---------------------------------------------------------- Reports of Boundary Layer Campaign Working Groups Meetings Snowmass, Colorado, June 28-30, 1992 ---------------------------------------------------------- Report of GEM Working Group 4: Data Task Force John Olson and Ted Rosenberg (co-chairs) During the Snowmass meeting, three sessions of the WG4:DTF were held with descriptions of the agenda items given in the program distributed at the meeting and appended to this report. This report is a summary of the discussion and recommendations which resulted from the meetings and was presented in the plenary session by J. Olson. The overheads used in the presentation are also appended to this report. I. Introduction It is recognized that the formation of a data and model base is essential to the GEM program. In our discussions the consensus was that the cost of the data base should be borne as a necessary part of the infrastructure of the program. In order to emphasize the dual nature of the data base it was decided to call it the GEM Data/Model base (DMB). The issues which resulted from the sessions arranged themselves into two categories: those associated with the nature of the data/model base (e.g. formats, location, rules of access etc.), and those associated with the management of the data base including a plan and time-line for achieving the DMB. The discussion below is divided between these two categories. II. Issues concerning the nature of the DMB The primary issue under discussion was the nature of the GEM/DMB vis-a-vis other data and model archives. It is recognized that it would not be cost effective for the GEM community to try to duplicate the information currently assembled in such archives as those operated by NASA/NSSDC, NSF/NCAR (which includes the closely related CEDAR data base) and the Air Force modelling program. Also, there are other data and model archives which have come into being as certain members of the community have carried out their research. It is recommended that the GEM/DMB develop appropriate interfaces to these data and model bases in order to facilitate the flow of information to the GEM community. In this way the GEM/DMB becomes something of a "nodal" data base, residing in several locations. The final development of the DMB will require those defining its form to investigate the trade-offs between assembling some independent data and model sets which are unique to GEM and simply adjoining the GEM data to an existing data base such as that operated by CEDAR. Also, the close association of GEM with programs such as STEP which offer contact with the international community should be encouraged. There follows a series of issues concerning the nature of the DMB. One such issue is that of longevity. Should the DMB be established so as to match the lifetime of the GEM program, growing or diminishing as the program does, or should it look to supporting observations over longer periods, e.g. a solar cycle? Finally, the contents of the DMB should be defined in light of the desire to minimize the overlap with other archives while still offering relatively quick access to important modules. Some of the suggested items to be included in the DMB were: * a list of instruments, sites and data coverage * campaign data and notices * indices * field models and model input parameters * standard coordinate systems (e.g. common MLT definition) * tested, non-research modelling codes The establishment of a DMB would require the development of a catalog describing the DMB, its contents and how it may be accessed. The community must also develop "rules of the road" which establishes the protocol for accessing and using the data to insure the originators of the information are properly credited or offered the opportunity to participate in the research using the data. III. Issues concerning the management of the DMB It was recognized that GEM researchers are already collecting data, developing models and organizing campaigns. Thus there is an immediate need for a clearing-house of GEM information to be established to serve the needs of the community until the final DMB is organized. In this light, the DTF suggested a short-term and a long-term strategy to develop the DMB. III.A. The short-term: a clearing house for information. The GEM Messenger was established at UCLA as a newsletter for the GEM community. It was felt that an efficient way to satisfy the need for information exchange in the near term would be to create a bulletin board system on the same computer as used to distribute the GEM Messenger. Dr. Russell has kindly agreed to do that and Dr. Olson of this committee will work with him in this development. Once in place, it is expected this will fill the need for the basic communication needed by GEM researchers. III.B. The long-term: The role of the Data Task Force It was agreed that the entire GEM community should have the opportunity to voice their preferences for the DMB. To this end, the DTF will produce a survey of the community which can be used to define the DMB and its operation. Once completed and analyzed, the survey results should be passed to a committee which is appointed to implement the DMB. This committee would be charged with a) creating the proposal to NSF, b) implementation of the DMB and c) oversight of the DMB. At this point, the responsibilities of WG4 would be completed and it may be the easiest to simply allow WG4 to become the oversight committee. Dr. T. Eastman, Program Director at NSF, has indicated that a proposal for the DMB should be received at NSF before February 1993. With this deadline, the following timetable is proposed: * July 1, 1992 Draft survey distributed to WG4/DTF * July 15 Survey finalized and distributed * Sept. 1 Survey closes, review begins * Oct. 1 Final DMB guidelines adopted * Oct. 1 DMB committee appointed * Feb. 1, 1993 NSF proposal submitted by DMB committee. -------------------------------------------------------------------------- Appendix: Plenary Session Report: -------------------------------------------------------------------------- Overhead 1 DATA/MODEL BASE: ISSUES ----------------------- + Cost: 1+ salary + hardware + Nature: nodal establishing interfaces with - NSSDC - NGDC - CEDAR + Lifetime/Archive function + Location: Nodal (as above) + Data exchange format + "Rules of the Road"-acknowledgments + Contents of the data/model base - observation sites/ instrument lists/ data coverage - campaign notices and data - indices - field models, model inputs - standard coordinate systems (MLT) - tested, (non research) models - source codes --------------------------------------------------------------------- Overhead 2. DATA/MODEL BASE ISSUES: MANAGEMENT ---------------------------------- + Short-term vs Long-term management needs - Short term: coordination/clearing house function - use UCLA GEM Messenger format for bulletin board/distribution - post general messages, campaign dates, issues too long for GEM Messenger - Long term: The Data Task Force - A survey of the community - Formulation of a D/M structure - Formation of a DMB oversight committee - proposal preparation - oversight of the DMB - development of catalog - etc + Time scale - recommendations should be forwarded to NSF in proposal form by February 1993.