Minutes of the Jefferson Lab CLAS Collaboration May (4 - 6), 2000 Next meeting: September 21 - 23, 2000 at Jefferson Lab, CEBAF Center For you to do: 1) Check out the first Collaboration photograph, found on the Hall-B web page. 2) Save the above dates for the next meeting. 3) Read the next (and very close to final) draft of the phi photo- production paper when it appears for Collaboration review in a week or so. I will notify you when it is on the web-site. ________________________________________________________________________ The weather was bright and beautiful as the CLAS Collaboration met for its spring meeting at CEBAF Center. About 70 Collaboration members signed-in to the meeting over a period of three days. Amazingly, CLAS has now been in full-scale data-taking operation for two and a half years without a major failure or breakdown. At this time, also, the Collaboration is about to submit its first publication, with more in the pipeline, which is a perfectly reasonable time-scale for the production of results from a large new apparatus. Since the January meeting, CLAS has successfully completed the e1(d) run (3.1 billion triggers of electrons on hydrogen at 4.8 and 3.1 GeV, and on deuterium at 2.5 GeV, all of it 70% polarized beam), and is in the middle of a so-far successful e5 run (1 billion triggers of electrons on a dual hydrogen deuterium target). Hall Leader Bernhard Mecking reviewed the state of CLAS and sent electronic copies of his slides to everyone. The role of CLAS in the future 12 GeV era was highlighted, with the main idea being that if the luminosity of the detector can be upgraded into the range of 10^35 /cm^2/sec, then the Q^2 range that can be covered in exclusive processes at high energy can be competitive with the other installations in the world. A fairly large restructuring of the CLAS instrumentation would be necessary. Work on the physics motivation and on the detector aspects of this upgrade are going on this summer, and will be included in the White Paper the Laboratory is assembling. The case for the 12 GeV upgrade will be made to NSAC this fall. (More discussion below.) Chairman Reinhard Schumacher discussed where we are in terms of progress on analysis projects and on manpower. Notes were handed out and the tables are available on the Hall B web site. There are now 74 analysis projects on the books, of which 65 have data to analyze. There are 41 CLAS PhD projects, an increase of 14% since January; 6 of these are finished, with theses available to the Collaboration from the secretary (Linda Ceraul), and some are down-loadable from the CLAS web site. Preliminary results of 20 analysis projects have been shown at national and international conferences. One publication is about to be submitted to PRL, a second is at the draft stage, and a third will be in draft form by mid-June. At present 8 out of 35 member institutions have post-docs on site at the lab; [I think it would help if this number were larger, since a limiting factor in our rate of progress is simply the number of people in close collaboration which we have on-site to bring to bear on our common analysis issues.] The first publication will be on phi photoproduction at high t. A preliminary draft was made available last week on the CLAS web site. Jean-Marc Laget made the case for this paper to the plenary group. The Ad Hoc Committee reviewing this paper (V. Burkert, W. Hersman, G. Adams) gave their summary comments back to Jean-Marc and the other lead authors about one week before the collaboration meeting. The suggestions that were made were accepted as reasonable by the lead authors, and were not thought to require a time-consuming additional round of analysis. A number of procedural questions about our publication process, as well as specific questions about the paper were raised in the plenary discussion, enough that a separate plenary discussion was planned for Friday evening. That meeting attracted 18 people and was very lively. The end result was that every objection or question that was raised was answered to the satisfaction of those present. It led also to a change of wording in the conclusion about how strongly the CLAS high t data rule out certain models of the reaction based on pQCD, [plus several small changes in text or figure presentation]. The final draft of this paper will be on the CLAS-internal web site in one to two weeks, and when it appears all members will be informed that the "official" comment-clock has started. The status of the eta electroproduction analysis for looking at the S11(1535) was presented by Jim Mueller. With 45,000 etas to work with, many analysis details related to acceptance corrections, systematic errors in peak shape, background subtractions, and detector geometry were discussed. There are specific results for the excitation of the S11 as a function of W, and within a model there are results for the contributing multipoles. There was a focus-group meeting about this measurement two weeks ago, and it was felt that this analysis is close to publication. Volker Burkert discussed the other analysis near the publication stage, that of pi0 electroproduction near the Delta(1232), with emphasis on extracting reliable multipole contributions as a function of Q^2. Three different high-level analysis approaches are being applied to the data to look for consistency of the results and for unexpected systematic effects. Systematic effects induced by the choice of event generator (AO-based vs MAID-based) were discussed. It was stated that a PRL draft would be ready by mid-June. Burin Asavapibhop discussed his just-completed thesis measurement of the axial anomaly at the gamma-pi-pi-pi vertex, based on g1a data. He showed that CLAS data confirm the QCD prediction for this quantity, or alternatively that the number of colors is indeed three. This experiments provides by far the best measurement of this anomaly factor that has been made. More systematic checks may be needed before this result is ready for publication. Kim Egiyan gave the last plenary physics talk of the day, on the search for NN correlations in 12C. The e2 run group has plenty of data to work with. He showed preliminary results for the yield of events where the angle between the q vector and the knockout proton direction is very large, which gives sensitivity to either direct knockout of high momentum nucleons, and/or protons from 2 step processes such as Delta exchange or proton rescattering. Kinematic cuts may separate the contributions. In the middle of the afternoon the Collaboration took a break to participate in the annual "Jefferson Lab Run-Around". In this ~1.3 mile foot race around the site, no fewer than seven CLAS people finished with times close to or under 10 minutes, and at least as many others also finished the race in good form. Afterwards the Lab hosted a picnic behind the Residence Facility which was very well attended. The diligent CLAS physicists tore themselves away, however, and reconvened their meeting after this break. A long series of short talks itemized the progress which the 10 active run groups have made recently in the calibration and cooking of their data. Since the last collaboration meeting the e1b/c cooking has been (almost) completed, and the e2 cooking has been completed. Simply averaging over calendar time, both of these run groups averaged 20 million triggers analyzed per day with a version of the analysis code called [I believe] PROD1.9. This number spans the period of time when the access to the raw data files on the tape silo was the main bottle neck in the cooking procedure. Since then some improvements have been made by the Computer Center in the handling of tape requests, another set of tape drives and server have been added, and the Collaboration has become more economical in it's use of the tape queues. In addition, Eugene Pasyuk and Valeri Koubarovski, with help from Jim Mueller, Franz Klein, and Mac Mestayer have benchmarked and tested a series of modifications to the analysis code to speed it up. The mods changed the cuts and iterations made in the track finding and fitting, but not in the algorithm itself. The present version, called PROD-2-0-G2A was developed which runs about twice as fast as the previous cooking code when run on photon data. A detailed CLAS-Note is in preparation. Using this new code, the g2 cooking has recently started, and the g6(b) cooking has also started. Dieter Cords presented results of work on the on-line system. Reliability, robustness and ease-of-use have continued to improve. The present structure of the on-line system was reviewed. It was estimated (by B. Mecking) that the hall now runs with 85% availability, with only about 5% of the down-time due to the on-line system. Mark Ito then discussed the situation in the offline analysis. Farm utilization is now above 90%, up from about 50% at the last meeting. Present utilization is hard to beat, since the remaining 10% comes from dead computers and "end effects". There have been improvements in many areas: better disk usage, a second tape server, more clever staging of the data to be cooked. The analysis code improvements were discussed. More CPU's for the farm are on order: 50 more dual processors in addition to the present 75; the actual throughput should double. He reported on the recent software meeting at which the generation of new cooking prods/releases was discussed, as well as the development of a full-blown database to manage the calibration parameters for CLAS. Will Brooks and Kyungseon Joo reported progress with GSIM and acceptance, but the discussion was too detailed to summarize here. Will estimated that to manage CLAS's simulation needs we must have 1000 Linux-equivalent CPU's at 100% availability, while at the present we have about 350. Those we already have are sitting at about 25% usage due to lack of demand. He predicts a CPU crunch as more analysis projects get closer to producing results. Mac Mestayer showed a result from the B-field optimization work done by Russian colleagues. There are two versions of the CLAS B-field: the one everyone has cooked with so far, and the other generated last winter after it was discovered that the drawing for the cryostat had been mis-read by the people coding up the B-field for the analysis software. By looking at elastic electron scattering it a certain way it was found that the reconstructed beam energy was most accurate if the B-field used was an interpolation of the two versions of the field. Work with other data sets is necessary to corroborate this result. The newly started cooking epochs, for g2 and g6(b), use the good old magnetic field map. On Saturday the Chairpersons of the Physics Working Groups gave summaries of the PWG meetings; we will not summarize the summaries here. The proposal by Dan Carman et al to study hyperon double polarization observables will be submitted for PAC 18 by the deadline on June 1. This proposal was previously approved by the SoN group, but now Dan had some nice results from preliminary analysis which shows that the double polarization observables PL' and PT' are everywhere large and measurable. In addition, Volker Burkert gave an overview of all the spin-related results which are under analysis now. Especially interesting was a first-ever observation of beam asymmetry seen in single pion electroproduction; the effects are large and show striking changes with W throughout the resonance region. Various observables in double spin asymmetries from the analysis of Raffaella deVita were shown, and they also were seen to exhibit large values; it was felt that this analysis is rapidly maturing and may be ready for publication soonest within that set of measurements. The NSAC Long Range Planning exercise will commence this fall, and in preparation Jefferson Lab is writing a White Paper to bring together the physics case for a 12 GeV upgrade of the CEBAF beam. The role of CLAS in the 12 GeV era has been under discussion among various members of the Collaboration who have contributed to the Lab's workshops devoted to this topic. Paul Stoler reported where we stand today. A strong case can be made for CLAS in the area of exclusive reactions. There are two working groups which are developing this case. First, the area of high Q2, high W, low t electroproduction of pseudovector or pseudoscalar mesons one can access skewed parton distributions (SPD's) which are a measure of the transverse components of the quark wavefunction inside the nucleon. Michel Guidal is organizing this working group. Paul explained what he believed to be the physics interest of such measurements. The big question is at this time: how does one connect the measured cross sections to the SPD observables which can be calculated from theory? Second, in the area of high Q2, but low W and high t, one can look at baryon form factors under those conditions. Ron Gilman is organizing that working group. Common to these measurements is the need to achieve higher (no one is sure how much higher) Q2 values to enter the region where scaling and asymptotic effects occur. The likely changes to the CLAS spectrometer were touched upon by Bernhard Mecking on the first day of the meeting. They key feature is the following. To get to higher Q2, one considers that the event rates fall off as Q6. Thus one needs higher luminosity in order to complete experiments in a reasonable amount of time. To double the range in Q2 requires, then, a factor of 8 or so in luminosity. A starting point for detector upgrade is to see what is needed to raise the luminosity of CLAS by a factor of 10 without losing the (semi-) exclusive capability of the detector. Faster chambers with smaller drift cells are likely to be needed. How such a development can be carried out, i.e. who would do it, and how much it would cost, and on what time-scale, will not be worked out in detail for this White Paper. Personnel changes: ================= Gail Dodge of the Membership Committee and Brian Raue of the Speakers' Committee reported on a number of shifts in who is doing what within the Collaboration. New Members (voted upon by the full Collaboration at this meeting). We welcome each of these new collaborators: Serguei Pozdniakov, ITEP, working with the on-line group Jerry Feldman, GWU, working on g8, linearly polarized photons Kenneth Livingston, Glasgow, working on rho and omega photoproduction, g8, lin. pol. photons James Kellie, Glasgow, working on kaon, rho, and omega photoprod., g8, lin. pol. photons. In addition, Gail Dodge is stepping down as head of the Membership Committee. That committee selected Kevin Giovanetti as the next head. >From now on, the requests for membership information, the lists, and so forth, will be handled by Kevin. [We thank Gail for her effective service in this post for the last several years.] The members of the Speakers' Committee (CSC) serve for two year terms, with staggering between the two members from each Physics Working Group. It was necessary, therefore, to select one new representative from each PWG at this time, such that the new CSC can convene at the beginning of September. Real Photon: Marco Battalieri replaces Rory Miskimen SoN: Brian Raue will be replaced by one of two candidates to be elected by email in the next few weeks Multihadron: Derek Branford replaces Bill Hersman Riad Suleiman replaces Maurik Holtrop as an alternate There is the important but unofficial post of Shift Scheduler, a job which has been done by Arne Freyberger for several years. He is stepping out of that job, and we are very happy that Mahbub Khandaker has agreed to pick up this task. It is expected that he will be running the shift-assignment procedure in much the same way that Arne has, starting in June, for the data-taking period of September through December. [We are very grateful to Arne for doing this task so well for the last few years, and wish Mahbub lots of success in juggling the shift schedule from now on.] Any changes, additions or other clarifications of these minutes by the Collaboration members will be cheerfully considered. Amended minutes will be sent out only in case of egregious errors or omissions. [...]=chairman's editorials