Документ взят из кэша поисковой машины. Адрес оригинального документа : http://qfthep.sinp.msu.ru/talks2013/HiggsTalk_FINAL4.pdf Дата изменения: Mon Jun 24 00:21:35 2013 Дата индексирования: Thu Feb 27 20:22:39 2014 Кодировка:

Review of ATLAS Higgs WW Results with 25 fb-1 of Data
Nikolina Ilic, University of Toronto QFTHEP 2013, St. Petersburg, Russia


Outline


Introduction Backgrounds


W+jets and W, W*, WZ, ZZ Z/* +jets Top WW



Signal Extraction



Gluon Gluon Fusion (ggf) Vector Boson Fusion (VBF)
7TeV+8TeV Combined VBF and couplings



Results





2

Spin Conclusions
Nikolina Ilic 2013-06-23


Introduction
Gluon Gluon Fusion (ggf)
Vector Boson Fusion (VBF)

Search for in 0/1 jet channel

Search for in 2 jet channel

Search in 4 channels


Different Flavour (DF): / + Same Flavour (SF) : / +



Split by jet multiplicity: 0, 1, 2jet
3 Nikolina Ilic 2013-06-23


Introduction
Gluon Gluon Fusion (ggf)
Vector Boson Fusion (VBF)

Search for in 0/1 jet channel

Search for in 2 jet channel

Search in 4 channels


Different Flavour (DF): / + Same Flavour (SF) : / +



Large BR over wide range of MH Poor mass resolution

Split by jet multiplicity: 0, 1, 2jet
4 Nikolina Ilic 2013-06-23


Introduction
Gluon Gluon Fusion (ggf) SM WW Vector Boson Fusion (VBF) Top

W+jets

W,WZ, ZZ

Z/*+jets

Search for in 0/1 jet channel

Search for in 2 jet channel

Search in 4 channels


Different Flavour (DF): / + Same Flavour (SF) : / +



Split by jet multiplicity: 0, 1, 2jet
5

Large BR over wide range of MH Poor mass resolution Lots of background to dig through!
Nikolina Ilic 2013-06-23


Backgrounds: W+jet/ W, WZ, ZZ


vv
-

W+jets


Small but kinematic distributions are similar to signal Hard to model in MC so estimated from data









W+jet

misidentified as lepton

6

Nikolina Ilic

2013-06-23


Backgrounds: W+jet/ W, WZ, ZZ


W+jets


Small but kinematic distributions are similar to signal Hard to model in MC so estimated from data





-





W+jet

misidentified as lepton + - /



W, W* WZ, ZZ Small and estimated from MC Reduced with 2 lepton selection Background estimations checked in same sign validation region
7






+





+

3 leptons
Nikolina Ilic 2013-06-23


Backgrounds: Z/* +jets
Big for DF
/
- +

Big for SF

+



-

/

-


+

0 jet

+



-







Estimated from CR

Estimated from Data (SF), MC (DF)



Background reduction depends on flavour channel and jet multiplicity





Reduce Reduce Reduce emerge Reduce
8

*: m >10 (12) GeV for DF (SF) Z: remove events around the Z mass peak Z/*: variables > 20- 45 GeV. In Z/* events, the leptons back to back and from neutrinos cancel Z/*: pT >30 for 0jet
Nikolina Ilic 2013-06-23


Backgrounds: What is left?


After W+jets, W , WZ, ZZ and Z/* + jets background is reduced, what is left? Z/*+jets in SF is reduced with cuts on hadronic recoil, and the less pileup sensitive Must reduce top and SM WW backgrounds

Different Flavour

Same Flavour

Jet multiplicity
9 Nikolina Ilic

Jet multiplicity
2013-06-23


Backgrounds: Top
Single (Wt)

+



+





+







-



2nd from b decay or misidentified jet

miss these jets



Modelled by MC, corrected to data from a control region which contain b tagged jets To reduce top background

10

N

b-jet

= 0 in 1 jet, pT (all objects) < 45 GeV in 2 jet
Nikolina Ilic 2013-06-23


Backgrounds: WW
Biggest bkg for 0 jet

-

2nd Biggest bkg for 1 jet

- -


+



+ -


+

+







Modelled by MC and corrected to data from CR for 0/1j, taken from MC (Sherpa+MC@NLO) for 2j

11

Nikolina Ilic

2013-06-23


Backgrounds: WW
Biggest bkg for 0 jet

-

2nd Biggest bkg for 1 jet

- -


+



+ -


+

+





Modelled by MC and corrected to data from CR for 0/1j, taken from MC (Sherpa+MC@NLO) for 2j Opening angle between the two leptons ( ) is small for signal since Higgs is spin 0 and there is spin correlation between the two W bosons SM H vs



12

Transverse Plane



SR CR
Nikolina Ilic 2013-06-23


Signal Extraction: ggF


After selections to reduce backgrounds, signal extracted Split SR to improve sensitivity (different Signal/Bkg ratio)

SR 1
13

SR 2
Nikolina Ilic 2013-06-23


Signal Extraction: ggF


After selections to reduce backgrounds, signal extracted Split SR to improve Final likelihood fit is to m sensitivity (different Signal/Bkg ratio)

T

SR 1
14

SR 2
Nikolina Ilic 2013-06-23


Signal Extraction: VBF



VBF signal is WW+2 forward jets with a large rapidity gap
pT (all objects) < 45 GeV to reduce dominant background mjj > 500 GeV, rapidity(Ylead jet, sublead jet) > 2.8 No additional jets between the two forward ones to reduce and ggF Require leptons to be between forward jets Define signal region: m < 60, | | <1.8 and fit mT in 4 bins

x2000

SR
15 Nikolina Ilic 2013-06-23


Results

e

jet


jet


16



Nikolina Ilic

2013-06-23


Results: 7TeV+8TeV Combined


At mH =140 GeV


Obser ved significance of 4.1



Study excess of events at mH = 125 GeV that was observed by H/ZZ


Expected Significance : 3.7 Obser ved Significance : 3.8

17

Nikolina Ilic

2013-06-23


Results: 7TeV+8TeV Combined


Study signal strength, =
= . ± .



, at mH = 125 GeV
B-tagging efficiency Jet energy scale/resolution

= . ± . . ± . . . ± . . . ± . .
Signal Yield ( ) WW bkg normalisation

18

Nikolina Ilic

2013-06-23


Results: VBF and Couplings


At mH =125 GeV (ggF is background)


Expected Significance : 1.6 Obser ved Significance : 2.5 , = . ± .

19

Nikolina Ilic

2013-06-23


Results: VBF and Couplings


At mH =125 GeV (ggF is background)


Expected Significance : 1.6 Obser ved Significance : 2.5 , = . ± .



Consistent with SM value of 1

20

Nikolina Ilic

2013-06-23


Spin









Try to differentiate between Higgs with spin 0+ and graviton-like spin 2+ particle Focus on Different Flavour, 0 jet channel and 2012 data Use a Boosted Decision Tree (BDT) analysis, train 4 variables: mT , , m , pT For spin 2+ the fraction of vs production is unknown so scan is performed over 5 different fractions ( ) Data are compatible with spin 0+ 2+ excluded at 99% CL, excluded at 95% CL
21 Nikolina Ilic 2013-06-23


Conclusions


The HWW channel is presented in the mass range of 115 GeV - 200GeV with the full 25fb -1 7TeV+8 TeV dataset in 4 flavour channels split by 3 jet multiplicities The largest significance of 4.1 occurs at mH =140 GeV
For mH = 125 GeV: = . ± . ,
,




= . ± .
+



Spin analysis favours spin 0+ Higgs, and rejects spin 2 with > 95% CL

22

Nikolina Ilic

2013-06-23


References






The Atlas Collaboration, Study of spin properties of the Higgs-like boson in the HWW channel with 21fb-1 of = 8 TeV data collected with the ATLAS detector, ATLAS-CONF-2013-31, CERN, Geneva, 2013 The Atlas Collaboration, Measurement of the properties of the Higgs-like boson in the WW decay channel with the ATLAS detector using 25 fb-1 of proton-proton collision data, ATLAS-CONF-2013-030, CERN, Geneva, 2013 J. Jovicevic, Probing the Standard Model Higgs boson in the WW decay mode with the ATLAS detector at the LHC, Licentiate Thesis, KTH Engineering Sciences, Stockholm, Sweden, 2013
23 Nikolina Ilic 2013-06-23


BACKUP

24

Nikolina Ilic

2013-06-23


Results: 7TeV+8TeV Combined




Previous 7 TeV analysis improved with 2012 updates, better WW Modelling (MC@NLO Powheg), improved lepton definition for W+jets reduction Event Yields for 0.75 mH < mT< mH (mT < 1.2 mH for 2j)
7 TeV
improved

Signal mH = 125 24±5 7±2 1.4±0.2

Bkg 161±11 47±6 4.6±0.8

Data 154 62 2

8 TeV Signal
0 jet 1 jet 2 jet 97±20 40±13

Bkg 739±39 261±28 36±4

Data 831 309 55

mH = 125

0 jet 1 jet 2 jet

10.6±1.4



Leading systematics are theoretical systematics on signal yields, jet energy scale/resolution and b-tagging efficiency
25 Nikolina Ilic 2013-06-23


W+jets Estimation



Estimated using data driven method since hard to model in MC Control Region: select 1 good lepton and 1 lepton does not satisfy strict selection requirements ( Denominator Lepton )
Calculated from dijet events. ~45% uncertainty due to differences between di-jet and wjets, pileup and real lepton contamination from W/Z



Includes QCD background estimation, when both leptons are misidentified as jets A validation region with same sign leptons containing mostly W+jets is used to validate this method
26 Nikolina Ilic 2013-06-23


Z/* Estimation


Z/*


Background normalised to data using a different flavour CR defined by m < 80GeV and ||>2.8 The 2 jet CR in addition to above requires Nb-jet= 0 and pTto <45 GeV

t



Z/*


Small in different flavour, so taken from MC In same flavour, Njet1, f recoil is used to define CR. For N jet 2 Etmiss - m in the Z peak is used to estimate DY since the statistics are to low and there are too many jets to use f recoil method
quadrant opp. pileup weight




27 Nikolina Ilic 2013-06-23


Top Estimation


MC Top normalised to data yields in CR after subtracting non-top backgrounds in 1j and 2 j



0j :



=



+

0

,

0

=

0

в

(

. 2 2 )

=





From btagged CR



1j:

28



=



, =



is from simulation

CR region is defined using same selection but with Nbjet=1 and requirements on and m removed = 1.04± .02(stat) for 1j, = 0.59 ± .07(stat) for 2jet
Nikolina Ilic 2013-06-23


Spin


Do 2 BDT trainings for 0+ and 2+ and perform 2D fit on BDT0 and BDT2

0

+

2

+

Background

29

Nikolina Ilic

2013-06-23


Systematics

30

Nikolina Ilic

2013-06-23


Full Selection

31

Nikolina Ilic

2013-06-23


Monte Carlo

32

Nikolina Ilic

2013-06-23