1. Search for the SM Higgs Boson in the H  γγ Decay Channel and Calibration of the CMS Electromagnetic Calorimeter with π 0  γγ Decays Marat Gataullin, Vladimir Litvine, Yong Yang Marat Gataullin, Vladimir Litvine, Yong Yang DoE Review, July 25, 2007

2. Integrated luminosity for 5  discovery  Keys: Clean Photon ID, 0.7% Mass Resolution, Precise Calibration  Next Steps: NLO Monte Carlo Generator for Higgs & Backgrounds; Optimize S/B Separation; Study ECAL Calibration Effects Higgs Signal and Backgrounds Caltech + UCSD Optimized H   Analysis Fitting NN and Mass for Higgs and QCD

3. Higgs  : Vector Boson Fusion forward jets Photons from Higgs decay qqH → qq γγ M H = 120 GeV Jets from qq are at high rapidity and large Δη. Jet- tagging gives a background reduction of 95% M H M H After photon selection After photon selection After Jet Tagging 120 GeV 37.1%16% Selection Efficiency:

4. Higgs  : Vector Boson Fusion CMS Note 2006/097: Two scenarios considered (Caltech) CompHEP includes the complete set of tree level (leading order) diagrams for the partonic subprocess ug→  gu

5. 5 Modifications to the Properties of the Higgs Boson Manohar and Wise, Phys. Lett. B636 (2006) 107-113 Motivation  “Minimal” solution to the Hierarchy Puzzle, through new physics at  ~ 1 TeV LHC: Promising Scenario  gg →h already at one loop: a higher order process in perturbation theory  WW → h, ZZ → h etc. affected less  New Top-like mesons also could appear A New Effective Theory  New “dimension 6” operators that couple the gluon & higgs fields  0.01 < C G < 0.1 from Tevatron data and neutron dipole moment  Higgs gg fusion rate could be several times, or much less than in the SM … T T T

6. Effects on H    Discovery  SURF’06 (Yike Lu)  Higgs can be discovered much faster or slower !  We plan expansion to other decay channels  Can be inverted to measure the Higgs couplings  Working on combining with with H  ZZ*  4μ channel (also for the SM search)

7. Precise ECAL Calibration with π 0 Data after L1 TriggerOnline Farm  0 Calibration >10 kHz ~1 kHz  Level 1 trigger rate dominated by QCD: several π 0 ‘s/event Useful π 0  γγ decays selected online from such events.  Main advantage: high π 0 rate (nominal L1 rate is 100kHz !) No track reconstruction (no alignment) required.  “Design” calibration precision  better than 0.5% Achieving this would be crucial for a fast H  γγ detection.  Studies: Selected 0.3M π 0 from 5M fully simulated QCD events Scenario L =2x10 33 cm -2 s -1 and L1 rate of 17 kHz (end of 2008).  Also works at lower instantaneous luminosities, at the startup !  Alternative strategies ( W  e ν) require 5-10 fb: months/years.

8. π 0  γγ Selection Selection based on local, crystal-level variables — suitable for online  Kinematics: P T ( γ ) >1 GeV, P T (pair) > 3.5 GeV and η < 1.48 (barrel)  Simple cuts on photon shower shape and isolation to remove converted γ ’s Trigger Tower (5x5 crystals)

9. Selection Results rate of 1.5 kHzor 2,100 π 0 /crystal/day with S/B ≈ 2.0 π 0  γγ rate of 1.5 kHz or 2,100 π 0 /crystal/day with S/B ≈ 2.0 High-rapidity regions suffer both in rate and S/B (3  1) Cracks between supermodules give a -1.5% shift and selective readout: a -0.4% shift with a period of 5 crystals. Dedicated procedure corrects to a 0.1% level. CMS-IN 2007/002

10. Calibration Performance Precision is then fitted to N is the number a=27±1% and b=0.20±0.25% of π 0 /crystal Calibration performed using an iterative algorithm developed for the RFQ calibration at L3, where we achieved a 0.5% calibration precision

11. π 0  γγ in the Endcaps (preliminary)  π 0  γγ rate comparable to that in the barrel (selection cuts applied)  The same event selection approach, with slightly relaxed P T cuts  First results are promising: calibration will take only ~5 times longer than in the barrel. Working on the ECAL+Preshower analysis.  Currently this is the only viable calibration technique for the endcaps. 2x10 33 cm -2 s -1 10 32 cm -2 s -1

12. Calibration Studies in Test Beams π 0 decays produced through: π - +Al  π 0 +X (11/2006) Three different π - beam energies: 9, 20, and 50 GeV Consider only 9x8 crystal matrix: about 140 π 0 decays/crystal Caltech group co-lead this effort (with the University of Minnesota)

13. First Resonance Observed by CMS Improvement over the uncalibrated peak (L3 algorithm): 7%  5.5% Currently working on π 0 test beams for the endcaps (October/2007): redesigning the target and improving the trigger setup. π 0  γγ produced in upstream scintillators

14. Calibration Precision with 50 GeV Electrons For each crystal, electron energy spectra fitted to a Gaussian. Distributions of the obtained peak positions for 9x8 crystal matrix: Precision: 1.0±0.1% with 0.9±0.1% expected. Calibration with ~5 GeV photons works well for higher-energy showers! CMS-DN 2007/007

15. 15  μ + μ - γ  μ + μ - γ production through final-state radiation provides a valuable tool to calibrate and study the response of the CMS ECAL.  30,000 events with P T (γ) >10 GeV produced after 1 fb -1, allowing us to perform ring-by-ring ECAL calibration to 0.5% precision; faster and independent of tracker systematic effects, which will affect the Z  ee calibration.  Complementary to the ongoing of L1/HLT trigger efficiency study using Z+ γ events. μ + μ - γ Final State Topology Ashok Kumar, Jan Veverka

16. Conclusions and Outlook with full detector simulation:  Proof-of-principle achieved with full detector simulation: crystal-by-crystal calibration to 0.5% after 20-80 hours crystal-by-crystal calibration to 0.5% after 20-80 hours at L=2x10 33 cm -2 s -1 (50-200 hrs. at L=10 32 cm -2 s -1 : startup). at L=2x10 33 cm -2 s -1 (50-200 hrs. at L=10 32 cm -2 s -1 : startup). Other methods are much slower and tracker dependent. Other methods are much slower and tracker dependent.  Many months of work on understanding the ECAL performance. Very useful for our physics analyses ( performance. Very useful for our physics analyses ( H    Test beams demonstrated a 1% calibration precision with ~5 GeV photons: successful reconstruction of with ~5 GeV photons: successful reconstruction of 50 GeV electrons. No noticeable systematics (~0.3% from test beam). 50 GeV electrons. No noticeable systematics (~0.3% from test beam).  Caltech is playing a leading role in this multi-national effort: Detector Performance Group task led by Gataullin/Litvine.

17. Extra Slides Follow

18. CompHEP EW background: ud→  du This background topology is very similar to Higgs signal CompHEP EW 2  +2jets background has smaller cross section compare to QCD 2  +2jets background (300 fb vs 50 pb), but has long hard tails in p T distributions and many photons at small  from ladder diagrams like 3,4. These tails are much harder than for the CompHEP QCD 2  +2jets background sample

19. Correcting for Cracks and SRO Cracks between baskets/supermodules give a -1.5% shift. Selective readout: a -0.4% shift with a period of 5 crystals. Dedicated procedure developed to correct to 0.1-0.2% level.

20. Calibration Algorithm Iterative algorithm (successful L3/RFQ Calibration) (w i  fraction of shower energy deposited in this crystal)  Both photon energy and direction reconstructed using crystal level information (same as during selection).  After each iteration pairs are re-selected with new constants (typically 10-15 iterations to converge).  Miscalibration is done before selecting events (4%).  Calibration precision defined as R.M.S. of the product of the final and initial miscalibration constant.  Use only pairs from ±2σ window around fitted π 0 mass

21. Effects of Corrections Correcting for cracks and selective readout gives an improvement of 0.9% (in quadrature), and removing pileup eliminates an additional constant term of 0.6%.

22. Calibration Performance Precision is then fitted to a=27% and b=0.2%