5 DATA PRODUCTS, SERVICES, ARCHIVES AND ACCESS

TABLE 1 SHOWS A LIST OF THE MAIN PRODUCTS THAT SHOULD BE THE FOCUS FOR GLOBAL TERRESTRIAL CARBON OBSERVATION INITIATIVE DURING THE INITIAL PERIOD. AS NOTED IN SECTION 3.2, MANY OF THESE OR SIMILAR PRODUCTS ARE NOW GENERATED OR BEING PLANNED OVER VARIOUS GEOGRAPHIC AREAS (APPENDIX 2). The raw data originate from different agencies (satellite and in situ, different countries), and the preparation of individual products may require access to one or more other data types or input products. Also, the data and products exist at various levels, from raw observations to synthesized high level products generated by incorporating lower level products into models.

The above considerations imply that to be effective, the IGOS carbon cycle theme must be based on a coordinated approach to data and information system issues. Importantly, this would also facilitate secondary and spin-off uses of the products, thus enhancing the impact of the IGOS initiative. Appropriate models exists, for example, GCOS products are generated through cooperation among primary data collectors, data quality monitoring centres, and product generation/analysis centres (refer to GUAN^[20]). The IGOS Ad Hoc Working Group on Data and Information Systems and Services also considered the generic issues involved and developed principles that should be followed. These principles will need to be applied to the carbon observation theme as a whole as well as to the contributing operations of the individual participating agencies; a way forward is proposed in Chapter 6.

+ NOTES:

1. The table presents an overview of major products for terrestrial carbon estimation that are needed for the TCO initiative. Most of these employ satellite data. Other observations or products are also needed but not listed here. They include meteorological products soil databases, and others (Cihlar, Denning and Gosz (2000); refer also to Chapter 4.).

2. Many of the above products are now produced by space agencies, research projects, or observation networks. Most of the products are being continually improved through product validation, supporting research and model development, and the reprocessing of archived data. The agencies and projects involved in these typically have specific plans for improving the products. Importantly, new programmes are now being developed in various countries that will make further progress in these areas.

3. Where different geographic areas are shown (spatial extent), the quality and documentation of the products is higher at the site and regional levels. The improvements of global products hinges on progress being made through regional or campaign initiatives.

4. Current limitation is lack of consistent measurements over large areas; progress depends on improved satellite measurements.

5. Current products have coarse spatial and temporal resolution.

6. Experimental ecosystem productivity maps have been generated for specific areas and years, using the above products as input (see Ecosystem productivity, p.18). The NPP (annual gross carbon uptake) and NEP (annual source/sink distribution) products are presently model- dependent and are also limited by the quality of the existing input products. Plans are in place to generate global NPP products, using new satellite data.

All product generation, archiving and distribution centres should be accessible through Internet. The generally available low-speed Internet links should facilitate access to output products and to low-volumes of raw data/low level products. In addition, as many as possible centres should be connected with high performance, wide bandwidth links. High performance research and education networks (e.g. NASA NREN, Asia Pacific Advanced Network), now available in several countries, can be used to facilitate exchange of high volume data among the participants and the generation of products. Internet will facilitate product access to policy users, conventions, and others. In addition to existing agencies (Appendix 2), new contributors will also be able to join, such as: research agencies; operational agencies; World Data Centers for trace gases, meteorology, renewable resources and environment^[21]; and educational institutions. Special attention will need to be given to in situ data, as their timely availability is critical for various derived products and the widely distributed points-of-entry present special challenges. The detailed arrangements for product generation, quality control, access and archiving will need to be worked out among the contributors based on the principles defined by IGOS-DISS. Cihlar, Denning and Gosz (2000) described several existing tools and mechanisms that could be used in this process.

An important advantage of using Internet is the ongoing growth of up-to-date communication and data exchange capabilities, as the technical performance of individual sites undergoes ongoing enhancements. This will also facilitate the transition from research to operational status as the observation networks and products mature, since the technology employed will remain similar. Another advantage is the ability to make the products widely known and accessible to a large audience worldwide, thus increasing the impact of the actions by IGOS-P and other contributors. However, for the Internet-based approach to function effectively, it will be necessary that data handling and archiving policies are harmonized, and that data and products are easily located in the holdings of the various participants. A single-point entry should be established that provides an overview and links to other sites. Suitable technical capabilities have already been developed by CEOS/WGISS^[22], GOSIC^[23], ORNL^[24], and others. They will need to be employed in a coordinated manner by the agencies participating in the IGOS carbon cycle observation theme.

The required basic data products for the terrestrial domain have been defined elsewhere (Cihlar, Denning and Gosz, 2000; Chapter 4.). Agreements will have to be reached regarding data products, documentation, archiving contributions and responsibilities of the participating agencies, and access to data and products. Data documentation formats have been established by various initiatives (e.g. FDGC^[25]; GCMD^[26]) and are becoming widely used; these should be considered for adoption by the carbon observations. Short- and long-term archiving issues will need to be worked out among the participating agencies. IGOS-DISS recommendations (IGOS Ad Hoc Working Group on Data and Information Systems and Services, 2000) regarding access to data and products should be followed. Based on the experience of many groups, it is also essential that the data processing and archiving arrangements accommodate reprocessing of archived data, permitting generation of better products as the algorithms and models improve and thereby increasing the impact of the initial investments at a modest additional cost.

Assuring that the data and products have a defined and understandable quality assurance attached to them is highly important. Furthermore, for long-term continuity, the history of quality control needs to be well described. The objective should be to ensure that a potential user knows and can rely on the quality assurance assigned to a product or data set. For complex products this can be a difficult task but is essential. One means of achieving better quality control is to establish a standardised method for calibrating sensors used in the Observing System.

The issue here is for data providers to describe adequately and in a uniform way the quality of their data and products, and to subscribe to a uniform standard of calibration procedures for sensors.