC software developed by MEDS in the Fortran/VMS/GKS environment to an appropriate language and operating system for more general distribution. The MEDS QCA software (checks ship tracks, position and date-time, and speed made good) is in the process of being converted to C++ for the Unix workstation using MOTIF for the graphical user interface. It is expected that this software will be operational by the end of January 1994. The profile checking software (MEDS QCB) is also being converted to C++ for a Unix workstation, not including the visual profile check available in the MEDS software. This application is expected to be operational by the end of February 1994.
A duplicates identification or station matching module is also being developed in C++. This module uses a fuzzy date-time, fuzzy position test (15 minutes and 5 km) to identify duplicates. This module is scheduled to be completed in March 1994.
Software has also been written to convert to and from the NODC workstation format to the ASCII format in which the GTSPP data are supplied to NODC by MEDS. Several other inventory routines that interact with the Sybase database are also available.
Mr. Mikhailov, Mr. Searle, and Mr. Diggs volunteered to test the portability of the software.
7.3 QC REVIEW AND REVISIONS
Document IOC-WMO/IODE-IGOSS/GTSPP-III/16 was presented under this Agenda Item. The attention was drawn to the suggestions made by WOCE for ceasing to automatically flag failures of the constant profile test and the Levitus climatology. A number of other modifications of existing tests or new tests were proposed. A question was raised about the boundaries of the Mediterranean Sea in one of the regional tests.
Russia described work that was started in preparing a climatology for the North Atlantic. This work was encouraged by participants since such climatologies could be of use to GTSPP.
Since Document IOC-WMO/IODE-IGOSS/GTSPP-III/16 had not been circulated prior to the Meeting, it was not possible for the participants to respond with decisions about the proposal contained therein. The Meeting accepted MEDS offer to review the text to ensure the proposed actions for each test were very clear, to include the question of appropriate boundaries for the Mediterranean, and to distribute copies to science centres, WOCE IPO and the TOGA/WOCE SDC for their review keeping in mind that it is intended to implement the agreed modifications by the beginning of 1994.
7.4 PILOT OCEAN DATA NETWORK
The Fourteenth Session of the Committee on IODE asked GTSPP to consider the establishment of a pilot ocean data network to improve the flow of data within IODE and to improve the delivery of IODE services to users. The Meeting noted the rapidly expanding services and user base of Internet and decided that GTSPP should implement a number of test servers on Internet. This will allow GTSPP to gather the experience, to evaluate the needs of IODE and make appropriate recommendations to the Fifteenth Session of the Committee.
There are 3 types of servers that can now be implemented on Internet. The 3 are FTP anonymous accounts, Gopher Servers and WAIS Servers.
The FTP anonymous account allows users who know you are on the network with a certain Internet address to connect to your system and either list the contents of textual files on their screens or copy files from the account to their workstation. This allows the user to view a "read me" file that describes the content and purpose of other files available in the account, and then if they wish, to copy files of interest to their workstation for use by local applications. It can be used, for example, to put up an ASCII file of data, with a "read me" file that describes the content and format. Users would view the "read me" file and then if desired, copy the data file for use on their workstation.
The next step of sophistication would be to implement a Gopher server. Implementation of a Gopher server allows the user to register with a central Internet server so that potential users of GTSPP services could discover the existence and addresses of our servers while browsing classes of available servers. In addition, Gopher supports the searching of file hierarchies on the GTSPP servers and allows the execution of scripts that would support copying back or "mailing" files to the client or to other Internet addresses.
The third level of sophistication is the WAIS server which allows the user to run applications on the server that allow browsing and retrieval of subsets of the data that meet various retrieval criteria.
The Meeting decided that GTSPP should proceed with the establishment of 3 servers to gain knowledge of this technology and understand how it could best be applied to meet the needs of GTSPP users. The 3 servers would include a server in NODC that would hold authority lists for ship codes and other GTSPP codes, a server in the Australian Oceanographic Data Centre that would hold a GTSPP software inventory and documentation available to other centres, and a server in MEDS that would hold the MEDI catalogue. In addition, SIO is intending to implement a server for maps of temperature and salinity prepared with the GTSPP data. The Representative of SIO offered to include this server in the GTSPP study of the uses of these types of servers.
Implementation of the servers is to begin by using FTP anonymous accounts and then is to proceed to Gopher and WAIS as experience is gained, and as it is determined that there is a need to implement a more sophisticated server.
The representatives of WOCE and the TOGA/WOCE Subsurface Data Centre informed the Meeting of servers that are already available for some of the WOCE DACs and for SISMER, in Brest. They offered to provide lists of the servers with Internet addresses so that GTSPP participants implementing the servers described above would be able to explore how they have been implemented. It was recommended to publish them in the GTSPP reports.
With the experience gained from the implementation of the test servers, recommendations can be prepared for further GTSPP Ocean Network Services to be considered at the next Steering Committee meeting.
7.5 SOFTWARE STANDARDS AND EXCHANGE OF SOFTWARE
Document IOC-WMO/IODE-IGOSS/GTSPP-III/18 describing various concepts regarding the development, exchange and portability of software within the GTSPP community was presented by Mr. Searle of AODC. He noted the considerable investment in time and money expended by GTSPP participants on developing software. Much of this development activity was overlapping which results in an inefficient use of resources. The concept developed by OceanPC to provide freely available software could be built on, as an example, within GTSPP. An objective of this, is to eventually standardize on software which would have considerable benefits, including, reducing costs, improving data flow, reducing time to implement new systems.
A number of issues were discussed, including legal aspects and responsibilities related to exchanged software. The Meeting agreed that no responsibility would be accepted for any software and that maintenance would not be provided. The Meeting then discussed the different levels of software portability and agreed that there were 4 different levels. These consisted of:
(a) fully portable software (complete applications);
(b) portable software modules (stand alone routines);
(c) partially portable software (dependent on proprietary software);
(d) commercial software.
Mr. Searle described a software inventory that they have developed using ORACLE and it was suggested that this be expanded to support an inventory of the 4 categories of GTSPP software given above. A further category describing existing and planned development projects would also be included. The Meeting agreed with the suggestion that this inventory be available on Internet and perhaps form part of the Pilot Ocean Data Network (Item 7.4). GETADE have done some work already to compile a software inventory and Mr. Searle agreed to work with GETADE to develop the inventory which will be accessible via Internet as referred to in Agenda Item 7.4.
A discussion took place on the future directions of the GTSPP software development and it was recommended that FORTRAN, C and X-Windows be adopted as a standard software development environment. It was stressed that this development environment was recommended as a direction that would assist with increasing the portability of software but should not be considered a rule for the GTSPP participants.
Further discussion took place on the future directions for the GTSPP software and it was considered that if true portability could be achieved, development projects could be shared between centres. This would be of considerable benefit to all participants. In an effort to explore portability and the sharing of software, a number of experiments and investigations were proposed. These are:
(i) US NODC and AODC will investigate IFREMER, SCRIPPS and CSIRO visualizations software used in their quality control systems as a "visual" front-end to the GTSPP quality control software;
(ii) MEDS would look at establishing the capability to enable US NODC to use the MEDS software to convert high resolution CTD and delayed mode data to the GTSPP format;
(iii) MEDS would provide their duplicate checking software to SIO for inclusion in the SIO quality control software application and to provide access to the system as a service for use by the US NODC.
8. REVIEW OF THE ACTION PLANThe Meeting reviewed the implementation plan as it was presented in the GTSPP-II Summary Report and made comments on the state of its implementation. The Meeting also requested the Chairman of the Steering Committee and the Technical Secretary to prepare the implementation plan for future activities taking into account the findings of GTSPP-III and include it as Annex II to the Summary Report.
It was recommended that an IOC Circular Letter be issued to inform Member States and international organizations on the state and plans of the GTSPP implementation.
The Meeting considered also the need to update the GTSPP Project Plan, published in 1991, and agreed before any concrete decisions are taken to request Mr. Searle to review the document and report his view to GTSPP-IV on future actions, if any.
9. OTHER BUSINESS
(i) Marine Pollution Emergency Response Support System (MPERSS)The Technical Secretary introduced a new WMO project on the development of Marine Pollution Emergency Response Support System (MPERSS) which is intended to provide in a globally co-ordinated way, the meteorological and related oceanographic support required for marine pollution emergency response operations on the high seas. The Technical Secretary invited the participants to consider in what way GTSPP can contribute to this Project.
The Meeting noted that GTSPP is already collecting and providing necessary information which may be used to meet the requirements of MPERSS for the preparation of the required meteorological and oceanographic information which may be available to the competent marine pollution emergency response authorities.
The Meeting expressed readiness to provide any technical advice for collection, management and analysis of the values of temperature and salinity required as an input to models describing the movement, dispersion, dissipation and dissolution of marine pollution.
(ii) Software DemonstrationsMEDS presented 4 demonstrations of software which have been developed in support of its GTSPP activities.
The first was the ship tracking software used to confirm the quality of station position and time information as well as ship identification. The software runs interactively and presents a map showing station locations, and computes speeds between stations based on position and time. Various editing and flagging operations are permitted using a point and click interface. There is also the capability to correct ship identification errors and to combine the correctly identified stations with other stations with the same identifier.
The second demonstration showed software used by MEDS in assessing the quality of the subsurface profiles. The software plots temperature and salinity profiles as a function of depth and computes and displays densities based on these variables. As each profile is being displayed, the suite of tests described in the GTSPP Quality Control Manual are executed and if any failures result, the information is also portrayed. An operator has a number of display and editing options all driven through a point and click interface. All changes of quality flags or of values are recorded in the GTSPP record structures.
The third demonstration was of the station duplicates detection software. This software identifies stations collected within 15 minutes and 5 kilometers of each other. For these stations, the subsurface profiles are then inspected and if sufficiently similar, they are presented to an operator for review. The programme runs interactively displaying profiles only as needed to resolve possible duplications detected by the software.
The fourth programme demonstrated is used in the detection of possible duplications of cruises. Newly received cruises are examined to see which stations lie within 15 minutes and 5 kilometers of stations in another cruise. The software presents information of which other cruises have duplications to a target cruise and how many stations are duplicates. An operator can then display the cruises showing the stations duplicated and decide if the duplicate stations from one or the other cruise should be flagged or if in fact the stations are not duplicates. The software also uses a form of subsurface test which can compare same variable profiles measured at different depth resolution.
(iii) AOML QC ProceduresThe AOML quality control procedures are currently focussed on XBT data collected in the Atlantic Ocean. Real-time, delayed mode, historical and navy declassified data are all currently being quality controlled. The database is updated each month. Quality control procedures at AOML can be characterized as more subjective than objective because there are no automatic flags established by any of the quality control tests. Profiles that are identified as suspicious are examined individually and flagged as appropriate. The steps involved in the quality control process are listed below:
(a) A programme to find the duplicate profiles in the dataset is run. Duplicate profiles found are printed to an output file. The operator interactively views the profiles to determine if they are in fact duplicates. While in the interactive mode, the operator can mark the profile as a definite duplicate, have a listing of the values for each profile printed to screen, or choose to ignore the profile if it is determined that they are not duplicates;
(b) A sigma test is then run on the dataset. The programme typically uses a 3 sigma tolerance about the historical mean and a 60 C tolerance about the Levitus 10 degree climatology. A profile is displayed if the temperature at any depth in a profile differs from the mean profile of a Levitus climatology by more than the specified tolerance. The operator has the option to mark questionable profiles. This will save the profile in a file which is then used as input to other quality control programmes;
(c) The data is next run through a programme that examines all of the XBTs associated with a specific cruise. Profiles, as well as time and position locations, are checked for consistency with surrounding XBTs. The profiles can be viewed interactively or on hardcopy. The programme is generally run interactively to further examine and identify suspicious profiles for a specific cruise. Suspicious profiles observed in the previously run test (see step 2) are viewed during this programme also. The following graphical outputs, which are quite useful for quality control, can be generated by this programme: temperature vs depth profiles (waterfall plots) and/or geographic position of the XBT's for any cruise; geographic position of the XBT's and speed plots for any cruise; and vertical temperature section plots for various types of cruise tracks;
(d) The data is then run through a programme that reads a position file which contains the information about position or time corrections for an XBT profile. The profile is updated with the corrected position and/or time and the XBT profile is rewritten back to the database;
(e) At this time, a programme is run on the data which allows the operator to interactively examine and edit all XBT profiles. Suspicious profiles from the previous tests are highlighted so that the operator knows what specific test a profile has failed. The following options are available to the operator while running this programme: the operator can get a listing of the data values; truncate the data; request to view a specific XBT; zoom in on specific features in a profile; and change the graph settings. Flags are placed on the data by the operator while running this programme;
(f) In the last of the actual quality control tests, the data are run through a programme that interpolates the sea surface temperature at 150 m and the average temperature of the upper 400 m layer onto a 20 x 20 grid within the region (80 W - 20 E; 40 S - 60 N) and counts the resulting fields in order to identify suspicious profiles, i.e., those producing "bulls eyes". Only data values that have received a flag of acceptable (1) or inconsistent (2) are used as input for the interpolation. If a "bulls eye" is observed on the contour map generated by the programme, then another programme is run on the data. This programme finds all XBTs that fall inside of a circle of a specific radius, around a specific location. Those XBTs that are identified as suspicious are re-examined by running the XBT edit programme (see step 5) and are flagged as appropriate.
After all of the above tests are run, the quality control process is complete for that dataset. The data is then run through a final programme which prepares it for submission to the US NODC.
It should be noted that AOML is presently attempting to establish a flagging system consistent with that of CSIRO. In this system, not only is the profile flagged, but the questionable profile is given a code which indicates why the profile was assigned that particular code.
(iv) Presentation of the Global Centre QC ModelThe QC Manual was developed in a X-Windows environment and based on the GKS graphic library. Visual inspection of profiles uses two screens containing cruise map, waterfall, current profile vs climatology and statistical envelope (3 s.d.), headers window. They are all dynamically linked: the operator can point and click on any of these items. The current profile is coloured according to the quality flag attached to individual points (4 colours).
The operator is allowed to modify the flags attached by the automatic QC. Elementary corrections are allowed both in the header or the data for:
(a) date and position;
(b) spikes (interpolation or deletion);
(c) deletion of erroneous bottom values;
(d) deletion of start-up transient.
Each of these corrections updates the history table according to the rules proposed
by MEDS. Two additional codes were added to this table to provide information on the two
last above-mentioned corrections.
SIO QC SoftwareMr. Diggs of SIO gave a brief presentation on the development of their QC software. After a short overview of the programme requirements and logical design, examples of the actual point and click environment were shown followed by a discussion of some of the concerns of SIO, most of which were addressed under other agenda items. One of those concerns was the 5% error in the depth/fall rate of the XBTs. Prof. D. Kohnke noted that this fall rate error has been addressed by IGOSS and that a publication would be forthcoming.
SIO is developing this software using IDL and Fortran on Silicon Graphics and Digital Equipment hardware. Upon completion, this software will become part of the public domain.
NETCDFUnidata's netCDF interface is one of many formats under consideration by the GTSPP for data exchange. Mr. Diggs of SIO gave an overview of this interface. NetCDF was originally derived from NASA's CDF as a network extension of the Common Data Format. This interface allows scientists to exchange information that is self-describing in a heterogeneous hardware environment. Some advantages of netCDF appear to be its ability to handle meta data, the self-describing nature of the files, the wealth of software that will accept the format and the number of institutions that already use it including NOAA/PMEL, CSIRO, WHOI, NCAR and SIO.
The Meeting noted that the interface may have shortcomings representing randomly spaced observations with variable depth resolution. SIO will eventually attempt to put all of their delayed mode data into netCDF for internal use and will report on their progress. Another format, HDF will also be investigated by MEDS, US NODC and SIO.
10. APPROVAL OF THE SUMMARY REPORTThe Meeting approved the draft of the Summary Report and requested its Chairman and the Technical Secretary to make all editorial changes they deem to be necessary.
11. DATES AND PLACE OF NEXT MEETINGThe Meeting agreed that there is a need to have one more Session before the IODE Committee Session planned for the end of 1995 beginning of 1996. It would be most desirable if this Session be held jointly with the WOCE UOT-DAC Session in 1995 and requested its Chairman to have the necessary consultations with the WOCE officers in order to identify dates and place of the next Session and inform the IOC Secretariat on the results of this consultation by the middle of 1994. It was recommended that the session be held in mid-summer of 1995, if possible.
12. CLOSURE OF THE SESSION
The Session was closed at 15.00 on 19 November 1993.
ANNEX I - AGENDA
1. OPENING OF THE MEETING
2. ADMINISTRATIVE ARRANGEMENTS
2.1 ADOPTION OF THE AGENDA
2.2 DESIGNATION OF A RAPPORTEUR
2.3 CONDUCT OF THE SESSION
3. REAL-TIME SYSTEMS DEVELOPMENT
3.1 REVIEW OF THE EXISTING OPERATIONAL DATA FLOW
3.3 NAVY DECLASSIFIED DATA
3.4 DATA FLOW MONITORING
4. DELAYED MODE AND HISTORICAL DATA SYSTEMS DEVELOPMENT
4.1 PROGRESS ON DEVELOPMENT OF THE CONTINUOUSLY MANAGED DATABASE (CMD)
4.2 PROGRESS BY PARTICIPANTS IN IMPLEMENTING ASPECTS OF THE DELAYED MODE SYSTEM
4.3 QUALITY CONTROL AND INTEGRATION OF HISTORICAL DATA FILES INTO THE GTSPP
5. PUBLICATIONS AND PRODUCTS
5.1 QUARTERLY REPORTS
5.2 TIME-SERIES, CLIMATOLOGIES, DATASETS AND OTHER PRODUCTS
5.3 THE GTSPP CD-ROM
6. CO-OPERATION WITH AND SERVICES TO OTHER PROGRAMMES
6.1 WORLD OCEAN CIRCULATION EXPERIMENT (WOCE) / TROPICAL OCEANS AND GLOBAL ATMOSPHERE (TOGA)
6.2 GLOBAL OCEAN OBSERVING SYSTEMS (GOOS)
6.4 GLOBAL OCEANOGRAPHIC DATA ARCHEOLOGY AND RESCUE (GODAR)
7. SOFTWARE AND SYSTEM IMPROVEMENTS
7.1 FORMATS AND META DATA
7.2 CONVERSION OF THE GTSPP QC SOFTWARE TO C AND UNIX
7.3 QC REVIEW AND REVISIONS
7.4 PILOT OCEAN DATA NETWORK
7.5 SOFTWARE STANDARDS AND EXCHANGE OF SOFTWARE
8. REVIEW OF THE ACTION PLAN
9. OTHER BUSINESS
10. APPROVAL OF THE SUMMARY REPORT
11. DATES AND PLACE OF NEXT MEETING
12. CLOSURE OF THE SESSION
GTSPP IMPLEMENTATION PLAN FOR 1994-1995 - RECOMMENDATIONS
INTRODUCTIONThe implementation of GTSPP can be described in terms of the following 4 general tasks:
- Develop an improved real-time data capture to prevent loss of data; provide an awareness of the existence of data to assist with later acquisition of the data; provide more data to operational programmes and managers of research programmes.
- Implement a documented and uniform QC for data being archived in the CMD including a rational approach to the management of duplicates, and the inclusion of the necessary meta data in the database and associated files.
- Identify sources of delayed mode and historical temperature and salinity data, acquire and digitize it where necessary, QC and input it to the CMD.
- Develop, prepare and distribute data, data products and data flow monitoring products at all time scales to meet the needs of users at all time scales.
OVERVIEW PRIORITIESFor the coming intersessional period the general priorities are as follows:
- Develop first GTSPP CD-ROM products.
- Include TRACKOB data in the GTSPP data flow.
- Implement the remaining aspects of the GTSPP-WOCE UOT DAC data flows.
- Implement the historical data and the time-series aspects of the project with a series of tasks designed to acquire QC and archive both digitized and non-digitized data.
- Implement real-time data provision to at least one TOGA modelling centre.
- Define a role of GTSPP in GODAR project and for the IOC OceanPC in the GTSPP.
IMPLEMENTATION OF THE ELEMENTS OF GTSPP
Existing Operational Data Flow- To encourage GTSPP participants to use the term "low and high resolution data" in describing their datasets. (Chairman GTSPP, IOC).
TRACKOB- To distribute TRACKOB data in accordance with methods established by GTSPP. (All GTSPP members).
- To include TRACKOB data in the GTSPP monthly monitoring reports. (MEDS).
- To make available to the GTSPP track temperature and salinity data along with any other parameters observed simultaneously that are not reported via TRACKOB. (All GTSPP members, ORSTOM Brest).
- To formulate structures and methods for handling data collected using undulating instruments. (Chairman GETADE and GTSPP).
To survey IGOSS National Representatives in Member States to determine what surface observational data are being collected and reported in TRACKOB. (IGOSS NOC in Germany).
NAVY DECLASSIFIED DATA
- To contact national navies requesting them to declassify data including high resolution data. (All GTSPP members).
- To keep the Secretary IOC and the Chairman IODE informed through letters, on developments in declassifying ocean data. (All GTSPP members).
- To contact US Navy to obtain any available high resolution copies of declassified data (US NODC).
- To assign a unique identification code when the call sign is absent but the SHIP is known. (Chairman GETADE, Toga Sub-surface Data Centre, US NODC).
DATA FLOW MONITORING
- To prepare an annual report for the real-time data flow along the lines of the TOGA/WOCE data rescue document and circulate it to the NODCs of GTSPP members. (MEDS).
- To enhance the usage of electronic mail systems by WDCs for distribution of CSRs. (Chairman IODE, Directors of WDCs).
- To complete the development of the software system to assign the real-time data assembled by the GTSPP to WOCE/TOGA lines. (US NODC).
- To continue production of the monthly GTS data flow analysis. (MEDS).
- To prepare semi-annual reports with the description of the flow of data through the WOCE UOT centres. (US NODC and WOCE IPO).
- To provide annually an overview of the flow of real-time and delayed mode data, identify problems and priorities for solution, publish this information in the GTSPP Semi-annual Report Series. (MEDS).
- To tag real-time data which are replaced by delayed mode data. (US NODC).
- To accept and merge scientific QC information on RT data from DACs. (US NODC).
- To re-format 1990-1992 delayed mode data from original datasets to GTSPP format. (US NODC).
- To complete QC of 1990-1992 delayed mode data. (US NODC).
- To add QC'd delayed mode data from 1990-1992 to CMD. (US NODC).
- To distribute annual datasets for 1991-1992. (US NODC).
- To distribute annual datasets for 1993. (US NODC).
- To QC the real-time datasets for 1994 and onwards and to alert SIO and AOML of problems. (MEDS).
- To carry out a comparative quality control on the Indian Ocean data for a short period in 1994 to validate MEDS procedures for those data. (AODC).
- To review the data and meta data being supplied with the delayed mode upper ocean XBT data and to recommend how best to incorporate all fields into the GTSPP format. (Ad hoc Group under the Chairmanship of Mr. M. Simmons).
QC AND INTEGRATION OF HISTORICAL DATA FILES INTO GTSPP
- To pass all current archived data through the GTSPP QC procedures and duplicates management control. (All GTSPP members).
- To undertake upgrading of the archives in data quality and eliminating duplications for particular ocean areas which coincide with their national interests. (All GTSPP members).
- To provide software for data quality and eliminating duplicates developed under GTSPP to all GTSPP members. (MEDS and US NODC).
GTSPP QUARTERLY REPORTS
- To continue production of GTSPP reports on half a year basis. (MEDS).
- To include analyses produced by WOCE UOT Science Centres, not included in the IGOSS Products Bulletin, into the GTSPP reports. (Chairmen IODE and IGOSS).
- To include in the GTSPP reports list of available software and associated documentation, and information on regional and higher resolution climatologies. (Russian NODC).
- To investigate with the scientific community what products they might like to see. (MEDS, WOCE IPO, TOGA ITPO).
TIME-SERIES, CLIMATOLOGIES, DATASETS AND OTHER PRODUCTS
- To complete datasets from Ocean Weather Stations as agreed at the GTSPP-III meeting, quality control them and issue as a CD-ROM. (MEDS, GODAR Project Leader).
- To identify repeat sections for the geographical areas of interest. (All GTSPP members).
- To co-ordinate efforts in the preparation of time-series. (Russian NODC).
- To suggest GTSPP climate products as international standards for climatic characteristics (1 degree square arranged by months and standard levels of observations; climatic characteristics in the grid of the 1 degree square). (GTSPP Chairman).
- To implement a pilot sub-project on climate product development for an agreed-upon geographical area. (Russian NODC).
- Start publication of GTSPP CD-ROM series. (MEDS, US NODC, GEMIM).
- To provide reports of the WOCE UOT DAC meetings in electronic form for inclusion in the GTSPP CD-ROM. (WOCE Representative).
- To engage a consultant to develop a CDS/ISIS customized user interface to be used in conducting searches of bibliographic and directory databases. (IOC and GEMIM).
- To announce the availability of the GTSPP CD-ROM in the IMS and International WOCE Newsletter. (Chairman GTSPP, IOC).
- To provide CD-ROM copies to the IOC Secretariat for distribution. (MEDS, US NODC, IOC).
- To extend the GTSPP monitoring system along TOGA/WOCE lines in accordance with WOCE Report No. 106/93. (US NODC).
- To provide the WOCE IPO with statistics for the annual datasets (from 1990 onwards) for incorporation in the WOCE Data Management Handbook every 6 months. (US NODC).
- To write routines for conversion of the originator's data into the GTSPP format before passing the data to the Science Centres. (US NODC).
- To return the flagged real-time data which has undergone scientific QC to US NODC for incorporation into the GTSPP CMD. (AOML, SIO, AODC, CSIRO, US NODC).
- To supply MEDS with the names of the programmes under the auspices of which the data were collected in order to encode this information in the surface code group. (All GTSPP members, MEDS).
- To propose to GTSPP members simple electronic forms with information on datasets received and status of processing which may be circulated among GTSPP participants by mail or put on a common directory accessible by ftp. (TOGA Subsurface Data Centre).
- To summarize TOGA/WOCE and IGOSS/IODE achievements in ocean data management and in interpretation of these successes on the GOOS development in oceanographic data services. (MEDS, AODC).
- To include on the GTSPP CD-ROM improved data input and output software using the GTSPP format in order to enhance the integration of new data with the GTSPP data and accept the GTSPP meta data. (MEDS).
- To convert the internal format of OceanPC to that for larger collections of data. (OceanPC Project Leader).
- To recommend the ways for OceanPC to accept software that is written in other computer languages. (OceanPC Project Leader, GTSPP Chairman).
- To consider steps to be taken to make OceanPC an exchange format for GODAR rescue operations. (GODAR and OceanPC Project Leaders, GTSPP Chairman).
- To identify ways in which GODAR and GTSPP will co-operate by sharing processing responsibilities. (GODAR Project Leader, GTSPP Chairman).
- To assist GODAR in re-formatting and processing new data so that it becomes available more quickly. (All GTSPP members).
- To share QC C-code software with data centres participating in GODAR data processing. (US NODC, all GTSPP members).
- To help GODAR to accept, QC and input re-saved data in the GTSPP database. (AODC, MEDS, AOML).
- To develop, test and document QC software for GODAR purposes. (US NODC).
- To provide a PC programme for key entry of data for the GODAR project and description of meta data. (MEDS, US NODC).
FORMATS AND META DATA
- To work with DACs to test adding 80-character free text groups to the GTSPP format. (MEDS, US NODC).
- To add free text groups to the GTSPP format to meet requirements for meta data which accompanies delayed mode data.
CONVERSION OF THE GTSPP QC SOFTWARE TO C AND UNIX
- To complete the duplicates management software. (US NODC).
- To test the portability of the GTSPP QC software for the database conversions. (Russian NODC, AODC, SIO).
QC REVIEW AND REVISIONS
- To review the text of Documnet IOC-WMO/IODE/GTSPP-III/16 "Proposals for Revision to the GTSPP QC Procedures for Real-time and Delayed Mode Data" and to distribute copies to science centres, WOCE IPO and the TOGA/WOCE SOC for their review in order to start implementation of agreed modifications in 1994. (MEDS).
PILOT OCEAN DATA NETWORK
- To establish 3 servers to meet the needs of GTSPP users:
(i) US NODC server to hold authority lists for ship codes and other GTSPP codes;
(ii) AODC server to hold a GTSPP software inventory and documentation available to other centres;
(iii) MEDS server to hold the MEDI catalogue and implement them on Internet.
- To implement a server for maps of temperature and salinity prepared with the GTSPP data. (SIO).
- To publish servers that are already available for some of the WOCE DACs and for SISMER in Brest in the GTSPP reports. (MEDS, WOCE and TOGA representatives).
SOFTWARE STANDARDS AND EXCHANGE OF SOFTWARE
- To develop an inventory of the 4 categories of GTSPP software and make it accessible via Internet. (AODC, GETADE).
- To adopt FORTRAN, C and X-windows as a standard software development environment for the GTSPP software development. (All GTSPP members).
- To investigate IFREMER, SCRIPPS and CSIRO visualizations software used in their quality control systems. (US NODC, AODC).
- To provide duplicate checking software to SIO for inclusion in the SIO quality control software application and to provide access to the system as a service for use. (US NODC).
- To inform IOC Member States and international organizations by a Circular Letter on the state and plans of the GTSPP implementation. (IOC).
- To review the GTSPP Project Plan of 1991 and make proposals for the need of a new modified publication (AODC, GTSPP Chairman, IOC).
- To make arrangements for GTSPP-IV in 1995 jointly with the WOCE UOT-DAC session. (GTSPP Chairman, IOC).
- To consider ways to attract new Member States not actively participating in GTSPP to assist GTSPP by accepting a variety of responsibilities in the development of delayed mode flow monitoring and in the historical aspects of this project. (GTSPP Chairman, IOC).
ANNEX III - LIST OF PARTICIPANTS
Ms. H. CAMERON
A/Manager, Library Services
Fisheries & Oceans
1026-200 Kent St.
Ottawa, Ontario K1A 0E6
Tel: (613) 993 2926
Fax: (613) 996 9230 Attn. Library
Tlx: 0533 4228
Mr. S. DIGGS
Scripps Institute of Oceanography/PORD
University of California, San Diego
La Jolla, California 92093-0230
Tel: (619) 534 1108
Fax: (619) 534 0704
Mr. D. HAMILTON (Co-Rapporteur)
National Oceanographic Data Centre
1825 Connecticut Av. NW, Room 422
Washington, DC, 20235
Tel: (202) 606 4636
Fax: (202) 606 4586
Ms. M. HAMILTON (Co-Rapporteur)
National Oceanographic Data Centre
1825 Connecticut Av. NW, Room 422
Washington, DC, 20235 U.S.A.
Tel: (202) 606 4636
Fax: (202) 606 4586
Omnet: NODC.WDCA or M.HAMILTON.GTSPP
Ms. P. HOLLIDAY
WOCE International Project Office
Institute of Oceanographic Sciences
Deacon Laboratory, Wormley
Godalming, Surrey GU8 5UB
Tel: (44) 428 68 4141
Fax: (44) 428 68 3066
Mr. J.R. KEELEY
Chief, Ocean Information & Systems Development
Marine Environmental Data Service
Fisheries & Oceans
12082-200 Kent St., Ottawa
Ontario K1A 0E6
Tel: (613) 990 0246
Fax: (613) 990 5510
Tlx: 0533 4228
Prof. D. KOHNKE
Bundesamt fuer Seeschiffahrt und Hydrographie
Postfach 30 12 20
Tel: (40) 3190 3400
Fax: (40) 3190 5000
Tlx: 215448 hydro d
Mr. N.N. MIKHAILOV
Head, National Oceanographic Data Centre All-Russia Union
Research Institute of Hydrometeorological Information
World Data Centre B
Russian Federal Committee for Hydrometeorology
6, Korolev St., Obninsk
Kaluga Region 249020
Tel: (7) (095) 546 3910 Fax: (7) (095) 255 2225
Tlx: 412633 INFOR SU Omnet: WDCB1.USSR
Mr. J-P. REBERT
TOGA Subsurface Data Centre
Centre ORSTOM de Brest, IFREMER
Tel: (33) 98 22 45 13
Fax: (33) 98 22 45 14
Mr. B. SEARLE
Australian Oceanographic Data Centre
Level 2, MHQ Annex
Maritime Headquarters, Wylde St.
Potts Point, NSW 2011
Tel: (61) (2) 563 4801
Fax: (61) (2) 563 4820
Mr. M. SIMMONS
National Oceanographic Data Centre
1825 Connecticut Av. NW, Room 422
Washington, DC, 20235
Tel: (202) 606 4591
Fax: (202) 606 4586
Ms. A-M. WILBURN
4301 Rickenbacker Causeway
Miami, Florida 33149
Tel: (305) 361 4336
Fax: (305) 361 4582
Dr. J.R. WILSON (Chairman)
Director, Marine Environmental Data Service
Fisheries & Oceans
12082-200 Kent St.
Ottawa, Ontario K1A 0E6
Tel: (613) 990 0264
Fax: (613) 990 5510
INTERGOVERNMENTAL OCEANOGRAPHIC COMMISSION
Mr. I. OLIOUNINE
IOC Senior Assistant Secretary
1, rue Miollis
75732 Paris Cedex 15
Tel: (33) (1) 45 68 39 63
Fax: (33) (1) 40 56 93 16
Tlx: 204461 UNESCO
LIST OF DOCUMENTS
Document Code Title
GTSPP-III/1 add. Timetable
GTSPP-III/2 Annotated Agenda
GTSPP-III/3 Not allocated
GTSPP-III/4 List of Documents
GTSPP-III/5 Reports of Participants on Progress in Implementing GTSPP
GTSPP-III/6 Overview of the Present GTSPP Data Flow
GTSPP-III/7 Monitoring of Real-Time and Delayed Mode Data Flows in the GTSPP
GTSPP-III/8 Progress Implementing the CMD in the US NODC
GTSPP-III/9 Progress in Implementing the Delayed Mode Aspect of the GTSPP
GTSPP-III/10 Review and Proposals for Further Developing the Historical Data Aspects of the GTSPP
GTSPP-III/11 Proposal for GTSPP Activities in Relation to Time-Series Data Sets
GTSPP-III/12 Draft Proposal for Publication of GTSPP CD-ROM Series
GTSPP-III/13 A Review of GTSPP Services to WOCE and TOGA
GTSPP-III/14 GTSPP Activities in Support of the GOOS
GTSPP-III/15 Proposal for GTSPP/Ocean PC Cooperation
GTSPP-III/16 Proposals for Revision to the GTSPP QC Procedures for Real-Time and Delayed Mode Data
GTSPP-III/17 Proposal for Implementing Pilot Ocean Data and Information Services on the Internet
GTSPP-III/18 Draft General High Level Standard for GTSPP Portable Software
GTSPP-III/19 Compilation of Existing GTSPP Software Available for Exchange
GTSPP-III/20 Review of GTSPP Formats and New Requirements
GTSPP-III/21 A Proposed Meta Data Strategy for the GTSPP
INFORMATION AND OTHER DOCUMENTSIOC/INF-863 Summary Report of the Second Session of IOC-WMO Steering Group on GTSPP (Obninsk, USSR, 15-19 July 1991) IOC WMO GTSPP Project Plan
IOC/IODE-XIV/3 Summary Report of the Fourteenth Session of the IOC Committee on IODE (Paris, France, 1-9 December 1992) GTSPP Quarterly Reports (latest issues)
IOC/GOOS-I/3 Summary Report of the First Session of the IOC Committee for GOOS (Paris, France, 16-19 February 1993)
IOC Manual & Manual of Quality Control Procedures for Validation of Oceanographic Data Guides No. 20
IOC/MIM-IV/3 Draft Summary Report of GE-MIM IV, Washington, DC, 6-8 October 1993
PROGRESS BY PARTICIPANTS IN IMPLEMENTING ASPECTS OF THE DELAYED MODE SYSTEM
U.S. NODCThe US NODC reported that it developed the Continuously Managed Database, first on a Sharebase database machine, and recently as a Sybase database on a Unix Workstation. The delayed mode data from 1990-1991 were quality controlled at NODC, using GTSPP software contributed by MEDS. These data were added to the CMD, but later removed at the request of WOCE UOT DACs. NODC has also developed data retrieval and inventory software which will be useful in managing delayed mode data in the CMD.
Prof. Kohnke reported that the DOD have been sending delayed mode data to the US NODC on a regular basis, for the purpose of addition to the CMD. He also expects closer working relations between those in Germany who manage operational data and those who manage delayed mode data. The GTSPP quarterly reports will be used in the future to improve data flows from Germany.
CANADAMr. Keeley reported on the progress which has taken place in Canada since the Obninsk meeting in 1991. It was reported that the increase in BATHY reports was not from ships but rather due to the increasing use of moored buoys.
A similar figure showed that the TESAC reports have increased but the volume of salinity data exchanged is still very small. MEDS has been assisting the Institute Maurice Lamontagne in sending the TESAC data over the GTS since August 1993. MEDS is also assisting Bedford Institute in the exchange of BATHY and TESACs through Inmarsat, as well as processing the historical data collected by the VOS Skogafoss.
MEDS receives approximately 1,000 stations each month from the Japanese fisheries and defense vessels and plans to include this data in the GTSPP data flow.
The US Navy is sending approximately 1,000 low resolution stations each month. These data represent the declassified messages which were never reported over the GTS.
MEDS has arranged to send the US Naval Oceanographic Office in Bay St. Louis, the real-time files 3 times each week. MEDS also plans to begin sending the real-time data files to the United Kingdom Meteorological Office for use in ocean modelling.
MEDS has processed 23,000 new delayed mode stations for inclusion in the CMD at NODC. TRACKOB data have not yet been included in the real-time data flow, which it had been hoped would be accomplished by this time.
As a final item MEDS is assisting the Argentine data centre with BATHY and TESAC statistics from the GTS.
RUSSIABeginning in November 1993, NODC of Russian intends to begin the monthly transfer of GTS data (received at RIHMI-WDC as RNODC IGOSS (BATHY/TESAC) and Navy Oceanographic Data Centre from the ships of the Russian Navy Hydrographical Service (BATHY)) to include them into GTSPP database in MEDS. RNODC IGOSS BATHY/TESAC data for January-August 1993 and Navy BATHY data have been submitted to MEDS. These data are recorded in the GE-TADE format.
It is necessary to co-ordinate the technical details of data transfer on a regular basis. NODC of Russia needs the financial support for regular exchange of data via Internet.
Now NODC Russia finished the preparation of the complete reference database with descriptions of R/Vs cruises, carried out for the historical period. Upon completion of the necessary actions on the base preparation (translation of dictionaries and help-files from Russian into English) Russia NODC is ready at the beginning of 1994, to exchange reference databases with USA NODC and ICES to verify data catalogues. It is efficient to define more exactly the technical details of this exchange (content of cruise description, format) as well as peculiar features and dates of carrying out the verification/comparing.
Russia NODC intends to begin the preparation of the long-term time-series datasets. The test data set for OWS "CHARLEY" has been made on the basis of all data available at Russia NODC/WDC-B1. Data are prepared in GE-TADE format. An additional QC of this dataset is required. This dataset is available for use under GTSPP. After checking of OWS "CHARLEY" dataset, Russia NODC proposes to prepare dataset with 10-year time-series of observations, made in the 70s - 80s, according to the national observation project "Sections" and submit them for the international exchange under GTSPP.
A member of Russian oceanographic organizations expressed their intentions to become users of the Global GTSPP database.
An Internet connection has not yet been implemented at the Russian NODC.
ORSTOM/BRESTThe implementation of the TOGA/WOCE Global Centre at Brest began in early 1993. The database contains real-time and delayed mode sub-surface data received since 1985 in 30 N - 30 S area for the 3 oceans. Current holdings are approximately 230,000 data with 60% being delayed mode.
The GTSPP real-time datasets are received monthly from the NODC.
The recent migration from binary to ASCII GTSPP format occurred while the system was being developed, therefore the entrance formats are written in binary. These modifications should be complete by the end of this year.
The documentation of the QC procedures are currently in French. MEDS has offered to translate the documentation to English.
The principles and codes proposed by MEDS to manage the History table have been adopted but because of the constraints to manage the delayed mode data, the history information has been reduced.
AUSTRALIAThe Australian Oceanography Data Centre (AODC) has recently implemented a computer system to manage, process, analyze and display ocean temperature and salinity data. This system is built around a commercial Geographic Information System (GIS) which uses ORACLE as its core management component. This system manages data in the originators format. Presently the available US NODC formats, including the GTSPP format and CSIRO and RAN data formats can be managed. Data flow mechanisms are well established with links to the Australian Bureau of Meteorology and via Internet to US NODC. Agreements exist with CSIRO to obtain data from their research vessel RV Franklin and their VOS lines. Arrangements also exist with Ocean Sciences Institute for data exchange.
Quality control procedures are dependent on development of software by CSIRO Division of Oceanography (AODC are assisting) which is taking longer than expected. It will be installed early in 1996. Real-time quality control procedures are in place but are not up to the GTSPP standard. This will be provided by the CSIRO software.
The Royal Australian Navy (RAN) is aiming to establish an oceanographic programme over the next few years. One component of this will consist of a survey of existing datasets from national and international agencies that are not held by AODC or presently available internationally. It is anticipated that an acquisition programme will follow.
AODC is establishing close relationships with oceanographic agencies within both Malaysia and Indonesia. These relationships should assist with the acquisition of delayed mode data from S.E. Asia.
CHINAParticipation in the Global Temperature and Salinity Pilot Project in China is by the National Marine Data and Information Service.
As background, the international oceanographic data management in the National Marine Data and Information Service (NMDIS) of China is an effort to promote and improve data management activities in China and the international society. China has positively participated in national and international oceanographic data observation and management programmes, for example, WOCE, TOGA, TOGA COARE, and the China-Japan Kuroshio Programme. As a member state of IODE, the WDC-D Oceanography in NMDIS provided a number of oceanographic datasets to Member States. All the data exchanged from the Member States and collected in China were checked by using serial quality control procedures and were merged into existing datasets. The oceanographic datasets are in a data management and services system developed by NMDIS in 1986 to ensure and improve the capability of the oceanographic data exchange and management and the ability to contribute to the Global Temperature and Salinity Pilot Project (GTSPP).
Recognizing that the GTSPP significantly improves the current IGOSS/IODE data flow in real-time data exchange, the compilation of the global temperature and salinity dataset, the circulation of newly developed standardized quality control algorithms, and the introduction of new state of the art in communications and computer applications, etc., China expressed its interest in the GTSPP and its willingness to participate in and contribute to the GTSPP positively at the Session of the IOC Assembly (7-21 March 1991, Paris). In 1992, NMDIS sent an expert to the US NODC for 6 months of joint-research on the GTSPP. He worked extensively with the GTSPP database and designed the algorithm and the programmes of Voluntary Observing Ships (VOS) XBT track lines for utilization with WOCE and TOGA data. He also developed the computer programmes on UNIX workstations to perform the oceanographic computations using UNESCO algorithms. NMDIS has the ability to contribute to GTSPP with its experience gained through activities as the WDC-D Oceanography in following areas.
- To digitize the data in manuscript and/or analog form and acquire the data in digital form not presently available to the international scientific community.
- To undertake quality control checking according to the GTSPP Quality Control Manual on the data collected.
- To develop data and information products.
- To provide advice and assistance to the GTSPP in the areas of data management, quality control, communication, and development.
Concerning quality control of historical data, NMDIS hopes to co-operate with other participants of the GTSPP to develop the standardized quality control algorithms for the historical data.
NMDIS hopes to co-operate with the US NODC to develop the following information and data products for the GTSPP. The standard of software development will be kept consistent with the needs of the GTSPP and the OceanPC.
(i) Cruise summary reports
(ii) Monthly data inventory reports
(iii) Monthly data distribution maps
(iv) GTSPP SST maps, standard depths ST maps, and anomaly value maps of them.
(v) Derived data products of the GTSPP database
- temperature, salinity, and density for standard depths
- climatic temperature and salinity characteristics by different degree grid points
(5 degree square, one degree square, etc.)
IGOSS AND IODE SYSTEMS DATA MANAGEMENT OBJECTIVES IN SERVICE FOR GOOS
(A Draft Proposal as submitted by Mr. B. Searle, Australia)
IntroductionData Management, Data Distribution and Products and Services are key areas of the Global Ocean Observing System (GOOS). These components, particularly the products and services are the most publicly visible and therefore must be implemented in a way that the public (and politicians) acknowledge as "value for money".
In general terms, the data generated by the observing system has two primary uses. The first is the acquisition of data in near real-time to generate both synoptic and a predictive range of products. This is an obvious area of "public benefit". The second is the archival of all GOOS derived data and the generation of time-series databases and climatological products. The second compartment is directed more at the scientific community and in some instances they generate products and services through research. These products may or may not have direct 'public value'.
The maintenance of quality control is an integral component of any system whether it is operational or research oriented. Quality control is usually not a highly visible activity but both the public and the research community have expectations that certain standards are met. Australian GOOS activities must include suitable quality control mechanisms in all areas of activity.
There is international consensus that the future Global Ocean Observing System (GOOS) will be developed where possible from existing meteorological and oceanographic service programmes. The possibility to do this for the areas of data management and dissemination is considerable, given a relatively well established international infrastructure.
REAL-TIME DATA DISTRIBUTION SYSTEM
BackgroundThe Intergovernmental Oceanographic Commission (IOC) and the World Meteorological Organization (WMO) operate, either jointly or alone, 5 major international ocean observing and data management systems which incorporate a real-time data distribution component. These comprise:
a. Joint IOC-WMO Integrated Global Ocean Observing (IGOSS).
b. Joint IOC/WMO Global Temperature and Salinity Pilot Project (GTSPP).
c. IOC Global Sea-Level Observing System.
d. Joint WMO-IOC Drifting Buoy Programme.
e. World Weather Watch.
A number of activities are undertaken by Australian agencies in support of these programmes. These include the CSIRO Division of Oceanography's Ship-of-Opportunity Programme (supported by the Navy), National Tidal Facility (NTF) Tide and Sea-Level Monitoring Network, Australian Oceanographic Data Centre (AODC) data management programme, Bureau of Meteorology's drifting buoy and meteorological observing programme. Other marine monitoring activities are undertaken on local or regional scales.
To make available all data collected within the GOOS programme to the widest range of users, at the highest quality and in as close to real-time as technologically possible.The objective of AGOOS is to build on the existing real-time data distribution activities by integrating the data collected by smaller scale programmes conducted at State or local levels and provide a more co-ordinated and robust real-time data distribution capability. This will require the development of a data management framework under which these activities can be unified. This framework must include an operational communications network for the acquisition and dissemination of data and products, improved management capabilities incorporating quality control and more effective product generation.
There is a requirement for an operational communications network, similar to AARNet but used exclusively for operational purposes. This communications network is needed to support a real-time distribution system that provides for the acquisition of data and dissemination of products from existing and future monitoring programmes. An important component of the goal is the need to provide more data in a timely manner, to both national and global users to support the generation of products. Australia is already an important contributor to global activities because of its geographic location.
StatusThe major (national and international scale) mentioned above, generally have well established data distribution systems. However, the existing smaller scale activities and developments to provide for presently unsupported requirements need to be integrated into an Australian wide systems, with links to the appropriate international data distribution systems.
Classification of ElementsA real-time distribution has 5 main components:
1. Data collection mechanisms (this will be addressed by the other AGOOS Goals).
3. Quality control.
4. Data processing.
5. Data distribution.
The communications component is the area requiring the most development since the other activities are already undertaken to some degree and require only high level integration and co-ordination.
DATA ARCHIVING AND MANAGEMENT
BackgroundOne of the primary purposes of GOOS is the development of datasets that can be used for a wide range of purposes including determining marine environmental trends and to assist government in developing environmentally sustainable economic policies. Given that a significant level of resources will be devoted to the collection of marine data, it is essential that once collected, the data be available for as wide a range of user applications as possible.
The management of data collected by GOOS will cover a range of scientific fields from living to non-living resources. The types of instruments used and the expertise required to understand and effectively manage the data is extremely broad. There is no single agency in Australia that has the data management skills and scientific expertise to manage all GOOS-related marine data. The only effective mechanism for a national scale capability is to create a series of AGOOS data management centres within existing (operational?) agencies that have the necessary specialist skills for particular data types and data management expertise.
A distributed data network linked by an operational communication facility (Note: AARNet is a research oriented capability) and co-ordinated by a "committee" of user agencies is required to facilitate this objective.
To create an easily accessible distributed data network to archive, manage, safeguard and disseminate marine data collected as part of the GOOS monitoring activities, in the national and international interest.To achieve this goal the following activities need to be undertaken:
1. Form a GOOS data management group to establish data management and archiving principles, provide a standard (within the Common Spatial Data Transfer Standard) for the exchange of marine data and ensure the co-ordination of developing efforts within agencies.
2. Strengthen the existing marine data management capabilities.
3. Expand the types of data managed to include biological, chemical and geological, as appropriate.
4. Undertake research and development to provide data management capabilities to effectively store the more complex or less standardized data types, such as biological data.
5. Develop a co-ordinated infrastructure of distributed data management agencies responsible for the management of specialist data types.
6. Create an electronically accessible marine data and information inventory to improve the access to the databases.
StatusThe management of "bluewater" physical oceanographic data is being undertaken by the AODC (and CSIRO Oceanography via WOCE) who also co-ordinate the international exchange of this data via IODE and other programmes. Marine meteorological data is also effectively managed by the Bureau of Meteorology and other agencies manage data on local or state scales. Various national and state agencies have marine data capabilities but these are generally developed to support in-house requirements and would need to be extensively enhanced to make them suitable to form part of a national operational data management network. Also, at the present time, there is no appropriate communications network to allow the creation of a "seamless" marine data management system.
In general, the other types of data, particularly biological data are not yet managed in a co-ordinated way. This is primarily due to the lack of standards in data management and inadequacies of mainstream computing technology to cope with complex data types. To overcome this difficulty, research and development into managing complex data structures in a spatial data management environment and appropriate standards to facilitate data exchange will be required.
PROVISION OF REAL-TIME SERVICES AND PRODUCTS
BackgroundA large number of marine-user groups have a requirement for real-time products and services. These range from coastal engineering activities to recreational boating, fishing, the off-shore resource industry and meteorology.
The services and products generated through an AGOOS monitoring programme must be focussed on specific user needs. This component will be one of the more visible outcomes of AGOOS and regardless of the quality of the science, the applicability of products to the user will be used to judge the success of AGOSS activities. Following from this, the continued funding support for an on-going monitoring activity will possibly be dependant on the governments (plus industry and public's) perceived success of the products and services generated.
It is very important that the provision of products and services is based on sound
scientific principles but the activity should not be operated by scientific agencies. This
production and dissemination role must be carried out by an operational agency('s) with an
interest in supporting one or more of the required maritime activities. At this stage the main
national scale operational agencies are the Bureau of Meteorology and the Royal
To provide timely and operationally significant products to support the full range of maritime activities with an emphasis on supporting both industry and government, in developing marine oriented environmentally sustainable development and economic processes.
StatusA number of products used to support a range of marine activities are presently produced by the Bureau of Meteorology and the Meteorological and Oceanographic (METOC) Services of the RAN. These include weekly analyses of the thermal and current structure of the Western Tasman Sea, weekly wave, SST, surface wind information for the Australian region.
The RAN is in the process of significantly upgrading its capabilities and additional capabilities generally based on new generation satellites will be implemented by agencies such as the Bureau of Meteorology as the data becomes available in operational mode. There is considerable potential for producing a large number of operational products in the near future.
Classification of ElementsThe elements for producing real-time products and services include:
1. Access to "real-time" databases.
2. Scientific support in the development of products.
3. On-going commitment from operational agencies.
4. An effective product dissemination mechanism.
MAINTENANCE OF QUALITY CONTROL
BackgroundThe quality control of marine data is a fundamental element of the AGOOS activity. Having access to data is of limited value unless information is known about that data such as, the instrument used, its level of calibration and the quality control procedures conducted on the dataset. It is essential that the users of the AGOOS datasets be aware of the quality of that data.
A continuous quality programme is required to support all aspects of the AGOOS data requirements. Quick and effective quality procedures are needed to support real-time users and a slower, more detailed examination of the data is needed to support the scientific users of the data. These two levels of quality control can be established by using the principle of the Continuously Managed Database (CMD). The CMD is updated as soon as data is received by the managing agency. It is then run through "quick look" automated checks and it is available to the users within a few hours of receipt. In slower time, computer supported manual reviews in conjunction with more sophisticated statistical techniques are undertaken to more accurately qualify data.
An important concept in the quality control process is that NO data should be thrown out. Data may be labelled "unacceptable" but should not be removed from the database for any reason. The quality control procedures should also incorporate scientific validation and must address both real-time and non real-time requirements.
To provide a scientifically acceptable quality control mechanism to create validated datasets to meet both real-time and delayed mode data and information needs.
StatusSuitable quality control procedures have been adopted by the agencies undertaking activities within their existing monitoring or data collection programmes, such as sea-level monitoring and the Ship-of-Opportunity programme. These quality control mechanisms have evolved over lengthy periods of development and in some cases, conform to internationally recognized standards. However, a considerable effort will be required to establish acceptable procedures for AGOSS components not as yet as well co-ordinated as the so existing programmes.
Classification of Elements1. Operational agencies.
2. Scientific agencies.
LIST OF ACRONYMS
AARNET Australian Universities Academic Research Network
AGOOS Australian Global Ocean Observing System
AODC Australian Oceanographic Data Centre
AOMC Australian Oceanographic Marine Centre
AOML Atlantic Oceanographic & Meteorological Laboratory
BATHY Bathythermograph Report
BODC British Oceanographic Data Centre
BSH Bundesamt für Seeschiffahrt und Hydrographie (Germany)
BUFR Binary Universal Form for Representation of Meteorological Data
CDF Coded Digital Facsimile
CDS Computerized Documentation System
CMD Continuously Managed Database
COARE Coupled Open Atmosphere Response Experiment
CSIRO Commonwealth Scientific & Industrial Research Organization
CSR Cruise Summary Report
CTD Conductivity, Temperature, Depth
DAC Data Assembly Centre
DOD Deutsches Ozeanographisches Datazentrum (Germany)
ESA European Space Agency
FNOC Fleet Numerical Oceanography Centre (USA)
GE Group of Experts
GEMIM Group of Experts on Marine Information Management
GETADE Group of Experts on Technical Aspects of Data Exchange
GIS Geographic Information System
GODAR Global Oceanographic Data Archeology & Rescue
GOOS Global Ocean Observing System
GTS Global Telecommunication System
GTSPP Global Temperature Salinity Pilot Project
HDF Hierarchical Data Format
ICSU International Council of Scientific Unions
IFREMER Institut Français de Recherche pour l'Exploitation de la Mer
IGOSS Integrated Global Ocean Services System
IMS International Marine Science Newsletter
INF Information Document (IOC)
IOC Intergovernmental Oceanographic Commission
IODE International Oceanographic Data & Information Exchange
ITPO International Toga Project Office
JGOFS Joint Global Ocean Flux Study
MEDI Marine Environmental Data Information Referral System
MEDS Marine Environmental Data Service (Canada)
METOC Bureau of Meteorology & Oceanographic Services (Australia)
MPERSS Marine Pollution Emergency Response Support System
NCAR National Centre for Atmospheric Research (USA)
NESDIS National Environmental Satellite, Data & Information Service (USA)
NMDIS National Marine Data & Information Service (China)
NOAA National Oceanic & Atmospheric Administration (USA)
NOC National Oceanographic Centre
NODC National Oceanographic Data Centre
NTF National Tidal Facility (Australia)
OCEANPC Ocean Personal Computer Project
ORSTOM Institut Français de Recherche Scientifique pour le Developpement en
OWS Ocean Weather Station
PICES North Pacific Marine Science Organization
PMEL Pacific Marine Environmental Laboratory
RIHMI All-Russian Union Research Institute of Hydrometeorological Information
RNODC Responsible National Oceanographic Data Centre
SIO Scripps Institute of Oceanography (USA)
SISMER Marine Scientific Information Systems
SOC Specialized Oceanographic Centre
TESAC Temperature, Salinity & Currents Report from a Sea Station
TRACKOB Report of Marine Surface Observations along a Ship's Track
UNCED United Nations Conference on Environment & Development
UOT Upper Ocean Thermal Project
VOS Voluntary Observing Ship
WDC World Data Centre
WHOI Woods Hole Oceanographic Institution (USA)
WMO World Meteorological Organization
WOCE-IPO World Ocean Circulation Experiment-International Project Office
XBT Expendable Bathythermograph