The JPO Director briefly reviewed the seven actions originating from the previous DIMP meeting, and noted where progress had been made, and where attention is still required. For example, the issues of a GCOS Data and Information System and the registration of data sets remain items for discussion (see 5.5). Several of the action items were addressed through the development of specific proposals during the intersessional period. Unfortunately the two submitted (e.g., a workshop for data centres, an information centre) were not funded during the year. Dr Spence invited further discussion on them later in the meeting. With regard to the item on data principles, the participants accepted and endorsed the guidelines proposed by the Chairman at earlier meetings (see Annex IV). A potential project on quality control for ocean observations was developed during the interim, but not submitted as yet (see 5.4). The various other items raised by the Director were dealt with in appropriate agenda items to follow.
The JPO Director noted the discussions on DIMP at the two most recent meetings of the JSTC. Their recommendations to DIMP had mostly been addressed, although a continuing close interaction between the two groups was stressed. He invited the participants to consider mechanisms whereby the panel provides its recommendations to the G3OS. For the GCOS JSTC, the panel Chairman plans to attend a special session at which the various GCOS panel chairmen will report their progress, plans, and problems, which will be tabled for discussion at the JSTC plenary session on subsequent days. Reports by the Chairman to the GSC and GTOS SC were encouraged.
The Chairman reported on the workshop on “Indices and Indicators for Climate Extremes”, that was proposed at DIMP-II. The workshop, sponsored by CLIVAR, GCOS and WMO, was hosted by the U.S. National Oceanic and Atmospheric Administration (NOAA) National Climatic Data Center (NCDC), in Asheville, NC, 3-6 June, 1997. Nearly 100 scientists and specialists from the insurance and reinsurance industry from 23 countries participated in the meeting. It addressed a number of issues including:
Generating more comprehensive and reliable information for the IPCC (Year 2000) assessment of climate change related to the question “Is the climate becoming more extreme?”;
Identifying critical data sets required to address changes in climate extremes;
Identifying, suggesting, and finding ways to correct inhomogeneities in the databases required to address the question of changing climate extremes;
Identifying specific indices or indicators that capture the multidimensional aspects of changes in climate extremes and can be easily calculated and readily understood;
Finding ways to ensure that access to high-quality observations continue for those indices and indicators of primary interest;
Identifying the type of information the insurance and reinsurance industry requires regarding changes in climate extremes.
The Chairman reported that the meeting accomplished a number of specific and tangible goals. About 20 papers were identified for publication in Climatic Change (pending peer review) and a reference textbook. Publication is scheduled for September 1998. Agreement was reached on a minimum set of indicators of climate extremes and the data required to calculate them. Regional rapporteurs and working groups were established for developing data sets and coordinating calculation of indices and indicators. The assistance of the CCl/CLIVAR Climate Change Detection Working Group was obtained to address special data homogeneity needs and to resolve research issues needed to calculate the indices and indicators agreed at the meeting.
A number of action items remain for follow-up. These include to:
Formalize regional task groups and rapporteurs with a letter from the GCOS/CLIVAR offices;
Solicit the cooperation of projects such as the System for Analysis, Research and Training (START), the Asia-Pacific Network for Global Change Research (APN), the Inter-American Institute for Global Change Research (IAI), and WMO’s DARE to enhance data coverage within several data-sparse regions (e.g., South America, Africa, Southeast Asia, etc.);
Identify location and sponsorship (funding sources) for a follow-up workshop in early 1999;
Identify computation and data archive centres for indices and indicators;
Continue the communication with and encourage the active participation of the insurance and reinsurance industry.
Mr Scholefield updated the panel on discussions held in Melbourne the preceding week at an informal meeting of WGCCD. The WG endorsed the outcome of the indices workshop and particularly supported the development of indices for climate change.
In the following discussion, participants urged that non-meteorological indices also be considered, and that industrial contacts be made judiciously. Additionally, it was noted that the OOPC is developing a suite of brochures to describe the products of use to users, and some of them will incorporate indices. The meeting, In situ Observations for the Global Observing Systems (GCOS-28), flagged indicators and indices as important elements to ensure the end users receive useful information. The role of the JDIMP should be to ensure that the appropriate data are available for use in developing the indices.
Mr Keeley noted that the last DIMP meeting discussed a project to assemble an integrated time series of Sea Surface Temperature (SST) from a number of sources. The proposal was presented to WOCE Panels and to the Chairman of the OOPC. The response was somewhat subdued. The Chairman of OOPC suggested an alternate theme - to quantify the value added to SST data as a result of quality control. This SST project was developed further resulting in the suggestion of a workshop to determine how such a valuation could be done. This idea was re-circulated once more and was more favourably received. Two members of the WOCE Upper Ocean Thermal group volunteered to undertake such a study in the limited domain of work they were starting. Some preliminary results from one of the studies undertaken by Dr R. Molinari of the NOAA Atlantic Ocean Marine Laboratory (AOML) in Miami were noted. He first calculated mean vertically averaged temperatures in 2 x 2.5 degree rectangles in the North Atlantic. The calculations used data with no quality control applied, but outliers removed by a 2-s test. Then he made the same calculation using data that had received quality control. Anomalies from these means were then calculated for each box. As might be expected, results using these two techniques did not differ greatly. The main difference was that the data with quality control applied tended to show greater variation and more abrupt changes from positive to negative anomalies (and vice versa). More work is required to further quantify these results and to evaluate the differences for this and for other products. Factors which could be considered in future studies include: 1) the impact of quality control on the definition of space and time scales, and 2) uncertainties in these derived from data with and without quality control applied. It is expected that there may be no single alternative to quality control, such as removal of outliers by a 2-s test[6], but the value of quality control may be dependent on the products to be generated.
It was agreed that no immediate action by the panel is needed. Further studies are expected to be done by others and reports of these results will be made to the next meeting. After this work is completed, the panel may wish to reconsider the need for a workshop to more fully explore this question, or simply to request a document that expresses the principles that apply when data with or without quality control are used for a variety of products. One agreed principle was stated - original data must be retained to enable future work to be done effectively.
At the DIMP-II meeting, a project was proposed to begin implementation of an on-line “data system” for climate data. One of the goals of the project was to implement a data system from a list of candidate systems through a workshop ‘vetting’ approach. Although such a process may provide some desirable learning opportunities, the Chairman noted that it now seems quite bold to have planned to select a single “data system”. At this time, he observed that there had been limited progress in moving forward on this workshop and with the new responsibilities of JDIMP, noted that it may be undesirable to carry through with it. Past experience with major data systems (e.g., the National Aeronautics and Space Administration (NASA) Earth Observation System Data Information System (EOSDIS) was focused on a single set of complementary space-based measurements and was much more limited in data sources than is currently of interest to JDIMP) has made it clear that it is very difficult to tie any computer data management system to existing technology or data transfer protocols, especially since data translation software is now abundant and inexpensive. The Chairman posed the issue: What do we mean by a “data system”? The one outlined in the GCOS Data and Information Management Plan was quite generic and included a variety of approaches with appropriate linkages. He invited the panel to discuss the issue and decide whether it wants to provide the oversight for linkages and capabilities among existing data systems or to encourage development of a distributed system with common software. He noted the difficulty and suggested that the latter approach may not be desirable. He challenged the panel to decide whether the workshop proposed at DIMP-II remains viable for JDIMP.
Participants noted that using the term “data system” served to delineate what was in the system and what was not. Since the G3OS are building from existing components, it may be counterproductive to be exclusive. Rather, what should be developed is an effective linkage among the various data sources and providers to serve the users’ needs. Thus, a more “integrative” approach was supported. In particular, a workshop as proposed earlier was not supported by the panel.
Dr F. Webster described the proposal made to establish a GCOS Information Center (IC). He began with a review of the recommendations from DIMP-II where it was envisioned that such a unit would be part of a distributed data and information system. It would allow users to locate and access data and information holdings at participating centres. The IC would provide search capabilities and access to a world-wide set of observations and information; it would point to other centres and systems; it would provide access to data, but would not hold data itself. He argued that the time is ripe to start such an activity, as it would provide immediate evidence of a G3OS data system. He noted that the data centres are ready and would agree to serve as a core for the system development. Once having it in place, others would be encouraged to join. As it evolves, the services would broaden. The specific proposal included the tasks of establishing an internet site which would contain information about data centre holdings, and would provide an ombudsman to assist users as appropriate. Dr Webster invited the panel to comment on these needs and to determine if a project for an IC should be advanced, and if so, how it should incorporate the non-climate elements of GOOS and GTOS.
Participants noted that spokespersons for the non-climate elements should be invited to comment on the proposal, and in fact, the various panels and working groups of GOOS and GTOS should be invited to review the proposal and make comments before it is submitted for funding. Particular issues were raised with the prospect of an ombudsman function, the diversity of data sets to be incorporated from the GTOS perspective, and with the potential for the IC to enjoy a long-term commitment and truly be ‘operational’.
After considerable discussion, it was agreed that Dr Webster would modify the proposal to incorporate non-climate aspects, and that it would be circulated to principals of GOOS and GTOS for their input. A discussion of the proposal would be scheduled at the JSTC-VII and a final decision taken on the project.
Mr V. Vent-Schmidt provided information on the status of the implementation of the GSN and reported on the second CBS/CCl expert meeting which took place at De Bilt in the Netherlands, 25-27 June 1997. The proposed network of about 1000 stations had been distributed to the National Meteorological and Hydrological Service (NMHS) to obtain their agreement and commitment. The experts reviewed the available responses from WMO Members which generated a number of questions such as the automation of stations, and the use of data from cooperative stations. The experts felt it necessary in future correspondence to clarify the requirements for a GSN station such as the routine monthly transmission via the GTS (e.g., using the report of monthly means and totals from a land station (CLIMAT) messages), and the provision of historical records together with appropriate metadata.
The experts accepted the proposal of member countries for substitution of and/or additional stations if they meet the requirements. The experts made specific recommendations to be discussed at the Panel meeting and the CCl meeting in August 1997. These include the establishment of a website for the GSN and the designation of a site as a repository of the historical climate data. The experts also felt it necessary to ensure the status of the GSN data as being “essential data” to be certain it is covered by the provisions of WMO Resolution 40.
The revised version of the GSN will be presented by the relevant rapporteur to WMO and the WMO Regions for approval. As far as data handling is concerned, Mr Vent-Schmidt proposed that the Deutscher Wetterdienst monitor the GSN data flow and include considerations of data availability and data quality with special regard to precipitation. A back-up centre should be invited to provide a similar service with regard to air temperature. The draft report of the meeting prepared by the rapporteur was available at the meeting[7].
The GCOS Upper-Air Network, consisting of about 150 stations, has been agreed upon by the appropriate National Meteorological Services through the WMO. The collection of data is being monitored on behalf of GCOS by the European Centre for Medium-range Weather Forecasts (ECMWF). Further discussion will occur at the AOPC-III in the UK in August.
Dr A. Jose reminded the panel that one of its key issues in the strategy for GCOS is the increased participation of developing countries. She also recalled that during the DIMP-II, it was recommended that a seminar/workshop be held in the Southeast Asian Region. A project document on this topic was to have been prepared, but little progress has been made. Dr Jose encouraged the panel to revitalize and support this project. She noted that prospective participants would be end-users of data products and information, particularly the socio-economic sectors, policy-makers, and environmental scientists. At national levels, this awareness activity would enhance support for their own governments in improving existing observation systems and possible eventual participation in the G3OS. Dr Jose provided the participants with some information on initiatives existing in the Southeast Asian Region which may be useful toward development of the seminar proposal.
In discussion, it was noted that the G3OS programmes must promote capacity building in developing countries. However, it was noted that all of the G3OS sponsoring organizations have active and effective programmes to address specific issues for their member states and constituencies. These, along with programmes such as START, APN, and IAI, should be utilized by the G3OS as the foundation for work on behalf of developing countries.
|
[6] It was noted that the
'ozone hole' was not identified earlier due to the use of such
criteria. [7] The final report (GCOS-35) is now available on the GCOS homepage or from the JPO. |
