Summary of CLAS Collaboration Meeting - June 3-5, 1999 S. Dytman The new CLAS Chairman will be Reinhard Schumacher starting Sept. 1. All the best to him. I will spend a lot of time with him this summer to assure an orderly transition. Next collaboration meeting will be in the week between the Santorini conference in Greece and the DNP meeting in Asilomar. It will be Oct. 14-16, 1999. Save the dates. Reinhard will be the organizer. This meeting was a somewhat bifurcated meeting. We had sessions on the ongoing task of taking quality data and the increasingly important task of aiming for our first paper. I will push hard to prove the quality of our data and improve the analysis where necessary. I know that many in the collaboration are already doing this, but I feel we need a sharper focus and better organization of this effort. I will send a separate summary of this situation (task force report) after I have consulted with a few people. The first session was devoted to reports of the e1 and e2 runs, a few important technical reports, and a series of talks discussing the status of tracking reconstruction. It appears that both recent runs (including the FIRST time CLAS has run with a A>1 target with the primary goal of a physics measurement) were successful. Volker Burkert (e1 ECO) reported major success for the e1 run. Mark Ito was COP. I remember 1.7 billion triggers at 1.5 and 2.5 GeV along with 2 billion at a few energies slightly larger than 4 GeV. This quadruples our data at each energy, the primary goal of the run. The primary problem was a new lower mass LH2 target that didn't fill correctly. A threshold of 260 MeV in the EC was used, giving an upper limit of 3 GeV in W. This added to the inefficiency of the trigger because many triggers came from radiative tail electrons. However, Volker says the real limit was the apparent tracking efficiency measured by a drop in the overall ep rate of 10% at 7 nA. I believe they ran at 6 nA or 7 x 10^33 luminosity. The dead time was about 10%. The DAQ system could have run with larger rate and a larger deadtime. Stepan Stepanyan was ECO and gave the e2 report. Mac Mestayer was COP. They ran with 3He, 4He, 12C and 56Fe targets for a total of 2.3 billion triggers (8.33 Terabytes of data). They had an overall data taking efficiency of 55% for 520 hours DAQ time and considered the run a big success. They ran with luminosity of 0.6-0.9 x 10^33 per atom (larger than 10^34 per nucleon). They ran with a much higher EC threshold than e1, 1.5 GeV at 4 GeV beam energy. Elton Smith reported on progress of the Quaconn (Quality Constants) Comm. This is a self-declared group of mostly Hall B staff. They have taken on the critical job of collecting calibration procedures, map storage procedures, and documentation for the collaboration. To test robustness of present methods, new students were charged with the task of calibrating the e2 data. Excellent success was reported. They define their task to not include general recruitment of manpower, but they will hopefully define the necessary jobs that are required to do the critical tasks. Gerry Gilfoyle gave a brief report on work he and Arne Freyberger did. Impressive time histories of CLAS data parameters are available on the web. If people were checking these more diligently, he showed a few places where failures would have been found more quickly. He didn't give the web address, here it is: http://claspc10.cebaf.gov/TIMELINE/ Latifa, Mac, and I organized a session on tracking and related topics. Franz Klein and Jim Mueller spoke on the present status of resolutions and efficiency. Franz showed that we have about 50% larger resolution than design with the unknown component probably due to geometry/B field errors. Better resolution (about 40%) is possible with 6 superlayers required, but at a large cost in efficiency. Jim showed new results for tracking efficiency that are much better than those in recent CLAS Analysis notes. He showed that careful determination of the denominator is important to get an objective number. For example, he agrees with the Saclay estimate of 15% apparent inefficiency for protons in gamma p->pi+ pi- p events, but ONE-THIRD of that value are really events that were not identified correctly so that the real inefficiency is 10%. Hit-based inefficiency is about 0.5% and time-based inefficiency is about 5% for a 4-5 particle final state and the remaining few percent is due to inefficiencies in other counters (TOF scin). I am personally very impressed with these results. Although they show we have reasonable values and understanding, further improvements will surely occur! Cole Smith has carefully studied ep elastic cross sections including radiative corrections. He sees agreement with the world's data at about the 5% level across a broad range of angles. This has very positive and somewhat negative aspects, but gives a solid picture of where we are. Mac Mestayer gave a summary of where we are in understanding the geometry. Unfortunately, the understanding of the magnetic field is intertwined because we don't have a useful field map. Geometry map values are presently fixed at those determined by Rob Feurbach with the assumption that region 1 can't change location. These values are INTERNALLY CONSISTENT, but do not put the elastic peak at the right energy. Recent work by Konstantin Mikhailov goes further, but additional assumptions are required. I feel this work is in the right direction, but needs more people and ideas. I will be very brief on the PWG progress, even though we spent an entire day presenting and discussing results. Electron beam data for many reactions now have cross sections with all the major corrections. Work has advanced to the stage of testing systematic effects. Analyses of photon beam data are now producing cross sections for the first time. The statistical accuracy and kinematic range compared to previous data are very impressive. The second major technical subject was acceptance. This session was organized by Will Brooks who is heading the acceptance effort. Steve Barrow showed results for a comparison of a strict application of single particle acceptance and a full GSIM acceptance. The former acceptance calculation applied the standard Volker Burkert fiducial cuts (no event reconstruction) while the latter did a full reconstruction. The acceptance at small relative angles was much smaller in the GSIM acceptance. This is due to tracking inefficiencies as discussed above. A major component of the time-based tracking inefficiency is the difficulty in disentangling 2 tracks that are close to each other. It should be noted that even this acceptance calculation is incomplete because the noise environment was not simulated. Burin Asavabhipop showed that binning choice matters. My understanding is that one must choose bin size to be larger than the width of the resolution function in that variable. Otherwise, there is an ambiguity in which bin to place the event in the acceptance calc. Will summarized the progress of a group putting together a CLAS acceptance document. They outline what appears to me to be a massive tome covering ALL aspects of the problem. He also made a call for PWGs to survey their groups for anticipated GSIM needs. I was hoping for some assessment on quality of the code, in particular whether it is good enough for publication quality acceptance. Saturday morning was filled with reports of various committees and discussion of new charter by-laws. First, Bernhard reported his view of our status. He discussed the 1999 and 2000 schedules which are on the web. He showed an interesting table of existing analyses. (We need to augment this list!) He also noted that CLAS improvements are largely tied to 8 GeV (and higher) upgrades. Work continues on calorimeter and Cerenkov improvements, but there is no work presently in progress or anticipated in tracking upgrades. Steve Dytman reported on the Coordinating Committee work. They meet every 2-3 weeks to discuss issues such as schedule and PWG policies (e.g. standards for and posting of CLAS results and submissions to PANIC). Recent work has been on organization of a Mini-symposium emphasizing CLAS results at the Asilomar fall DNP meeting. Time is tight, we want to have roughly 10 talks on CLAS work. Your best strategy is to submit abstracts to your PWG asap so that we can try to include your talk in that session. Questions about this should go to me, Latifa, Ritchie, or Briscoe; the situation is still fluid. Another major recent topic has been trying to find a way to encourage progress toward first CLAS papers. I have tried to push this at SON working group meeting in May with uncertain results. This is not a simple situation, people seem threatened by certain aspects. My only goal has been to find a way to choose the efforts that are required and organize them in a way that is politically uncontroversial. The most positive way to frame the issue is apparently in terms of the issues common to many analyses. Later in the session, I proposed to organize task forces around 4 issues that many people have talked about and the reaction was very positive. The issues are acceptance, geometry/B field, tracking efficiency, and photon normalization. I will discuss these in the task force report. Mac Mestayer reported on work of the Service Work Committee. They recently sent out evaluations of the service work proposed by each group. Almost all were rated 'average' and requests will be made for more documentation and some shifts in effort. At this time, Mac said the sum of the proposed work is equal to what is required to keep CLAS going is "about right". I thought that in the past and in the responses I read that they emphasize the (sometimes) unfilled jobs of detector pager carrier, calibrators, and data cookers. Mac sought a further charge from the collaboration to define the new jobs and to start to seek people to commit to them. The collaboration in attendance expressed dissatisfaction with the lack of progress of this committee and fully supported the need to move ahead quickly. Jim O'Brien (CLAS Archivist) felt the extension of the charge they sought was already included in the original language. Mac outlined his view of the CLAS work in terms of the work before data taking, during calibration and cooking, and to get final physics results and the roles of the various entities - JLab staff, run groups, and defined collaboration wide jobs. He proposed the addition of 2 new collaboration jobs- Software Coordinator and Analysis Coordinator. The SC would do much of the work already in the realm of the existing SC, Mark Ito, but would give that job wider backing. The AC would give overall guidance to aspects of analysis such as setting standards for publication, sponsoring comparisons of competing analysis methods, and others I forget. Although there were many words spoken against the proposals (e.g. reinforcing the strong role of run groups in calibrating and cooking), I feel he had some good ideas in there. Right now, no one is doing many of the tasks he mentioned and all are required. I also discuss aspects of this in the task force report. Rory Miskimen (chair of the CLAS Speakers Comm) discussed their recent work. They reviewed 9 proposals for PANIC submissions, approved 7 of them and merged 2. There were various problems with a few of the submissions that can be attributed to varying extent to all involved. We must all work harder to improve the evaluation. Deadlines must be respected and all abstracts will require more documentation from the submitter and from the PWG reviewers. For PANIC, all 7 talks submitted to the organizers will be given; this is a major accomplishment! After discussion last time, the collaboration supported the need for a way to approve analyses with sufficient rigor to be useful on a resume even though PAC never sees it. I proposed a mechanism that is very similar to the way most PWG's evaluate proposals. The people at the meeting had a few clarifying suggestions, but approved the basic concept by a wide margin. I will send some specific words in a few weeks; in the meantime, we will prepare for submissions for CAA (CLAS Approved Analyses) at the next collaboration meeting. (Suggestions for a neater name will be appreciated.) We approved 2 new full members, Alexei Stavinsky and Konstantin Mikhailov (both ITEP). Congratulations to both! Gail Dodge has been Chairman of this committee for a few years and has done a marvelous job. However, she will leave the job in one year and we are looking for a successor. Gail raised some membership issues. There will be proposals next meeting regarding limited and term members. Talk to Paul Stoler if you want input into the former, Rory Miskimen is working on the latter. Finally, I want to express the pleasure I have in thinking about the large strides we have taken as a collaboration in the last 2 years. We have been coping with the difficult task of finding the right ways to advance toward the (still) expected flood of exciting physics results while balancing group coherence and individual initiative. It has never been easy and sometimes frustrating as I learned what could and could not be done. One of my comforts has been in John Domingo's analogy to herding cats. Still, we are clearly getting there and the future looks good.