Ghosts in Run-10 200GeV

1. Ghosts in Run-10 200GeV MihaelMakek WIS 30-Mar-11

2. Contents • DC Ghosts • RICH ring sharing

3. I. DC Ghosts Cuts: |phi1-phi2|<0.05 AND |z1-z2|<0.5 • Treating the DC ghosts: • Select the pairs which fall inside the window • Randomly discard on of the tracks and remove it from the buffer • The distributions show no visible centrality dependance • Using the same cuts for all centralities

4. I. DC Ghosts • The distributions after randomly removing one of the tracks

5. I. DC Ghosts Left: the total number of electron tracks per event vs. centrality Right: the percentage of the tracks removed by DC ghost cut

6. I. DC Ghosts Left: the +- yield removed by the DC ghost cut (foreground) Right: the fraction of the remaining +- background after the DC ghost cut (as for efficiency correction)

7. II. RICH ring sharing Cuts: |phi1-phi2|<0.08 AND |z1-z2|<25 • The peaks fitted with a double Gaussian - the widths show no significant centrality dependance • Using the same cuts for all centralities

8. II. RICH ring sharing Counting the events with the RICH ghosts: Left: the percentage of the ghost events in total events Right: the percentage of the ghost events in events with ntrk > 1

9. II. RICH ring sharing • Use HBD as the first step of the rejection • In the remaining cases remove the event Left: N+- with CA only Middle:N+- with CA + HBD  The fraction of the ghost yield in the total yield drops from 0.032 to 0.024 when applying HBD (for 40 < centrality < 100) Right:N+- with CA + HBD and ghost event rejection The mass spectra for: 40 < centrality < 100

10. II. RICH ring sharing Left: the yield removed by the RICH ghost cut (foreground) Right: the fraction of the remaining +- background after the RICH ghost cut (as for efficiency correction)

11. II. RICH ring sharing • Consistency check: we want see that the fraction of ghost events obtained from the counters corresponds to the ghost yield obtained from the mass spectra: • The numbers are in agreement (see the next slide)

12. II. RICH ring sharing From the event counters: From the mass spectra: CA: CA+HBD:

13. Backup slides

14. Electron track distributions CA + projection cut CA + projection cut + HBD matching

15. The data set, event cuts, eID cuts • Run-10 – 200 GeV Au+Au data • EWG files: 530/828 runs, the fraction of runs with HBD gain calibrated • Events 3.5B out of 7.0B in +/-20 cm vertex • Event cuts: • abs(bbcz) < 20 cm • Track selection: • track quality = 31, 51, 63 • eID cuts: • n0 > 2 • sqrt(emcsdphi*emcsdphi + emcsdz*emcsdz) < 3 [#] • dep > -2 [#] • chi2/npe0 < 10 • disp < 5 • prob > 0.01 • |zed| < 75 [#]EMC recalibrators for 62GeV - Deepali M. Makek - WIS

16. Approach to ghosts • In the DC: • Select pairs with |f1-f2| < 0.05 && |zed1-zed2|< 0.5 • Randomly remove one of the tracks • In the RICH: • Most of the ring sharing pairs are formed by an electron and a hadron • Most of the electrons are coming from the HBD backplane and can be rejected by looking for a hit in the HBD (R~10) • Most of the hadrons can be rejected by the HBD (R~10) • We expect that HBD can reduce the number of ghost pairs up to a factor of ~100 • We want to use the HBD as the first step for the rejection of the ring sharing pairs M. Makek - WIS