Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
2. Information for Global Change Global change studies are interdisciplinary and the issues associated with such studies are often long term. Data and information will be needed, not as an end in themselves, but as a means for gaining understanding of the global system. Many requests to the data management system will be for derived products such as analyses or edited data collections in asso- ciation with descriptive text or graphical material, rather than raw observational data. Thus, the system must provide such products (here described as information) as well as data. It must keep the raw data as well, in order to provide the material for future reanalyses. To fulfill the role of information management, it must play an active role in the generation, acquisition, quality control, dissemination, and retention of value-added products. Priority Requirements For most global change studies, regional and global data and information will be required. No one nation, agency, or institution will be able to gather the appropriate data without cooperation from other nations, agencies, and institutions. For the United States this means that any one agency will often need the cooperation of others to produce datasets for global change. Moreover, the U.S. Global Change Research Program (USGCRP) will require international cooperation and data exchange. A policy will be needed that ensures unrestricted exchange of global change data across national boundaries. The principal scientific activities having data and information system requirements are as follows: 1. Long-term measurements and derivation of products. Long- term measurements of the global environment will form a core part of 12
INFORMATION FOR GLOBAL CHANGE 13 the USGCRP. Products and information derived from the data should play a principal role in judging the state of global change. Measurement teams should be responsible for end-to-end integration from initial measurements to final products. The team should include data managers. The process should be forward-looking, with an emphasis on obtaining reliable measurements. It should include a review process that takes into account the need to document long-term global change. Examples of such activities are the long-term studies of surface temperature. 2. Process studies. A process study may have its own data management system. The relevance of such a project comes from the final step of codifying the understanding through new algorithms or new measurement techniques. Additional funding outside the project will be needed to guarantee, with a scientifically knowledgeable representative of the archival system, the useful transfer and preservation of the project data. An example of a process study is the Tropical Ocean-Global Atmosphere experiment . 3. Model assimilation. Model assimilation involves the use of physical laws as represented in a numerical model to reduce a set of disparate but related data to a uniform representation with known levels of credibility. The information content of the derived product comes in part from the data and in part from the nature of the numerical model that is used. The distinguishing characteristic in this process is the integration of disparate datasets from multiple data sources and frequently the need for those data to be available in near real time. Product archiving is the final data management activity. A critical factor is documentation of the model output . 4. Non-global-change data. A number of existing data sources and holdings are not obviously relevant to present high-priority global change science activities. Yet they are clearly a valuable resource with major potential value in the future. To increase the chances for wise stewardship, we need to set in place disciplinary panels to advise data
14 INFORMATION FOR GLOBAL CHANGE center managers on data acquisition and retention policies and priorities. Data needed for broad global change research run the gamut from site-specific to truly global datasets. Disciplines involved include earth, atmosphere, and ocean sciences. Many datasets will serve multiple disciplines and will need multiple talents to assemble. Time series data of all types are used prominently in global change research to detect past and present trends. Proxy data may serve when direct measurements are impossible. The need for precision is important, as some of the predicted changes will be small and will take place over long periods of time. Often datasets will be enormous, due to the resolution and spatial scale needed to address global issues. A common attribute of many of the datasets is that they must be accessible and usable by future generations. Analyzing global change is a long-term process. The term "data archaeology" is sometimes used to refer to research done using old direct or proxy data that have been unearthed in archives. As distinct from data archaeology, which refers to numer- ical data recorded by humans, the natural environment may retain information (often called "proxy data") about global processes, including climate. Such information may be recorded in the geological record (glacier ice, for example). The natural record may be analyzed to infer and measure characteristics of past climate and to decipher the importance of decadal-scale present global change. Aspects of this activity affecting development of a global change data management system include: ⢠Preservation of large volumes of material (e.g., sediment cores and ice cores). ⢠Development of algorithms that construct measured, ordered sequences of physical and chemical variables (isotopes, dust content, for example) to a time series of environmental parameters (surface temperature). ⢠Development of criteria for subjecting samples to destructive analysis.
INFORMATION FOR GLOBAL CHANGE 15 The global change data and information management system will be created with limited resources. The high levels of quality assurance, documentation, and long-term archiving described herein cannot practically apply to all research data now collected. Priorities must be set. Research data in the system must be reviewed. This review will have to be done by central coordinating committees and data centers working closely with the global change research community. Functions of a Data and Information Management System The prime function of any data and information management system is the stewardship of the data and information, with all its ramifications. The data and information needed for global change research are costly to acquire. Their safekeeping must not be left to chance. Data system functions may vary depending on programmatic goals, attributes of the thematic data or programmatic issues, and researcher needs. Most components of a data management system will have these elements. The functions of the system include, but may not be limited to, the following: ⢠Providing a programmatic focus for data management. A data management system component should focus the flow of information necessary to conduct global change research. ⢠Data identification and acquisition. The system should take an active role with scientists in identifying datasets useful for global change research. ⢠Standardization of procedures. Standards for quality assurance, documentation, and distribution should be similar among system components.
16 INFORMATION FOR GLOBAL CHANGE ⢠Data quality assurance. High standards of data quality assurance must be employed to maximize the application of data to answer global change questions. ⢠Data preservation. Long-term stewardship of research data must be assured. ⢠Data documentation. Datasets must be completely documented (the documen- tation is often termed "metadata") to ensure their utility for present and future researchers. ⢠Selective data retrieval. It must be possible to retrieve selectively data relevant to a user's needs. ⢠Data distribution. Data must be easily accessible by the world research com- munity with as few restrictions (including cost) as possible. ⢠Derived data products. The system must be able to integrate data within and across disciplines to create derived data products for use by the research community and policymakers.
