************************************************** * Minutes of the EW task force Meeting June 21st ************************************************** 1. An organizational point: Monica has moved the distribution list fron VM to UNIX removing all VM addresses. She was notified with almost one week of delay that the message with the minutes of the previos meeting had not been delivered to Bolek, Ioana, Brigitte, Pere and Edwige. Some addresses were modified in the meanwhile. However regularly take a look at the EW page on WWW, where the full information is always kept (minutes, dates of coming meetings plus other useful pieces of information) in case such problems occurr again. 2. Luminosities 95 Preliminary values for the luminosities were provided by Brigitte: P-2 121 runs 8097.0 +- 16.65 nb-1 236606 uncorrected events P 86 runs 4849.8 +- 13.2 nb-1 136014 uncorrected events P+2 141 runs 9345.3 +- 18.6 nb-1 251740 uncorrected events These errors are statistical only. Preliminary systematic errors are being evaluated. The lumi/run and lumi/bunch/run will be sent to scanbook soon. The correlations with previous years are not yet estimated. 3. Andrea has generated 1million events with KORL07 to study the effect of multiphotons in the evaluation of the acceptance and to understand the discrepancy between two different MC productions (runs 944 and 1757) leading to two different acceptances (see minutes of may 21st). 4. Pere showed a plot giving the (small) effect of the acollinearity cut on s'/s. This plot was obtained from a production of s-channel only events generated at the Peak. Henri remarked however that the effect should have been checked off_peak, where the asymmetry is much larger. Henri has estimated this same effect in an approximate way using Zfitter. Zfitter however uses costheta instead of costheta*. It should therefore be checked the the difference when using costheta or costheta* is smaller than the correction calculated by Henri. The decision on whether updating all 93-94-95 asymmetries or only adding the 95 bhabha asymmetries will be taken at the EW meeting during the Aleph week, after discussing the Bhabha results. 5. Alain informed us on the situation concerning the LEP energies for 93 and 95. The details are given in a text appended to these minutes. As a result of this discussion, we decided that at the LEP energy meeting (which took place a few hours after our meeting) we would incourage the LEP energy group to a. release FINAL 93 energies and preliminary 95 b. if they do not really want to release final 93 energies, at least give preliminary 93 and 95 Thanks to Eric the LEP energy spread and its error should be ready soon, reducing the error on the width from 1 to 0.2 MeV. Today Alain informed us on the conclusions of the LEP energy meeting, which are given below: ************************************************************************* Pending unforeseen difficulties, there will be energy files for both 1993 and 1995. The error for 1993 will be somewhat larger than that for 1995, because of the lack of certainty that the conditions were the same. The magnetic model of 1995 will be used, but the error increased to 2-3 MeV , typically 50% of the effect, this being more or less what one gets if one estimates the error on the basis of only one experiment. Both files are labelled PRELIMINARY. The RF for 1995 will use results of the phasing measurements that take place these days. Date for energy files are end of this week, beginning of next week. Concerning the other experiments, L3 will be ready to use the energies, ALEPH also-- These were the 2 experiments who REQUESTED energies. DELPHI wil use them if available -- and would like some write-up. OPAL seems to have difficulties to be ready. *************************************************************************** +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Message from Alain. to: ALEPH members interested in LEP energies. ----------------------------------------- This is the situation concerning the LEP energies for 1993 and 1995. 1. for 1993 the only issue concerns a reevaluation of the magnetic model to take into account the rise during fills observed in 1995 with the NMR's and with BOF-EOF or long term experiments. The strategy consists in building a model of the rise from the part of 1995 where conditions were similar to 1993 (ie no bending modulation). The rise can be cross-calibrated with a relative precision of about 20% by comparison with polarization in 6 EOF-BOF or long term experiments. It amounts to a 5 Mev shift on the center-of-mass energy between the luminosity-weighted average energy and that at the time of the polarization calibrations. An additional uncertainty should be assigned to the effect in 1993 because one is not certain that all conditions were identical. This fear is based on the fact that only part of the rise can be associated with trains, the remaining part being more mysterious. Even if the source were the same, one could argue that ground conditions might have changed from year to year, so that the effect on LEP magnets could be different. It is my opinion that, due to the lack of instrumentation, it is highly unlikely that future experiments will shed light on the past at that level. But how can one set a limit on that? (besides checking Mz_1993 vs Mz_1995) One possibility is to estimate the 1993 rise from the 1993 long term experiments. There are three long term experiments but two of them are prolongations of end-of-fill MDs, and do not really apply to the first 10 hours of fills, where most data are recorded. This leaves us with essentially one experiment, the temperature experiment, which indeed provided first evidence for unexpected jumps. With only one experiment the error is given by the non-reproducibility of the rise from one fill to another. This can be extracted from the non reproducibility of 1995 fills which is well measured. While on average the luminosity weighted average energy is lower than the energy at the time of polarization calibrations by 2.5 Mev/beam, this number fluctuates by 1.2 MeV. This would lead to an additional uncorrelated error for 1993 of the order of 2.4 MeV center-of-mass. e.g. on the Z mass. (the corresponding 1995 error being 1 MeV on Mz) A useful test of the rise in 1993 was suggested by G. Quast: one could compare the Z mass measured during the first 3 hours of each fill to that measured using all other data. Although the precision of this kind of test is limited to about 3 MeV, it certainly should be done. An important question concerns our certainty that the Z width is not affected. This can be tested for 1995 by comparing the NMR rise at peak-2 and peak+2. The rise is consistent with being identical with a precision of the order of 0.5 MeV. Another interesting test is the fact that the energy difference between the energy points is the same if one uses the NMRs to measure it or if one uses the polarization calibration, within a precision of 1MeV. This suggests that the additional uncertainty on the Z width from the magnetic rise should be of order 0.5 MeV. ALTHOUGH these error estimates and the precise energy files are presently being refined, I think that -- we have now all elements to finalize the 1993 energies. -- I dont see how the 1996 data can help us understand what might have changed between 1993 and 1995. Therefore I would strongly support issuing FINAL 1993 energies as soon as the final cross-checks are completed, without further delay. (as always one cannot be too dogmatic about this.) G. Quast's test can be performed on the existing energy files, (Arnaud can you do it?) Incidently, it would be useful to perform it also in 1995. 2. 1995 energies. the situation for 1995 is simpler for what concerns the magnetic model, and this work is essentially completed, with a precision better than for 1993 (more data). There are however three new aspects in 1995: -- RF corrections. The RF configuration has varied widely in 1995 and the logging of changes is somewhat patchy. Mike Hildreth and others have done a tremendous amount of work trying to understand the situation. ALEPH vertex data have proven to be very useful. The large discrepancy shown in the ALEPH WW meeting originated in miscalibrated OPAL data, as comparison with ALEPH data revealed. The situation is much better now. Nevertheless, the final file for RF corrections is far from ready, beam loading issues from bunch to bunch remain somewhat obscure. It is hoped that a solution will be found before the end of June. Because we alternated carefully between -2 and +2, it is expected that no serious implication for the width will occur, but this remains to be demonstrated. -- dispersion effects. Here the situation is much better, John Yamartino and Patrick Puzo are in the process of finalizing the energy corrections, and most of all the measurements of OSVD. A slight complication arises because of correlation in Vernier positions between experiments. Altogether, however, the remaining corrections and errors are of the order of 0.3 MeV or less. -- correctors. Contrary to 1993 where care was taken to load reproducibly the horizontal orbit, there are some diffences in the integral of horizontal magnetic field between normal data taking and polarization data, of the order of 1MeV at one energy 0 at the other. Jorg Wenninger is investigating what the possible effect could be. Expectation is: somewhere between 0 and 1 MeV. Unfortunately this is different at peak -2 and peak +2. Given the state of affairs, it seems absolutely impossible to have FINAL 1995 energies for Warsaw. I still think one should aim at PRELIMINARY energies but the possibility that the RF corrections wont be ready cannot be ignored. In order to avoid that everybody waits for everybody else, this goal should be maintained as long as possible. I should add that a final value of the LEP energy spread and its error should be ready within days, thanks to Eric. The error is reduced down to 1MeV on Sigma_E, giving about 0.2 MeV on the width. For 1993/1994, this should be final. For 1995 the final values of OSVD's are needed but I dont expect this to be a problem. We all know that deadlines of conferences are artificial, but they constitute a good excuse to converge. The Warsaw deadline is as good as any other.