1. Multiplicity as a measure of Centrality in Richard S Hollis University of Illinois at Chicago

2. Collaboration (May 2004) Birger Back, Mark Baker, Maarten Ballintijn, Donald Barton, Russell Betts, Abigail Bickley, Richard Bindel, Wit Busza (Spokesperson), Alan Carroll, Zhengwei Chai, Patrick Decowski, Edmundo García, Tomasz Gburek, Nigel George, Kristjan Gulbrandsen, Clive Halliwell, Joshua Hamblen, Adam Harrington, Michael Hauer, Conor Henderson, David Hofman, Richard Hollis, Roman Hołyński, Burt Holzman, Aneta Iordanova, Jay Kane, Nazim Khan, Piotr Kulinich, Chia Ming Kuo, Willis Lin, Steven Manly, Alice Mignerey, Gerrit van Nieuwenhuizen, Rachid Nouicer, Andrzej Olszewski, Robert Pak, Inkyu Park, Heinz Pernegger, Corey Reed, Michael Ricci, Christof Roland, Gunther Roland, Joe Sagerer, Helen Seals, Iouri Sedykh, Wojtek Skulski, Chadd Smith, Maciej Stankiewicz, Peter Steinberg, George Stephans, Andrei Sukhanov, Marguerite Belt Tonjes, Adam Trzupek, Carla Vale, Siarhei Vaurynovich, Robin Verdier, Gábor Veres, Edward Wenger, Frank Wolfs, Barbara Wosiek, Krzysztof Woźniak, Alan Wuosmaa, Bolek Wysłouch ARGONNE NATIONAL LABORATORYBROOKHAVEN NATIONAL LABORATORY INSTITUTE OF NUCLEAR PHYSICS, KRAKOWMASSACHUSETTS INSTITUTE OF TECHNOLOGY NATIONAL CENTRAL UNIVERSITY, TAIWANUNIVERSITY OF ILLINOIS AT CHICAGO UNIVERSITY OF MARYLANDUNIVERSITY OF ROCHESTER Richard Hollis, UICFocus on Multiplicity, Bari 2004

3. Outline Considerations – Centrality in 200GeV Au+Au –Event Selection –Detector Efficiency –Choose Data η region to measure centrality –Event Generator Simulations – N part Other Collision Systems –d+Au 200 GeV –Au+Au 19.6 GeV Richard Hollis, UICFocus on Multiplicity, Bari 2004

4. Paddle Counters and ZDCs ZDC (hidden) Richard Hollis, UICFocus on Multiplicity, Bari 2004 The Detector Ring Counters Octagon and Spectrometer

5. A clean event selection ensures consistency of good events Only use data within ±4ns (~±60cm) from Paddles –Red points → outside this cut the acceptance changes Further restrictions requires ZDC signal Richard Hollis, UICFocus on Multiplicity, Bari 2004 Beam-gas collision peaks Considerations: Event Selection 200GeV Au+Au Collisions Schematic Diagram of Paddles

6. Considerations: Event Selection A clean event selection ensure consistency of good events Only use data within ±4ns (~±60cm) from Paddles –Red points → outside this cut the acceptance changes Only use Silicon vertex range of ±10cm for physics measurements (blue) Richard Hollis, UICFocus on Multiplicity, Bari 2004 Increase >4ns due to change in acceptance Paddle time difference widens due to low multiplicity → smearing effects Flat over the region used

7. Considerations: Detector Efficiency Trigger system incurs small bias for peripheral events –Have to account for the missing cross- section Richard Hollis, UICFocus on Multiplicity, Bari 2004 Have two main trigger types (in Au+Au) One or more hit in each paddle array –Estimated to have 97% efficiency More than 2 hits in each paddle array –Estimated to have 88% efficiency

8. Considerations: Detector Efficiency Trigger system incurs small bias for peripheral events –Have to account for the missing cross- section Estimate the efficiency –Using Data and MC simulations Richard Hollis, UICFocus on Multiplicity, Bari 2004 Unbiased Hijing sample 97% efficient trigger (red) Data 88% efficient trigger (blue) Data Number of hit paddle segments Similar plateau observed in data

9. Considerations: Data Centrality Region Paddles are located in 3.2 < |η| < 4.5 Monotonic anti- correlation with neutral spectators Region appears to be usable Richard Hollis, UICFocus on Multiplicity, Bari 2004 200 GeV Au+Au data for 0-25% cross-section

10. Centrality Determination Summary of Information –Have clean event selection –Estimated the trigger efficiency –Centrality determined from Paddles (3.2<|η|<4.5) –Observed spectators are monotonic with that signal Need one more piece –Does MC predict monotonicity between N part and Paddle signal? Richard Hollis, UICFocus on Multiplicity, Bari 2004

11. Centrality Determination Nicely correlated –Now have all the information needed Hard work is done –Divide the data into cross-section bins –Use MC to estimate for each bin Richard Hollis, UICFocus on Multiplicity, Bari 2004 Only use data where (Si) vertexing finding efficiency is 100% (top 50% of cross-section)

12. Centrality Complete 200 GeV Au+Au is well established Now investigate the centrality of d+Au collisions at 200 GeV

13. Centrality at 200GeV d+Au Same Considerations –Details of analysis is different Event Selection –Clean-up by requiring a valid silicon vertex Very Low multiplicity Events –Paddle timing smeared Not always neutron in both ZDCs Efficiency –Used a shape matching algorithm between Data and Simulations (HIJING or AMPT) –Efficiency includes Trigger and Vertex finding efficiency –Estimated to be 82.5% Richard Hollis, UICFocus on Multiplicity, Bari 2004 Hijing + GEANT Data Shapes agree reasonably in High multiplicity region Data inefficient for peripheral events EOct is the summed charge deposited in the Octagon detector

14. Unique PHOBOS η coverage –Many regions to pick from –Not just the ‘paddles’ All regions were used –same basic algorithm –Sum the charge deposited in these regions (from Silicon) d+Au Data Centrality Regions Richard Hollis, UICFocus on Multiplicity, Bari 2004 EOct ERing ETot EdHem EAuHem

15. All Centrality methods agree when reconstructing the min-bias distribution Unique PHOBOS η coverage –Many regions to pick from –Not just the ‘paddles’ All regions were used –same basic algorithm –Sum the charge deposited in these regions (from Silicon) d+Au Data Centrality Regions Richard Hollis, UICFocus on Multiplicity, Bari 2004

16. d+Au Data Centrality Biases Two types of potential biases 1.Auto-correlation biases –Measurement of centrality interferes with physics measurement –Causes a change in shape of dN/dη distribution 2.Trigger-Biases –Due to inefficiency for peripheral collisions –Changes of a bin Richard Hollis, UICFocus on Multiplicity, Bari 2004

17. Auto-Correlation Biases Mid-rapidity Centrality, EOct –Shown for 3 cross-section bins Form a Truth distribution –Divide the N part distribution into bins with the same as EOct –Repeat analysis Can see a clear bias developing due to the centrality determination Richard Hollis, UICFocus on Multiplicity, Bari 2004 AMPT+GEANT Simulations EOct Npart

18. Auto-Correlation Biases Mid-rapidity Centrality, ERing –Shown for 3 cross- section bins Bias is smaller Richard Hollis, UICFocus on Multiplicity, Bari 2004 AMPT+GEANT Simulations ERing Npart

19. Centrality at 200GeV d+Au With these two things –can make the centrality dependence Can now integrate these distributions for the total charge –Scale by N ch pp Results consistent with expectations from low energy data Richard Hollis, UICFocus on Multiplicity, Bari 2004

20. Centrality at 200GeV d+Au With these two things –can make the centrality dependence Can now integrate these distributions for the total charge –Scale by N ch pp Results consistent with expectations from low energy data Richard Hollis, UICFocus on Multiplicity, Bari 2004

21. d+Au Centrality Complete Final Example: Au+Au collisions at 19.6 GeV

22. Centrality at 19.6GeV Au+Au Hit number of Paddles efficiency method –Works at 62.4 GeV –Does not work at 19.6 GeV Too low multiplicity No plateau observed Multiplicity in paddle’s region too low Richard Hollis, UICFocus on Multiplicity, Bari 2004 See the same plateau

23. Centrality at 19.6GeV Au+Au Use summed charge in octagon –EOct as in d+Au Could this introduce a bias? Make new Centrality measures to check this Richard Hollis, UICFocus on Multiplicity, Bari 2004

24. Centrality at 200GeV Au+Au Measure same results for –midrapidity methods (a) –away from midrapidity (b) For midrapidity yields for the top 50% of the cross- section Aneta Iordanova will talk about this measurement in the next talk Richard Hollis, UICFocus on Multiplicity, Bari 2004 200GeV from (a) 200GeV from (b)

25. Conclusions We have established methods for centrality determination –From 19.6 to 200 GeV Au+Au –From d+Au to Au+Au collision systems Have an array of complementary techniques for centrality cross-checks Techniques used introduce little or no bias onto the physics measurements –Any biases introduced can be estimated and corrected for from MC studies Richard Hollis, UICFocus on Multiplicity, Bari 2004