- Measurement of Rb above the Z peak ---------------------------------- Mario has looked at the HE data at 161 and 172 GeV (not yet at 183, mainly because of the known bug in the database), and at the Z data (1.2 pb-1) taken in 1996 (for comparison and systematics estimation). Since the statistics are small and background fluctuations can be high, one should take care of the purity. So, Mario first has shown a comparison of b-tag efficiencies and purities at 161 and 172 GeV for a double hemisphere tag and an event tag. Result: With similar or even slightly higher efficiencies (of the order of 22 and 41%, depending on the set of cuts) he obtains better purities (93-97%) with the event tag. The statistical error is the same for both methods. For the hemisphere tag the systematic error related to the background can be high, however, the efficiency can be obtained from the data. For the event tag the background contribution is small, and the error on the efficiency is smaller than the statistical error. All this suggests to use the event tag. A comparison of different energy runs shows that the efficiency as a function of the prob. cut is similar at all energies within 5%. The minimum error is obtained for an efficiency of about 40%. This estimate stems from a systematic error taken from the Z sample, and a statistical error calculated from the efficiency. For this efficiency, the difference in purity for the two tagging methods is about 8%. To get an estimate of the systematic error, he compares the measurement of Rb performed with the 1996 Z data to the Standard Model value, again as a function of the prob. cut. The problematic part now is the extrapolation of this error to the higher energies, which is still under investigation. At 161 GeV he measures a bbar x-section (s'/s>0.9) of 4.8 +- 1.0(stat) pb, (24 events), where he expects 5.69 pb from Zfitter (Effb=41%,Purity=93%). At 172 GeV he measures a bbar x-section (s'/s>0.9) of 3.8 +- 0.9(stat) pb, (18 events), where he expects 4.59 pb from Zfitter (Effb=42%,Purity=95%). The result at the 1996 Z data is : Rb=0.210+-0.004(stat) (3965 events), Effb=47%,Purity=97%, to be compared to the world average of Rb=0.2170+-0.0009. From these measurements he extracts limits on lepto-quarks (which couple to b-quarks), which are improved w.r.t. to our old results. Conclusions: ------------ The difference in efficiency at the high energies and the fact that the background eff. is not well described has to be understood. The latter aspect is a further reason to go to high purity. He doesn't apply the VDET smearing in the b-tagging. Should be done. Dieter suggests to look at the 183 GeV data where we have high statistics. This will of course be done, however, for final numbers one has to wait for a debugged MC production. The impact of the database error is not very clear, but it should be noted that in a Higgs HA-analysis differences in efficiency of a couple of % have been found. There is already a MC set corrected for the bug (by the Wisconsin group), which is available for tests. At the end the decision is taken not to show Rb at the LEPC, but one should go for a finalized analysis to be shown at the winter conferences.
- Current status of leptonic x-sections and Afb, Z' limits -------------------------------------------------------- Marc is analyzing all leptonic channels. In this presentation he focused on the muonic channel (tau channel not ready and strange features for Bhabhas, see later). He takes all the available data (PERF/MAYB) above the Z peak, the available MC at 162,172 and 183 GeV, and points out that there is no MC available at 133 GeV. A study of the trigger efficiency has still to be done. Muon Channel: ------------- After a standard selection for good tracks a muon preselection is applied. He asks for at least 2 and a maximum of 8 good tracks, then takes the two most energetic ones, which have to be antitagged as electrons, and at least one of them has to be tagged as a muon. Note that for the particle id he is using an extended version (to be used at LEP2) of the known EPIDOR/ARBOR packages. A note on this will be released soon. Then a further selection is applied. The sum of momenta (from the TPC) has to exceed 60 GeV, missing mass =< invariant mass. This last cut gives efficiencies down to the lowest energies. For the low energy part a cut on sqrt(s'/s)>0.1 or 0.122 is applied, for the high energy part one needs sqrt(s'/s)>0.9 or 0.85 and InvMass>87%*Elep. He points out that he has found some runs in the 172 run where one HCAL sector was all fired up. Therefore there is no mu/pi id. possible. To be followed up. Frederic asks about the motivation for a different selection w.r.t. to the previous one (by Elsa Merle). Marc answers that for the moment no rigorous comparison between the two analyses has been done, which will be necessary for a decision to be taken. Marie-Noelle mentions that the gamma-gamma -> mu mu background description is problematic. Marc answers that he doesn't find a bad agreement. This has to be understood. In particular, it has to be checked if Marc and Elsa are using the same MC. In addition, there should be an exchange of event lists between them in order to understand the various disagreements between their analyses. For the rejection of the cosmics background he tries to use data only and cuts on two (practically uncorrelated variables) : a) Sum of all ITC hits of the two muon candidates b) The minimum of the distances d1 for both candidates, where d1 is like the d0 measure but corrected for the actual beam position. When plotting these two variables on a plane, the signal region is defined as a) >= 8 and b) =< 0.2. The remaining three fields are cosmics. Under the assumption that the variables are not correlated, it is possible to extract the contamination of the signal region by looking at the event ratio in the other regions. At 161 this gives a contamination of 0.4 events, at 172 GeV further detailed studies are under way. In the end he shows tables with the results for the muonic x-sections and Afb, efficiencies and backgrounds, at 161 and 172 GeV. Efficiencies from 80 to 91% are obtained, with backgrounds of 4-5 (0.2-0.9)% for the inclusive (exclusive) sample. The agreement with the SM predictions (Zfitter) is reasonable. He cannot use the 183 MC because of the known bug in the database, which affects the muon momentum a lot. He also points out that there should be a decision taken on the final s' definition to be used. Bhabha : ------- He uses a similar preselection as in the muon case, with the obvious inversion of the particle id requirements. Then the sum of the e+,e- momenta (from TPC) has to exceed 20 GeV, the invar. mass must be greater than 60 GeV, the total charge must vanish, |cos(theta*)|< 0.9 or -0.9
- Discussion ---------- Marie-Noelle: ------------- She would like to present the x-sections to the BEW group before the LEPC. Thus, Guenther will organize a meeting before the next Thursday meeting, at the beginning of November. As soon as we have x-sections, the contact term fits can be done easily by Frederic. Henri-Videau: ------------- We should first see what will be shown at the next LEPC, i.e., just detector performances or also physics results. Frederic: --------- He wonders if there is anybody working on a light quark selection. It seems that there is nobody working on it. Dieter: ------- He proposes to go for a LEP jamboree before the next winter conferences. There is general agreement in the group to do so.
- Next meeting ---> around the 5th of November
Guenther Dissertori
Sun October 26 , 18:50:38 MET DST 1997