| QCD Update "Photon Run I Update" Dana Partos |
|
March 23, 2001
|
|
Summary: The CES and CPR are two main methods for calculating the photon cross section. Ideally, both methods should give the same cross section. In Run 1a the two cross sections agreed to 2%. Originally, the neither the 1b or 630 GeV cross sections agreed. We recalibrated the background efficiencies for the two methods using pi0's from charged rho decays. We discovered that there was a significant discrepancy between the measured and expected CPR and CES pion efficiencies. We applied the recalibrated efficiencies to the CES and CPR cross sections to both data sets and the CES and CPR methods were brought together. At the time the 1b and 630 GeV photon cross sections were blessed there were several things left to do in our analysis. The main things were recalibrating the signal efficiencies using the photons from eta decays, measuring the trigger efficiencies, applying lateral shower leakage recalibrating the signal efficiencies using the photons from eta decays, measuring the trigger efficiencies, applying lateral shower leakage corrections, finding the source of the discrepancy between the measured and expected CPR and CES efficiencies, and calculating the CPR and CES systematic errors. This talk will describe the studies that were done to find the source of the CES/CPR discrepancies and the CPR systematics. The four possible sources are problems with the detector, and incorrect accounting of the material in front of the CPR, the underlying event, and the background mixture. The first check was of the CES energy calibration. The CES strip and wire energies were found to be 25% too low. Once the CES energy was recalibrated and a Z vertex cut was applied, the CES discrepancy disappeared, but the CPR discreapancy increased. We checked that the CPR chambers were working correctly by studying the CPR pulse height distributions of the rhos, QFL, and W electrons. We are satisfied that the CPR overall is working correctly. In addition, we used min bias data to check the CPR efficiency as a function of wedge number. Again, no problems were found with the detector. The material in front of the CPR was last measured before 1a. Since then the CDT was taken out and the Time Of Flight detector was added. The additional material from the TOF should increase the CPR efficiency. We measured the CPR efficiency of photon candidates as a function of phi. Since the TOF is in a limited phi range, we should see a bump in the CPR efficiency, but we did not. The underlying event was measured using min bias data. It was previously believed that the CPR efficiency only depends on the luminosity. We found that the CPR efficiency actually depends on the number of vertices, and the vertex distribution for min bias data is different from that of photon data. Unfortunately, the new underlying event measurement did not resolve our problem. We then looked at the CPR efficiency of non-isolated photon data to be sure that the min bias data accurately mimics the rho's underlying event. There is a difference between the two data samples, but it does nothing to resolve the discrepancy. Since our CPR efficiency prediction from the rhos assumes that the events in the rho peak are all pions. We checked to see if the events in the rho peak are all pions. We checked to see if the events in the peak could come from other mesons. Even if only 75% of the events in in the rho peak are all pions. We checked to see if the events in the peak could come from other mesons. Even if only 75% of the events in the peak are pions, the discrepancy would still not go away. Finally, we performed the rho analysis on the 630 GeV data to see if the 630 data had the same problem with the CPR efficiencies. The difference between the measured and expected values was identical to the difference in the 1b rho data. This left us with the decision of how to correct the photon and background CPR efficiencies. We know that the background efficiency needs to be corrected by the difference between the measured and expected rho efficiencies, but we don't know the what the photon efficiency correction should be. We decided to use the photon efficiency correction that best agrees with the CES cross section and then use other changes to the photon efficiency to get the systematic error. |
| Postscript (PS) file. |