cms theory outreach

CMS Theory Outreach Maurizio Pierini CERN Tuesday, October 30, 12 - PowerPoint PPT Presentation

CMS Theory Outreach Maurizio Pierini CERN Tuesday, October 30, 12 Outline I will discuss two examples of SUSY analyses and how they could be implemented in a phenomenology study I start with a simple cut&count case (SS Dilepton +


  1. CMS Theory Outreach Maurizio Pierini CERN Tuesday, October 30, 12

  2. Outline • I will discuss two examples of SUSY analyses and how they could be implemented in a phenomenology study • I start with a simple cut&count case (SS Dilepton + btag) • I then go to a combination of 6 mutually-exclusive razor analyses (Razor) • For both I discuss how to derive the ingredients (efficiency and shape) and how to get the limit out • When possible, I show a comparison between the outreach and the official results Tuesday, October 30, 12

  3. SS Dilep + Btag Tuesday, October 30, 12

  4. The SS+Btag Analysis -1 for the 10 events in the baseline region (SR0). CMS, s = 7 TeV, L = 4.98 fb int 80 100 120 140 160 180 200 200 (GeV) (GeV) Expected BG ee 180 miss > 80 GeV miss • 160 e µ T Data T E Look for events with two jets E 140 -1 T µ µ = 4.98 fb T H 120 • 100 int Requite two SS leptons (any flavor) = 7 TeV, L 80 60 60 s • 40 40 CMS, Ask for two btags 20 20 Figure 1: Top plot: distribution of E miss vs. H 0 0 T 1 2 1.8 1.6 1.4 1.2 0.8 0.6 0.4 0.2 0 0 100 200 300 400 500 H (GeV) 600 • Events / 10 GeV Counting experiment in eight overlapping H (GeV) 100 200 300 400 500 600 T 0 regions of the HT vs MHT plane 0.1 0.2 • Quote the best of the eight limits 0.3 Events / 10 GeV T E > 0 GeV miss 0.4 Data 0.5 Expected BG 0.6 line of the table includes both tagged and untagged jets. int CMS, s = 7 TeV, L = 4.98 fb -1 SR0 SR1 SR2 SR3 SR4 SR5 SR6 SR7 SR8 No. of jets ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 3 ≥ 2 No. of b-tags ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 2 ≥ 3 ≥ 2 Lepton charges + + / − − + + / − − + + / − − + + / − − + + / − − + + / − − + + / − − + + / − − ++ E miss > 0 GeV > 30 GeV > 30 GeV > 120 GeV > 50 GeV > 50 GeV > 120 GeV > 50 GeV > 0 GeV T H T > 80 GeV > 80 GeV > 80 GeV > 200 GeV > 200 GeV > 320 GeV > 320 GeV > 200 GeV > 320 GeV 1.4 ± 0.3 1.1 ± 0.2 0.5 ± 0.1 0.05 ± 0.01 0.3 ± 0.1 0.12 ± 0.03 0.03 ± 0.01 0.008 ± 0.004 0.20 ± 0.05 Charge-flip BG 4.7 ± 2.6 3.4 ± 2.0 1.8 ± 1.2 0.3 ± 0.5 1.5 ± 1.1 0.8 ± 0.8 0.15 ± 0.45 0.15 ± 0.45 1.6 ± 1.1 Fake BG 4.0 ± 2.0 3.4 ± 1.7 2.2 ± 1.1 0.6 ± 0.3 2.1 ± 1.0 1.1 ± 0.5 0.4 ± 0.2 0.12 ± 0.06 1.5 ± 0.8 Rare SM BG Total BG 10.2 ± 3.3 7.9 ± 2.6 4.5 ± 1.7 1.0 ± 0.6 3.9 ± 1.5 2.0 ± 1.0 0.6 ± 0.5 0.3 ± 0.5 3.3 ± 1.4 Event yield 10 7 5 2 5 2 0 0 3 N UL (12% unc.) 9.1 7.2 6.8 5.1 7.2 4.7 2.8 2.8 5.2 N UL (20% unc.) 9.5 7.6 7.2 5.3 7.5 4.8 2.8 2.8 5.4 N UL (30% unc.) 10.1 7.9 7.5 5.7 8.0 5.1 2.8 2.8 5.7 Tuesday, October 30, 12

  5. The SS+Btag Results ¯ t t A1 ¯ t t A2 B1 W − ˜ t ¯ χ 0 P 1 b t 1 ˜ W − ˜ χ 0 t ∗ B2 P 1 ˜ 1 1 χ − P 1 1 ˜ χ 0 ˜ P 1 χ − g 1 ˜ ˜ 1 ˜ χ 0 b 1 ˜ g 1 g ˜ ˜ b 1 ˜ χ + ˜ χ + ˜ ˜ ˜ b ∗ 1 χ 0 t ∗ 1 1 ˜ ˜ χ 0 g ˜ χ 0 1 1 ˜ ˜ 1 b ∗ χ 0 1 g ˜ 1 1 P 2 P 2 g ˜ W + P 2 W + b ¯ ¯ t t P 2 ¯ t t ¯ t t • No excess found • Several models considered to put limits ~ • We will try to reproduce their A1 result for the specific case of m( χ 0 )=50 GeV -1 CMS, s = 7 TeV, L = 4.98 fb -1 CMS s = 7 TeV, L = 4.98 fb int 2 int 10 800 ) GeV x BR pb Same Sign dileptons with btag selection NLO+NLL ~ ~ ~ ~ ∼ ∼ 0 ± prod Model B2: g g → b b , m( b ) = 500 GeV, m( χ ) = 200 GeV, m( χ ) = 50 GeV NLO+NLL Exclusion = 1 700 σ σ ± σ 1 1 1 ~ ~ ~ ~ ∼ 0 Model A2: g g → t t , m( t ) = 530 GeV, m( χ ) = 50 GeV 10 1 1 0 1 1 ~ ~ ~ ~ ∼ 0 ∼ χ Model A2: g g t t , m( t ) = 280 GeV, m( ) = 50 GeV → χ 1 1 m( 1 600 ~ ~ σ ∼ 0 ∼ 0 Model A1: g g → 4top + 2 χ , m( χ ) = 50 GeV 1 1 1 500 Same Sign dileptons with btag selection 400 -1 10 300 -2 10 200 100 -3 10 500 600 700 800 900 1000 1100 1200 400 500 600 700 800 900 1000 1100 ~ m( g ) GeV ~ m( g ) GeV Tuesday, October 30, 12

  6. The Theory Outreach CMS provides the information in the paper to reproduce the analysis • Efficiency for each physics object, as a function of the pT of the generator-level particle CMS Simulation, s = 7 TeV CMS Simulation, s = 7 TeV Lepton eff btag eff Efficiency Efficiency 1 1 0.8 0.8 0.6 0.6 electrons 0.4 0.4 0.2 0.2 muons 0 0 20 40 60 80 100 120 140 160 180 200 0 50 100 150 200 250 300 350 400 Lepton p (GeV) b-quark p (GeV) T T Figure 2: Lepton selection efficiency as a function of p (left); b-jet tagging efficiency as a CMS Simulation, s = 7 TeV CMS Simulation, s = 7 TeV Efficiency Efficiency 1 1 HT MET 0.8 0.8 0.6 0.6 miss E > 30 GeV T H > 200 GeV T 0.4 0.4 miss E > 50 GeV T 0.2 0.2 H > 320 GeV T miss E > 120 GeV T 0 0 0 50 100 150 200 250 300 350 400 450 500 0 100 200 300 400 500 600 700 800 miss gen-H (GeV) gen-E (GeV) T T miss • Number of observed events, expected, and error on the expected • A set of yield UL for different assumed errors, to calibrate the statistical method used offline Tuesday, October 30, 12

  7. The procedure ∞ | | δ P(s) 0.18 To get the signal efficiency 0.16 0.14 • Generate SUSY events for a given point of the SMS 0.12 plane (I used pythia8 here) 0.1 0.08 • Reconstruct gen-jets (I used anti-Kt with R=0.5) 0.06 0.04 • Apply the analysis cuts at gen level 0.02 0 0 2 4 6 8 10 12 14 16 18 20 • Use hit OR miss to accept or reject the event s in SR0 To evaluate the limit Region quote UL my UL SR0 10.1 9.8 • Derive a posterior for the signal yield in each SR1 7.9 7.5 region SR2 7.5 6.1 • SR3 5.7 5.5 Convolute that with a given error on the signal (I used 305 everywhere) SR4 8.0 7.7 SR5 5.1 4.1 • Compute a 95% upper limit integrating the SR6 2.8 4.0 posterior SR7 2.8 2.0 • This gives the quoted right answer within one SR8 5.7 5.4 event (most probably due to binning effects) Tuesday, October 30, 12

  8. Comparison to official limit -1 CMS s = 7 TeV, L = 4.98 fb 2 int 10 x BR pb NLO+NLL ~ ~ ~ ~ ∼ ∼ 0 ± Model B2: g g b b , m( b ) = 500 GeV, m( ) = 200 GeV, m( ) = 50 GeV → χ χ 1 1 1 ~ ~ ~ ~ ∼ 0 Model A2: g g t t , m( t ) = 530 GeV, m( ) = 50 GeV → χ 10 1 1 1 ~ ~ ~ ~ ∼ 0 Model A2: g g t t , m( t ) = 280 GeV, m( ) = 50 GeV → χ 1 1 1 ~ ~ ∼ ∼ σ 0 0 Model A1: g g 4top + 2 , m( ) = 50 GeV → χ χ 1 1 1 Same Sign dileptons with btag selection -1 10 -2 10 -3 10 500 600 700 800 900 1000 1100 1200 ~ m( g ) GeV Tuesday, October 30, 12

  9. Comparison to official limit e -1 CMS, s = 7 TeV, L = 4.98 fb int 800 ) GeV Same Sign dileptons with btag selection prod NLO+NLL Exclusion = 1 σ σ ± σ 700 0 1 outreach limit ∼ χ m( 600 500 400 300 200 100 400 500 600 700 800 900 1000 1100 ~ m( g ) GeV Figure 6: Left plot: exclusion (95 % CL) in the There is some problem somewhere, but a factor-three of (unfortunately on the “wrong” side) is not too bad for 1 day of work Tuesday, October 30, 12

  10. Razor Tuesday, October 30, 12

  11. The Razor Frame • Two squarks decaying to quark and LSP . In their rest frames, they are two copies of the same monochromatic decay. In this y frame p(q) measures M Δ M 2 q − M 2 ˜ ˜ χ = 2 M ˜ M ∆ ≡ χ γ ∆ β ∆ , M ˜ q x • In the rest frame of the two incoming partons, the y two squarks recoil one against each other. • In the lab frame, the two squarks are boosted longitudinally. The LSPs x escape detection and the quarks are If we could see the LSPs, we could detected as two jets boost back by β L , β T , and β CM → β T In this frame, we would then get y |p j1 | = |p j2 | → β L Too many missing degrees of 11 freedom to do just this z Tuesday, October 30, 12

  12. The Razor Frame • In reality, the best we can do is to compensate the missing degrees of freedom with assumptions on the boost direction - The parton boost is forced to be RAZOR longitudinal CONDITION - The squark boost in the CM frame is |p Rj1 | = |p Rj2 | assumed to be transverse p Rj1 • β TCM We can then determine the two by requiring that the two jets p j1 p* j1 - β LR* have the same momentum after the transformation p* j2 p j2 • The transformed momentum defines the M R variable p Rj2 - β TCM q ( E j 1 + E j 2 ) 2 − ( p j 1 z + p j 2 z ) 2 , M R ≡ momentum p is determined from the massless 12 Tuesday, October 30, 12

  13. The Razor Variable • M R is boost invariant, even if defined from 3D momenta • No information on the MET is used M Δ a. u. • The peak of the M R distribution provides an estimate of M Δ 0 200 400 600 800 1000 1200 1400 1600 1800 2000 M [GeV] • R M Δ could be also estimated as the “edge” of M TR s a.u. ( p j 1 T + p j 2 p j 1 p j 2 M Δ T ) − ~ E miss E miss · ( ~ T + ~ T ) M R T T . T ≡ 2 0 200 400 600 800 1000 1200 • R M [GeV] M TR is defined using transverse quantities T and it is MET -related R ≡ M R • T The Razor (aka R) is defined as the ratio . M R of the two variables 13 Tuesday, October 30, 12

Recommend


More recommend