CHAMMP NEWSLETTER, VOL. 3 NUMBER 1 May 1993 Dave Bader, (301)903-4328 dc_bader@pnl.gov (My preference for all who have INTERNET) or D.BADER on OMNET SEMNTER-CHERVIN OCEAN MODELING RECOGNIZED BY "SCIENCE' AND SMITHSONIAN The Semtner Chervin ocean modeling work was highlighted recently in an article by Richard Kerr of "Science" magazine (Vol 260, pg 32, 2Apr 93). A side-box on the development of the parallel version of the code contains a paragraph on the LANL CM-5 implementation. Semtner and Chervin will be honored at the Computerworld Smithsonian Awards dinner on June 7, where they will receive the 1993 award for Breakthrough Compuational Science. This award is given annually for the best use of compuational science to:"significantly increase the possibilities for improvement in the human condition; solve or make notable progress on a previously intractable problem; set new, replicable standards for scientific endeavor; and create new technological tools with which to effect change." With this award comes a permanent display in the Smithsonian's National Museum of American History. CHAMMP ACCESS TO THE HIGH PERFORMANCE COMPUTING RESEARCH CENTERS Now that the Intel Paragon is running at the ORNL Center, access for CHAMMP researchers is available on both the Paragon and the CM-5 at LANL. Temporary policies for gaining access to these centers is as follows: 1. A complete, but concise statement of resource requirements for the next 3-6 months must be submitted to both CHAMMP and the HPCRC's. The statement should include enough technical information to judge the applicants capability to make use of the machine. Early users should be aware that consulting is limited and that problems will be encountered at the centers while they stabilize. Those interested in using the LANL CM-5 should obtain a copy of their electronic submission form. E-mail a request to scp@acl.lanl.gov to obtain a copy. 2. Proposals will be reviewed for relevance to CHAMMP goals and technical merit. Priority is given to CHAMMP Science Team PI's and Model Development Team participants who have massively parallel code development projects as part of their CHAMMP-funded activity. Proposals which are not directly relevant to CHAMMP research priorities will be rejected. THIS IS NOT A GENERAL ACCESS PROGRAM. CHAMMP FUTURE DIRECTION In the future, CHAMMP will sharpen its focus on the compuational and theoretical aspects of decade to century climate prediction. CHAMMP's contributions in developing a computational science component for climate modeling are now being recognized. The broad scope of climate modeling, however, requires that the program focus on those issues that are required to predict climate on the scales of interest to DOE to guide energy policy. It is hoped that other agencies will develop a stronger compuational component within their modeling programs so that a larger compuational science community actively involved in climate prediction develops. The first step in achieving this tighter focus was a short review held by JASON, CHAMMP's technical oversight committee, on Oct 26-27, 1992 in San Diego. The primary emphasis was long term climate prediction with coupled atmosphere -ocean and ocean GCMs. Although the JASON written report has not been finalized, preliminary discussions with several participants at the meeting were encouraging. PROPOSAL SOLICITATION A CHAMMP proposal solicitation will be announced in the "Federal Register" later this year. The review of these proposals will result in the selection of Science Team Projects commencing in the fall of 1994. PI's who have grants expiring in 1994 must recompete to continue their CHAMMP involvement. More information will follow later. PROJECT FUNDING All Non-DOE laboratory PI's should be aware that their annual reports and requests for next year's funding are past due, if not already submitted. Contact your institution's contracts office if you have any questions about procedure. CHAMMP SCIENCE TEAM MEETING The annual CHAMMP Science Team meeting was held in Monterey, CA, March 15-17, 1993. Despite the delays experienced by the East Coast crowd in getting to the meeting because of the "Storm of the Century" (Climate Change??), the meeting was very successful. Mike MacCracken is preparing electronic and paper summaries of the meeting. MODEL DEVELOPMENT EFFORT The model development effort should get into full swing with the availability of the HPCRC massively-parallel machines. The effort will concentrate on model validation and performance characterization. Metrics for the evaluation of these properties were the topics of discussion at one Science Team session, but no conclusions were reached. An ongoing discussion among participants will lead to more definitive measurement criteria and a first cut at quantitative and qualitative evaluation. MEETINGS Workshop announcements are periodically distributed via the same e-mail list used to distribute this newsletter. A brief listing follows. SIAM 1993 Annual Meeting, July 16, 1993, Philadelphia, PA Minisymposium: Improved Numerical Methods for General Circulation Modeling Developed under the Department of Energy's CHAMMP Program Four talks have been contributed. Contact Dave Bader. IAMAP Meeting, Yokohama Japan, July 1993 CHAMMP presentation by Mike MacCracken at the Global Climate Modeling Symposium Joint ARM-CHAMMP Workshop on Cloud and radiation Parameterizations Colorado State University, September 7-8, 1993, Ft. Collins, CO. Attendence at workshop is by invitation and participants are expected to perform a set of standard computations and submit results prior to the conference. Contact Dave Randall at CSU (randall@redfish.atmos.colostate.edu) CHAMMP Contacts Dave Bader, CHAMMP Program Director (301)903-4328 dc_bader@pnl.gov Mike MacCracken, CHAMMP Chief Scientist (510)422-1826 mmaccracken@llnl.gov Bob Malone, CHAMMP Director of Model Development (505)667-5925 rcm@lanl.gov