search for new phenomena in dijet events with the atlas
play

Search for new phenomena in dijet events with the ATLAS detector at - PowerPoint PPT Presentation

Search for new phenomena in dijet events with the ATLAS detector at s = 13 TeV Dengfeng Zhang(Tsinghua University) ILHC-ICTP2019, ICTP May 27-31, 2019 Outline Introduction to Dijet Analysis Dijet Resonance Analysis Dijet Angular


  1. Search for new phenomena in dijet events with the ATLAS detector at √ s = 13 TeV Dengfeng Zhang(Tsinghua University) ILHC-ICTP2019, ICTP May 27-31, 2019

  2. Outline • Introduction to Dijet Analysis • Dijet Resonance Analysis • Dijet Angular Analysis • Summary May 29, 2019 ILHC-ICTP2019@ICTP � 2

  3. Introduction • pp collisions at √ s = 13 TeV, providing a wide scope to search for new phenomena at ATLAS; • Final states including partons dominate in some BSM models; • Total dijet production rates for BSM signals can be large; • Two complementary analysis: resonance analysis based on m jj and angular analysis base on 𝜓 . χ = e 2| y *| , y * = ( y 1 − y 2 ) / 2 May 29, 2019 ILHC-ICTP2019@ICTP � 3

  4. Dijet Event Display Recorded highest mass dijet event: leading/sub-leading jet p T =3.74 TeV, |y*|=0.38, m jj =8.02 TeV. May 29, 2019 ILHC-ICTP2019@ICTP � 4

  5. Resonance Analysis

  6. Overview of Resonance Analysis • In SM, hadron collisions produce jet pairs primarily via 2 → 2 parton scattering processes governed by QCD; • QCD predicts a smoothly falling dijet invariant mass distribution; • New particles decaying to two jets may introduce local excesses. • Sensitive to resonant signals. • Benchmark Model: q *, Z ʹ , W ʹ , W*, QBH, etc. • Three presented results: 15.7 fb -1 , 37 fb -1 and 139 fb -1 in Run 2. May 29, 2019 ILHC-ICTP2019@ICTP � 6

  7. Event Selection and Background Estimation: 15.7 fb -1 Event Selection: Events ATLAS Preliminary 6 10 Data • GRL Fit 5 10 Statistical uncertainty on fit • LAr, Tile, SCT error rejected Function choice 4 10 • Core: Incomplete event rejected 3 10 2 • PV has at least two tracks 10 -1 s =13 TeV, 15.7 fb p -value = 0.67 10 • Pass HLT_j380 Fit Range: 1.1 - 7.1 TeV |y*| < 0.6 1 • ≥ 2 clean jets, Leading jet p T > 440 GeV Sub- Rel. Uncert. 0.4 0.2 leading jet p T >60GeV 0 − 0.2 − 0.4 • |y*| = |y 1 -y 2 |/2 < 0.6(1.2 for W*) 2 3 4 5 6 7 m [TeV] ATLAS-CONF-2016-069 jj • m jj > 1100 GeV(1717 GeV for W*) Global fitting with 3-parameters function on the m jj spectrum to estimate the background directly: f ( x ) = p 1 (1 − x ) p 2 x p 3 x = m jj / s May 29, 2019 ILHC-ICTP2019@ICTP � 7

  8. SearchPhase Results: 15.7 fb -1 • BumpHunter Algorithm is employed to search for local excess over the background. • No significant local excess. Events Events 6 ATLAS Preliminary ATLAS Preliminary 10 6 10 -1 -1 s =13 TeV, 15.7 fb s =13 TeV, 15.7 fb Data Data 5 10 Background fit Background fit 5 10 BumpHunter interval BumpHunter interval φ q *, m = 4.0 TeV W*(sin( )=0), m = 2.8 TeV 4 10 q * x W* 4 φ 10 q *, m = 5.0 TeV W*(sin( )=0), m = 3.8 TeV q * x W* 3 10 3 10 p-value: 0.67 p-value: 0.83 2 10 2 10 φ σ × σ × W*(sin( )=0), 30 q *, 3 10 10 x p -value = 0.67 p -value = 0.83 Fit Range: 1.1 - 7.1 TeV Fit Range: 1.7 - 7.6 TeV 1 1 |y*| < 0.6 |y*| < 1.2 Significance Significance 2 2 0 0 − − 2 2 2 3 4 5 6 7 8 2 3 4 5 6 7 8 m [TeV] m [TeV] jj jj ATLAS-CONF-2016-069 May 29, 2019 ILHC-ICTP2019@ICTP � 8

  9. Limit Setting: 15.7 fb -1 Bayesian method to set upper limits at 95% C.L. on Acceptance*Xs*Br. 1 BR [pb] BR [pb] ATLAS Preliminary ATLAS Preliminary -1 -1 s =13 TeV, 15.7 fb s =13 TeV, 15.7 fb 1 |y*| < 0.6 |y*| < 1.2 × × A A − × × 1 10 σ σ − 1 10 ATLAS-CONF-2016-069 − 2 10 φ q * W*(sin( )=0) x − 3 10 Observed 95% CL upper limit Observed 95% CL upper limit − 2 Expected 95% CL upper limit Expected 95% CL upper limit 10 68% and 95% bands 68% and 95% bands − 4 10 2 4 6 2 2.5 3 3.5 4 M [TeV] M [TeV] q * W* 95% CL exclusion limit 2012@8 TeV Model Observed Expected Observed Expected Quantum Black Hole 8.7 TeV 8.7 TeV 5.66 TeV 5.66 TeV Excited quark 5.6 TeV 5.5 TeV 4.06 TeV 3.98 TeV W ′ 2.9 TeV 3.3 TeV 2.45 TeV 2.51 TeV W ∗ 3.3 TeV 3.3 TeV 1.75 TeV 1.95 TeV May 29, 2019 ILHC-ICTP2019@ICTP � 9

  10. Event Selection and Background Estimation: 37 fb -1 Event Selection: Events / Bin Same with last publication. ATLAS 7 10 • GRL Data 6 10 Fit • LAr, Tile, SCT error rejected Statistical uncertainty in fit 5 10 Function choice 4 • Core: Incomplete event rejected 10 3 10 • PV has at least two tracks 2 10 • Pass HLT_j380 -1 s =13 TeV, 37.0 fb 10 Fit Range: 1.1 - 8.2 TeV • ≥ 2 clean jets, Leading jet p T > 440 GeV Sub-leading 1 |y*| < 0.6 jet p T >60GeV − 1 10 Rel. Uncert. 0.4 • |y*| = |y 1 -y 2 |/2 < 0.6(1.2 for W*) 0.2 0 − 0.2 − 0.4 • m jj > 1100 GeV(1717 GeV for W*) 2 3 4 5 6 7 8 9 Phys. Rev. D 96 (2017) 052004 m [TeV] jj Sliding Window Fitting Method(SWiFt): • Slide over mass spectrum into smaller windows; • Perform fitting in each window; f ( x ) = p 1 (1 − x ) p 2 x p 3 , x = m jj / s • Stitch background fit value in each bin together for the full range prediction. May 29, 2019 ILHC-ICTP2019@ICTP � 10

  11. SearchPhase Results: 37 fb -1 • BumpHunter Algorithm is employed to search for local excess over the background. • No significant local excess. Events / Bin Events / Bin ATLAS ATLAS 7 7 10 -1 10 -1 s =13 TeV, 37.0 fb s =13 TeV, 37.0 fb Data Data 6 6 10 10 Background fit Background fit BumpHunter interval BumpHunter interval 5 5 10 10 φ q *, m = 4.0 TeV W*(sin( )=0), m = 2.8 TeV x q * W* φ q *, m = 5.0 TeV W*(sin( )=0), m = 3.8 TeV q * x 4 4 W* 10 10 3 3 10 10 p-value: 0.63 p-value: 0.83 2 2 10 10 φ σ × σ × W*(sin( )=0), 100 q *, 10 10 10 x p -value = 0.63 p -value = 0.83 1 1 Fit Range: 1.1 - 8.2 TeV Fit Range: 1.7 - 8.1 TeV |y*| < 0.6 |y*| < 1.2 − − 1 1 10 10 Significance Significance 2 2 0 0 − − 2 2 0.5 0.5 Data-MC Data-MC JES Uncertainty JES Uncertainty MC MC 0 0 − − 0.5 0.5 2 3 4 5 6 7 8 9 2 3 4 5 6 7 8 m [TeV] m [TeV] jj jj Phys. Rev. D 96 (2017) 052004 May 29, 2019 ILHC-ICTP2019@ICTP � 11

  12. SearchPhase Results: 37 fb -1 Bayesian method to set upper limits at 95% C.L. on Acceptance*Xs*Br. BR [pb] BR [pb] ATLAS ATLAS 1 -1 -1 s =13 TeV, 37.0 fb s =13 TeV, 37.0 fb 1 |y*| < 0.6 |y*| < 1.2 × × A A − × × 1 10 − 1 10 σ σ Phys. Rev. D 96 (2017) 052004 − 2 10 − 2 10 φ q * W*(sin( )=0) x − 3 10 Observed 95% CL upper limit Observed 95% CL upper limit Expected 95% CL upper limit Expected 95% CL upper limit ± σ ± σ ± σ ± σ Expected 1 and 2 Expected 1 and 2 − − 3 4 10 10 2 4 6 2 3 4 5 m [TeV] m [TeV] q * W* 95% CL exclusion limit 2016 Model Observed Expected Observed Expected Quantum Black Hole 8.9 TeV 8.9 TeV 8.7 TeV 8.7 TeV Excited quark 6.0 TeV 5.8 TeV 5.6 TeV 5.5 TeV W ′ 3.6 TeV 3.7 TeV 2.9 TeV 3.3 TeV W ∗ 3.4 TeV 3.6 TeV 3.3 TeV 3.3 TeV May 29, 2019 ILHC-ICTP2019@ICTP � 12

  13. Latest Resonance Analysis: 139 fb -1 Event Selection: Full Run2 Data Events ATLAS Preliminary • Good Run List (GRL) 8 10 -1 s =13 TeV, 139 fb 7 10 Data • LAr, Tile, SCT error rejected, 6 Background fit 10 BumpHunter interval 5 10 q *, m = 4.0 TeV • Core: Incomplete event rejected, q * q *, m = 5.0 TeV 4 10 q * 3 • PV has at least two tracks, 10 p-value: 0.8 2 10 • Pass HLT_j420, σ × q *, 0.1 10 p -value = 0.8 1 Fit Range: 1.1 - 8.1 TeV • ≥ 2 clean jets, Leading jet p T > 420 GeV, Sub- |y*| < 0.6 − 1 10 leading jet p T > 150 GeV, Significance 2 4 6 8 2 0 • |y*|=|y 1 -y 2 |/2<0.6 − 2 2 3 4 5 6 7 8 • |m jj |>1100 GeV m [TeV] ATLAS-CONF-2019-007 jj • Sliding Window Fitting Method(SWiFt) is still robust: f ( x ) = p 1 (1 − x ) p 2 x p 3 + p 4 lnx , x = m jj / s • BumpHunter Algorithm is used to search for local excess over the background. • No significant local excess. May 29, 2019 ILHC-ICTP2019@ICTP � 13

  14. Latest Resonance Analysis: 139 fb -1 • CL s technique implemented in HistFitter framework to set upper limits at C.L. of 95%. BR [pb] Theory 1 Observed 95% CL Expected 95% CL ± σ × 1 ± σ A 2 − 1 10 × σ M q* <6.7 TeV − 2 10 − 3 10 -1 s = 13 TeV, 139 fb − 4 10 ATLAS Preliminary 2 3 4 5 6 7 8 m [TeV] q* ATLAS-CONF-2019-007 More results of other channels and Dib-jet analysis are coming soon for one paper. May 29, 2019 ILHC-ICTP2019@ICTP � 14

  15. Angular Analysis

  16. Overview of Angular Analysis • In SM, t-channel dominates the parton scattering process, most dijet productions occur at small angles, differential cross section tends to be flat; • BSM predicts additional dijet production at large angles. χ = e 2| y *| , y * = ( y 1 − y 2 ) / 2 2 ln( E + p z y = 1 ) E − p z • Sensitive to resonant/non-resonant signals; • Benchmark Models: CI, QBH, etc. • Two published results: 15.7 fb -1 , 37 fb -1. May 29, 2019 ILHC-ICTP2019@ICTP � 16

Recommend


More recommend