Cross section Measurement of pp Zbb Z process

Cross section Measurement of pp → Z+bb, Z→ process at CMS Aruna Kumar Nayak TIFR, Mumbai, India (On behalf of CMS Collaboration) 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 1

Introduction Ø Measurement of Zbb production is an important test of QCD calculation Ø Background to Higgs discovery channels at LHC, like SM H → ZZ → 4 l, SUSY bb. F, F → tt (mm) Dominant at LHC Ø Z + 1 b-jet has been measured both at CDF & D 0 Ø The possibility of observing and measuring the production of Z + 2 b-jet at LHC has been studied aiming at early 100 pb-1 of CMS data. ~ 15% of bb. Z total s 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 2

bb. Z measurement can help reducing uncertainty of bb. H cross-section calculation 4 flavor scheme 5 flavor scheme (Z) Comparison: hep-ph/0405302 NLO in 4 -flavour scheme: NLO in 5 -flavour scheme: At least one b in detector acceptance S. Dittmaier, M. Kramer, M. Spira 2003 J. Campbell, R. K. Ellis, F. Maltoni, S. Willenbrock 2003 ~ 20 % uncertainty due to scale ~20% uncertainty due to scale ~10 % uncertainty due to PDF Mh=120 Ge. V Cut on p. Tb ~10 % uncertainty due to PDF

Cross section and Event generation Signal bb (Zbb) : Comp. HEP events with p. T(b) > 10 Ge. V, |h|(b) < 10 , mll > 40 Ge. V, |h|(l) < 2. 5 were generated and fully simulated in CMS 100 pb-1 calibration and mis-alignment Cross section calculated using MCFM, NLO s (llbb) = 45. 9 pb , l = e, m, t PDF : CTEQ 6 M, scale m. R = m. F = MZ LO cross section calculated using PDF : CTEQ 6 L 1 and same values for scale K (NLO) = 1. 51 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 4

Cross section and Event generation Backgrounds tt~ + n jets, n >= 0 : Generated using ALPGEN Cross section normalized to NLO inclusive tt~ cross section 840 pb llcc + n Jets, n>= 0 (Zcc) : Generated using ALPGEN Normalized on NLO s (using MCFM) 13. 29 pb, K factor = 1. 46 p. T(c) > 20 Ge. V, |h|(c) < 5, mll > 40 Ge. V with cuts : ll + n Jets, n >= 2, (Zjj) : Generated using ALPGEN Normalized on NLO s (using MCFM) 714 pb , K factor = 1. 02 with cuts : p. T(j) > 20 Ge. V, |h|(j) < 5, mll > 40 Ge. V 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 5

Initial Event selections: Trigger selection : single isolated electron or muon Level-1 threshold 12 Ge. V, 7 Ge. V & High-Level threshold 15 Ge. V, 11 Ge. V Corresponds to low luminosity period L = 1032 cm-2 s-1 Lepton Selection : Two high p. T, isolated, opposite charged leptons |h|(e) < 2. 5, |h|(m) < 2. 0, lepton p. T > 20 Ge. V Jets Selection : Two or more jets with corrected ET > 30 Ge. V , |h| < 2. 4 Jet corrected using Monte Carlo jet energy correction. 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 6

b-Jet Tagging Lepton, jet selections + double b-tagging with b-discriminator > 0. b-discriminator of 2 nd highest discriminator jet Jets are tagged using “Track Counting b-tagging” Which uses the 3 -dimentional impact parameter significance , of 3 rd highest significance track, as the b-tagging discriminator Effective to supress the Z+jets backgrounds. 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 7

b-tag efficiency b-tagging efficiency for b, c, light jets after applying cut on b-discriminator > 2. 5 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 8

ETmiss selection Lepton, jet selections + double b-tagging with b-discriminator > 0 Missing ET reconstructed from calorimeter and corrected for Jet Energy scale and muons. Type-1 ETx, ymiss = - (ETx, ycalo + Sjets(ETx, ycorr – ETx, yraw)) Muon corr. = - (Smuons (px, y – Ex, y (calo. deposit))) Effective to supress the tt~+jets backgrounds Cut ETmiss < 50 Ge. V 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 9

Event Selection details Two Leptons, p. T > 20 Ge. V, |h|(e) < 2. 5 , |h|(m) < 2. 0 Two or more Jets , ET > 30 Ge. V , |h| < 2. 4 Two b-tagged Jets Missing ET < 50 Ge. V Initial and final cross sections after all selections Process Name s NLO (pb) Final s (fb) Electron Muon Zbb 46 176 212 tt~ + jets 840 173 178 Z+jets 714 5. 5 Zcc+jets 13. 3 4. 3 5. 1 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 10

1 Expected Measurement for 100 pbevents scaled to 100 pb-1 Purity of b-tagging in Zbb events 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 11

1 Expected Measurement for 100 pb. Z → mm final state Z → ee final state 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 12

tt~ background Estimation Dilepton mass region Signal : 75 -105 Ge. V (Z) Side band : 0 -75 Ge. V & 105 – above (no Z) NZ(tt) = (e. Z(tt)/eno. Z(tt)) X Nno. Z(tt) DNZ(tt)/NZ(tt)= 1/√Nno. Z(tt) NZ(tt) = expected no. of tt~ events in signal region Nno. Z(tt) = measured no. of tt~ events out side signal region e. Z(tt) = selection efficiency of tt~ in signal region eno. Z(tt) = selection efficiency of tt~ outside signal region DNZ(tt) = uncertainty of the expected number of tt~ events in the signal region. Uncertainty on e. Z(tt)/eno. Z(tt) is negligible compared to the statistical uncertainty on Nno. Z. 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 13

Systematics Uncertainty due to Background and double b-tagging. NZbb and DNZbb are determined as follows. NZbefore b-tag = NZjj + NZcc + NZbb NZafter b-tag = el X NZjj + ec X NZcc + eb X NZbb Where, NZbefore b-tag = measured number of Z/g* → ll events after all selections except b-tagging under Z mass peak (75 -105 Ge. V). Contribution of tt~ is negligible (~1%). NZafter b-tag = measured number of Z/g* → ll events after all selections including b-tagging with tt~ subtracted NZjj is unknown number of ll+jets (u, d, s, g) events before double b-tagging. NZcc is unknown number of Zcc events before double b-tagging. NZbb is unknown number of Zbb events before double b-tagging. (after all selections except b-tagging) eb, ec, el are the efficiency of double b-tagging for Zbb, Zcc and Z+jets events ( Ratio of number of events before and after double b-tagging) 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 14

Contd. . . Using D 0 analysis approach, Use the Ratio where is ratio of selection efficiencies Solving the equations The Uncertainties on NZbb is calculated from uncertainties of NZafter b-tag (uncertainty due to tt~ subtraction), d. R and uncertainties on eb, ec, el 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 15

Total Uncertainty on Measurement Source of uncertainty Value used (%) d(s(Zbb)) (%) Jet energy scale (JES) 7 7. 6 Type 1 missing ET scale 10 (unclustered ETmiss) + 7 (JES) 7. 4 MC p. Tjet, hjet dependence -10, +0 b-tagging of b-jets (deb) 8 16 mistagging of c-jets (dec) 8 0. 5 mistagging of light jets (del) 7. 6 0. 5 NZafter b-tag due to tt~ subtraction 4 4. 6 R (Zcc / Zjj) 5 0. 4 lepton selections 0. 5 luminosity 10 10 Total cross section is expected to be measured in 100 pb-1 of data with uncertainty ds = +21%, - 25% (syst. ) , +/- 15% (stat. ) 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 16

Conclusion Ø The possibilty to measure the bb. Z, Z → ll process at CMS has been studied, aiming the early 100 pb-1 of data, with robust selection of leptons and jets. Ø Possible methods to measure backgrounds from data has been discussed. Ø The statistical and systematic uncertainties has been studied. 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 17

backup 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 18

Systematics due to JES and MET scale Uncertainty due to JES : Calculated by shifting Jet Energy by 7% Rel. Uncert. = (emax – emin)/2 e Uncertainty due to MET scale : Calculated by shifting Raw MET unclustered in jets 10% and corrected Jet energy by 7% Rel. Uncert. = (emax – emin)/2 e ET(x, y)miss. corr = acalo X ET(x, y)miss. calo + jets(ajet X ET(x, y)jet. corr – acalo X ET(x, y)jet. raw) acalo = ajet = 1 in ideal calibrations. acalo was varied to 10% and ajet was varied to 7%. Total systematic uncertainty due to JES is 7. 6% Total systematic uncertainty due to Missing ET scale is 7. 4% 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 19

Uncertainties due to lepton selections : total efficiency for lepton etotal = e 2 off-line X etrigger with 10 pb-1 the syst. uncert. of etotal on e+e- final state is 1%. (CMS PAS EWK-07 -001) for 100 pb-1 , rescaled by √ 10 => 0. 3% for m+m- the systematic uncertainty is 0. 5% We use 0. 5% for both ee and mm final state. Uncertainty due to MC p. T and h of jet dependence for signal events It is defined from the comparison of ratio of LO and NLO MCFM cross sections for different set of cuts on p. T (jet) (10 and 30 Ge. V) and h (jet) (10 and 2. 4). p. T(jet) > 10 Ge. V, |h|(jet) < 10 : Comp. HEP generator level cuts p. T(jet) > 30 Ge. V, |h|(jet) < 2. 4 : experimental like selection cuts The LO ratio is 10% bigger than the NLO ratio. Statistical Uncertainty : Defined as DNsel /Nsel = 1 / √Nsel , where Nsel = 46 , measured number of events after all selections. => d. Nsel = 14. 7% 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 20

eb = 0. 0654, ec = 0. 0051, el = 0. 00025 NZcc = R X NZjj The uncertainty on the central value The uncertainty on d. NZafter b-tag = 4%, ratio is +/- 3% due to scale variation around m. R = m. F = MZ from MZ /2 to 2 MZ due to JES is +/- 1. 6% and ETmiss scale is +/- 3. 6% uncertainty on NZafter b-tag due to tt~ subtraction. Uncertainties on eb, c, l are approximated as 2 X Deb, c, l deb = 8% (per jet) (CMS PAS BTV-07 -001) dec = 8% (per jet) used the same value as for b-jets. del = 7. 6% (per jet) (CMS PAS BTV-07 -002) 29 th Sep. - 4 th Oct. 08 Physics at LHC, Split 21