Center for Cosmology and Particle Physics

Combined ATLAS Standard Model Higgs Search with 1 fb-1 of Data at 7 TeV

Kyle Cranmer, New York University

on behalf of the ATLAS Collaboration
Kyle Cranmer (NYU) EPS-HEP 2011 - ATLAS Higgs Combination 1

Introduction
Understanding of electroweak symmetry breaking is a major goal of the LHC physics program

Center for Cosmology and Particle Physics

V () = 2 ||2 + ||4

Initial focus: search for the Standard Model Higgs

drove the design of both the ATLAS and CMS detectors stresses every major sub-system
Goldstone modes

ATLAS and LHC are running great!

Analyses here are based on up to 1.2 High pile-up environment

New Heavy Gauge Bosons: Lepton+ETMiss Dileptons Compositeness: dimuons Extra Dimensions Same-sign dimuons (ADD) Dileptons (RS) Diphotons (RS) Diphotons+ETMiss (UED) Leptoquarks Highly Ionising Particles

fb-1

(become longitudinal polarizations of massive W,Z)

vacuum expectation fluctuations

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

Channels Included in the Combination

Enhance sensitivity by combining all available searches channels in the context of the SM Higgs hypothesis

Branching ratios
LHC HIGGS XS WG "(pp ! H+X) [pb] 2010

Center for Cosmology and Particle Physics

WW ZZ
pp ! H (N

bb

NLO

10

+NN

s= 7 TeV
LL Q CD +N LO EW )

10-1 !! cc

gg

pp !

pp

pp ZH

qqH

(NNLO

QCD

(N

+ NLO

NL

(N N

EW)

pp

QC

LO +N

QC LO

ttH

+N )

some channels are composed of subchannels and include control samples

Channel WH ZH 0-jet 1-jet 0-jet 1-jet veto IP 2 veto 2 2 0 1 0 1 >30 >30 >30 >30 >30 <50 MWW MWW MT Mllqq M4l btag (veto) Jets MET (GeV) Shape Mgg Mbb Mbb

10-2 "" 10-3 100 120 Z"

10-1

(N

LO

QC

EW

LO

EW

D)

10-2

140

160

180

200

100

200

300

400 500

MH [GeV]
Fig. 41: The SM Higgs production cross section at

1000 MH [GeV]

s = 7 TeV.

Mass Range
#H [GeV]

Fig. 35: SM Higgs branching ratios as a function of the Higgs-boson mass.

102

pp !

H (N

NLO

+N

s= 14 TeV
NLL QC D +N LO EW )

110-150 10 110-130
10
-1

(from sidebands)
10
pp ! qqH (NNLO

Top (3j - high Mbb) and W+jets (low Mbb)

pp !

pp

QCD

+ NLO

ZH

W H
O

EW)

(N

(N

NL

NL
D

pp !

QC

QC

ttH

+N

10 110-130
-2

Z+jets (low Mbb)

QC D)

(N

LO

LO

+ NL O EW EW ) )

100

200

300

500

1000 MH [GeV]

10-1

WW (ll) WW (lqq)

110-240 110-240 200-600 200-600 200-600 200-600 110-600

82

200 300 100 WW (control region Mll) 400 500

Fig. 42: The SM Higgs production cross section at

1000 MH [GeV]

Fig. 36: SM Higgs total width as a function of the Higgs-boson mass.

s = 14 TeV.

Top (from reverse btag) and WW (Mll CR)

114

W+jets (sidebands) W+jets (sidebands) VV(from MC) and top (MC and checks) Z+jets (from MC) and top (from MC) ZZ (from MC), Z+jets (MC) and top (CR)
3

ZZ (ll) ZZ (llqq) ZZ (4l)

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

LHC HIGGS XS WG 2010

(GeV/c2)

LHC HIGGS "(pp ! H+X) [pb] XS WG 2010

103

Main backgrounds
102

LHC HIGGS XS WG 2010

Detailed presentations on the channels

Center for Cosmology and Particle Physics

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

Low19, 2011 13channels mass : 20 July

H (110-150)
Events / 2 GeV
400 350 300
DRAFT DRAFT

DRAFT
)/SM @ 95% CL
102

Center for Cosmology and Particle Physics

Events / 5 GeV

Data 2011, s = 7 TeV, Ldt = 1.08 fb ATLAS Preliminary

Data

-1

Observed CLs limit Expected CLs limit July 20, 2011 16 : 39 DRAFT 8 ATLAS Preliminary 1 -1 Data 2011, s = 7 TeV, Ldt = 1.1 fb s = 7 TeV Data 2011, 2 7

ATLAS Preliminary 6 10

Ldt = Data fb 1.08

-1

BR(H

uly 19, 2011 13 : 20 2011 13 : 20

MC mH=120 GeV, 5xSM Total background (Fit)

5 4 3

mH=210 GeV, 1xSM

7
Total background

250
Events / 2 GeV
400 Data 2011, s 7 TeV, Ldt -1 Data 2011, s = 7 TeV, = Ldt = 1.08 fb = 1.08 fb
Data

Not reviewed, for internal circulation only

350 ATLAS Preliminary 150 ATLAS Preliminary Data 300 250 200 150 100

Events / 5 GeV

-1

8 7 6 5 4

Events / 5 GeV

200

s 7 TeV, fb-1 Data 7 2011, s = 7 TeV, = Ldt = 1.1 Ldt = 1.1 fb

H Data 2011,

-1

HZZ 4l

(*)

ATLAS Preliminary 6 ATLAS Preliminary

5 4 3 2 1

Data

Data

1
2 1
95% C.L. limit on / SM

100 50 100

MC mH 5xSM MC mH=120 GeV,=120 GeV, 5xSM Total background (Fit) Total background (Fit)

m =210 GeV, 1xSM mH=210 GeV,H1xSM Total background Total background

(*)

110
100
120

115
200

120
300

125

130
400

135
500

140

145

150

110

120

3 130 2 1

140

150

(*) H HZZ 4l ZZ 4l

160

m [GeV]

Events / 10 GeV

100

Events / 10 GeV

0 100

Events / 10 GeV

ATLAS Preliminary -1 250 Data 2011, s 7 TeV, Data 2011, s = 7 TeV, = Ldt = 1.1 Ldt = 1.1 fb fb
ATLAS Preliminary ATLAS Preliminary 200 150 100 50 0

200
150 100

Events / 10 GeV

-1

60 50 40 30 20 10

Events / 10 GeV

WH and ZH, Hbb (110-130) 250 Data Ldt = 1.1 fb

110 120 120 130 130 140 150 160 160 m s = 7 [GeV] 2011, [GeV]m TeV, 140 150

50 100 110

VH, H bb 600 60 100 -1 m4l [GeV]m4l [GeV] Data 2011, s = 7 TeV, Ldt = ATLAS Preliminary 1.1 fb 60 50 80 -1 ATLAS Preliminary Data 2011, s 7 TeV, fb-1 Data = 7 TeV, = Ldt = 1.1 Ldt = 1.1 fb Data Data 2011, s
0 100 200 200 300 300 400 400 500 500 600
-1

Observed (CLs) Expected (CLs) 1 2

Ldt=1.04 fbm,
-1

MH [GeV] 600
s=7 TeV [GeV]

4l

Data

Data

50 ATLAS Preliminary ATLAS Preliminary 40 30 20 10 0 50 100 100 150

mH=120 GeV, 20xSM Data Total background

Data

40 30 20

m =120 GeV, 20xSM mH=120 GeV,H20xSM Total background Total background

m =120 GeV, 20xSM mH=120 GeV,H20xSM Total background Total background

60 40

mH=120 GeV, 20xSM Total background

WH, WH, Hbb Hbb

WH, Hbb

b ZH, HbZH, Hbb

ZH, Hbb
20

50

50 100

100 150

50

150 200

200 250

250

50

150 200

200 250

10 250
110 115 120

m mbb [GeV] bb [GeV]

m mbb [GeV] bb [GeV]

10 GeV

30 -1 Data 2011, s 7 TeV, Ldt -1 Data 2011, s = 7 TeV, = Ldt = 1.04 fb = 1.04 fb

Kyle Cranmer (NYU)

10 GeV

0 GeV

50

100

18 150 16

18

16 Data 2011, s 7 TeV, Ldt -1 Data 2011, s = 7 TeV, = Ldt = 1.04 fb = 1.04 fb

EPS-HEPm [GeV] 2011 ATLAS Higgs Combination

-1

200

250

50

100

150

200

125 130 Higgs mass [GeV]

250

[GeV]

The H WW ll Channels
250 m4l [GeV] mbb [GeV]

0 100 0 50

200 100

300

400 150

500 200

600

50

100

150

Limit200
-1

250Center for

mbb [GeV] Particle Physics

Cosmology and

Events / 1010 GeV Events / GeV

60 30 50 25 40 20 30 15 20 10 10 5

Data 2011, s = 7 TeV, Ldt = 1.1 fb -1 Data 2011, s = 7 TeV, Ldt = 1.04 fb ATLAS Preliminary Data ATLAS Preliminary
mH=120 GeV, 20xSM

95% CL Limit on /!/!SM Events 10 GeV

-1

18

16 102 14 12 10

ATLAS Preliminary Data 2011, s = 7 TeV, Ldt = 1.04 fb

ATLAS Preliminary Observed
Data

H$WW $l#l# Ldt = 1.04 fb s = 7 TeV

-1

(*)

Data

DRAFT

Total background

mH=150 GeV, 1xSM

ZH, Hbb
Total background

H WWl 0j ckground Estimates for H+0l +jet

106
4 2

Expected 11 1! 2!

mH=150 GeV, 1xSM Total background

"

HWWll + 1j

miss 0 tons, high ET,rel , and no jets with pT 200 25 GeV and || < 04.5 are considered > 50 100 150 250 0 50 100 150 200 250 50 100 150 mbb [GeV] required to satisfy the following additional cuts: 1
mT [GeV]

200

250 mT [GeV]

tum of the dilepton system, |PT |, is required to be at least 30 GeV. -1

16 Data 2011, s = 7 TeV, Ldt = 1.04 fb

Events / 10 GeV

18

mT [GeV]

Not reviewed, for internal circulation only

mass is The invariant mass <Data GeV the < 65 GeV for predicted 50 Figure 14 required to satisfy mdistribution foror m candidate events selected, the total background and 1: ATLAS Preliminary 120 140 160 180 200 DRAFT 220 240 12 July 20, the regions mH < 170 GeV and m =150 170() he signal expected in the H , mH GeV, 1xSM respectively.(l bb) and ZH ( + 2011b18 : 59 and the H M [GeV] H ZZ GeV, the W H b), H WW ()10 channels. For the HTotal background channel the signal is multiplied by factor of 5 and is illustrated he a Higgs boson mass tend to emerge from120 interaction point bbsimilar a associated ATLAS Preliminary 8 or Higgs boson decay hypothesis of mH = the GeV. For the H inexcess >2 production channels, HWW ll 10 Observed Broad H WWl + in correlation present in the WWsystem due to the GeV,zero the multilicative factor is 20. For the Ldt = 1.04 fb 6 boson mass hypothesis is also l m 1j = 120 spin but nature of the Expected he Higgs H 126<mH<158 GeV/c2 8 1 pton ZZ () in the transverse 95% Exclusion: 158-186 GeV s = 7 TeV 4 H opening anglethe Higgs boson plane, , is is mH = 210 GeV and no multiplicative factor is mass hypothesis required to be less than 2 6 H < 170 For the H 170 GeV). channel the Higs boson mass hypothesis is mH = 150 GeV and pplied.2 GeV (mH WW () 0 4 o multiplicative factor is applied. 200 50 150 mT , is required to100 satisfy 0.75L mH < mT 250 mH , where transverse mass is < IMIT 37.
H

Statistical significance

(*)

-1

0 120 140 160 180 200 220 240 mT = (ET + ET ) (PT + Pmiss )2 , (2) T M [GeV] The combination selected, the total background and not only useful to optimally take advantage of the of Higgs boson search channels is andidate events Figure 15: The expected and observed signal signicances for various Higgs boson 2 , Cranmer (NYU) and P + 6 Kyle |Pmiss | discrimination of is the transversethe background Higgs Combination integrated luminosity of 1.04 fb butTeV. The observed signicances at n EPS-HEP 2011 - ATLAS thevarious independent channels, at 7 miss + mstatistical = E ull the signal from momentum of in dilepton with an
H
1

Systematic Uncertainties miss 2

50 40 30 20 10

Total background

The High Mass Channels

HZZ llqq
(*)

Center for Cosmology and Particle Physics

uly 19, 2011 13 : 20 0

200 300

H WW lqq
400 500 600 700

DRAFT
900

800

H ZZ llqq
90 80 70 60 50 40 30 20 10 0 200 300 400 500 600 700 800 900 HZZ llqq
(*)

8 19, 2011 13 : 20 July

H ZZ ll

-1

DRA

mlljj [GeV]
Events / 50 GeV Events / 25 GeV

35
3

-1 Data 2011, s = 7 TeV, Ldt = 1.04 fb Data 2011, s = 7 TeV, Ldt = 1.1 fb

10

ATLAS Preliminary

-1

Data 2011, s = 7 TeV, Ldt = 1.04 fb ATLAS Preliminary

Data

Events / 50 GeV

104 40

-1

40 35 30 25 20 15 10 5 0 0

Data 2011, s = 7 TeV, Ldt = 1.04 fb ATLAS Preliminary

Data

30 25 20 15 10

ATLAS Preliminary Data

DatamH=300 GeV, 1xSM mH=400 GeV, 50xSM Total background Total background

mH=300 GeV, 5xSM Total background

mH=300 GeV, 1xSM Total background

10

10

HZZ ll HWWlqq

(*)

HZZ ll

(*)

1 5

0 0 0200 100 200 300 1000 1200 1400 1600 1800 2000 400 600 800 400 500 600 700 800 900 m mT [GeV] [GeV]
WW

100

200

300 400

500

600

700

800 900 mT [GeV]

mlljj [GeV]

Events / 25 GeV 95% C.L. limit on / (SM)

95% C.L. limit Events SM GeV on / / 25

4.5

Data 2011, s = 7 TeV, Ldt = 1.04 fb

4 date eventsATLAS Preliminarytotal background and the selected, the Data 3.5 the tagged and untagged categories, and qq in 3 10 2.5 2
mH=300 GeV, 5xSM Total background
(*)

Events / 25 GeV 95% C.L. limit on / SM

104

-1

Data 2011, s = 7 TeV, Ldt = 1.1 fb

-1

4.5

Data 2011, s = 7 TeV, Ldt = 1.04 fb

25 10
3

102

20 15

Observed Expected 1 2

ATLAS Preliminary Ldt=1.04fb , s=7TeV H ZZ llqq

mH=400 GeV, 50xSM Total background Data
-1

ATLAS Preliminary

4 14 ATLAS Preliminary Observed 3.5 12 Expected 3 1 10 2.5 2 2 8

-1

Data ATLAS Preliminary

H Ldt=1.04fb , s=7TeV

-1 m =300 GeV, 5xSM

H ZZ Total background ll
(*)

H 10 HZZ llbb on of 1.5 ATLAS Preliminary systematic uncertainties between ZZ llbb channels. 1.5 HWWlqq 1 6 tructed 1 objects in the detector (electrons, photons, Expected 10 1 1 0.5 and on the s=7 TeV Monte Carlo simulation, most same 1 0.5 4 0Since in most channels the2backgrounds 50 200 400 600 800 1000 1200 1400 1600 1800 2000 elated. -1 200 300 400 500 600 700 800 900 0 -1 200 300 400 500 600 700 800 900 2 L dt=1.04 fb 10 Observed [GeV] mlljj nd-related systematic uncertainties are typically mWW [GeV] mlljj [GeV] 0 0 250 300 350 400 450 500 550 600 200 250 300 350 400 450 500 550 600 n the luminosity, which is fully correlatedM [GeV] among 200 250 300 350 400 450 500 550 600 H mH [GeV] y the signal estimates, except in channels where mH [GeV] in MonteThe invariant mass such as the H ZZ igure 2: Carlo simulation, distribution for the candidate events selected, the total background and the Figure 2: The invariant mass distribution for the can ignal expected in the H ZZ , the H ZZ qq in the tagged channelscategories, and and untagged able to exclude high-mass SM , the H ZZ signal expected in the H ZZ HZZ Higgs H WW qq. H WW qq.

EPS-HEP 2011 - ATLAS Higgs Combination lsoKyle Cranmer (NYU) be taken of the correlation of systematic uncertainties between channels. allows proper account to

DRAFT

The H ZZ 4l Channel (110 < MH < 600)

Events / 5 GeV
8 7 6 5 4 3 2 1 HZZ 4l
(*)

Center for Cosmology and Particle Physics

Data 2011, s = 7 TeV, Ldt = 1.1 fb ATLAS Preliminary

95% C.L. limit on /

-1

-1

SM

10

Data mH=210 GeV, 1xSM Total background

ATLAS Preliminary (*) H ZZ 4l -1 Ldt = 1.1 fb s=7 TeV

Observed CLs Expected CLs 1 2

SM

10

1
100 200 300 400 500 600 m4l [GeV]

160

GeV]

150 200 250 300 350 400 450 500 550 600 mH [GeV]

Events / 10 GeV

60 50 40 30 20 10

Data 2011, s = 7 TeV, Ldt = 1.1 fb ATLAS Preliminary

-1

Data mH=120 GeV, 20xSM Total background

SM

ZH, Hbb

250

50

100

150

200

250 mbb [GeV]

GeV]

Events / 10 GeV

18 16 14 12 Data 2011, s = 7 TeV, Ldt = 1.04 fb ATLAS Preliminary

Data mH=150 GeV, 1xSM

-1

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

95% SM 95% CL limit on !/ !CL Limit on !/ !SM

in most of the mass range, excepted in the very high mass range as described in Appendix A. interpretation in terms of limits is then given both using the CLs method [53] and in the Power ned Limits [10, 54] (PCL). The former should be taken as the actual result. In contrast to the of Ref. [10], the uncertainties on the cross section are included in the prole likelihood con. limits set by the individual channels used for the combination are shown in Fig. 3. Only the W () channel has a median result which is sensitive to the Standard Model Higgs boson, H ZZ excludes a larger mass range. The relative importance of the different channels ther they bring excesses or decits is illustrated in this gure.

Summary of individual channels limits

The situation prior to the combination

Center for Cosmology and Particle Physics

Two individual channels are able to exclude by themselves

ATLAS collaboration: Search for the Standard Model Higgs Boson

10

ATLAS w/ 35 pb-1
ATLAS
Observed, PCL Expected, PCL ATLAS collaboration: Search for the Standard Model Higgs Boson -1

Observed, before PCL 1!, PCL

95% CL limit on / SM

14

Exp. Obs.

Exp. Obs.

[arXiv:1106.2748] "
L dt ~ 35 pb

s=7 TeV

+ 2!, PCL

HH H WW l l l H WW l W,ZH bb bb W,ZH

H ZZZZ llll H llll H ZZZZ llqq H llqq H ZZZZ ll H ll H WW WW l qq H l qq

10
102

ATLAS

% L dt ~ 35 pb ,
-1

s=7 TeV

10

1 10 100
1

200

PCL limits 300

400

500

600

mH [GeV]

Expected Observed Expected Observed Fig. 10. The expected and observed upper limits on the total cross section dividedZZ#Hthe # llll by llll ZZexpected Standard Mo H#" " H#" " H# # boson cross section. This is a 95% PCL # WW## lWWgreen and yellow bands indicate ZZ#Hrange llqq which the limit is H# the llqqZZ# in # H limit.H The # l$l$ l$ $ H dottedl WW marks # qq H# ZZ#H $$ , # ll$ llsb to lie in the absence of a signal. The ne# WW H#$line# l$qq the results obtained using CL# ZZand $the application of t constraint gives the solid100 The limits are calculated at the masses marked with symbols and the lines between t line. 200 300 400 500 600 are to guide the eye.

mH [GeV]

ATLAS Preliminary 100

200

L dt ~ 1.0-1.2 fb , s=7 TeV

-1

Fig. 8. The expected and observed cross section limits, normalized to the Standard Model cross section, as a functio Observed, PCL Higgs boson mass for the individualATLAS search channels. The visually most apparent dierence between expected and ob Expected, PCL in the H W W qq channel, which has a decit approaching one sigma both at 320 GeV and 480 GeV. These re 2 the prole likelihood10 method with a power constraint (PCL). TeV limits are calculated at the masses marked with 1!, PCL s=7 The The lines between the points are to guide the eye. The grey horizontal-1 bands show thePCL + 2!, uncertainty on the Standard Mo dt ~ 35 pb section prediction, with the inner region highlighting theLcontribution of QCD Observed, CL scale uncertainties.

95% CL ! & BR (pb) Limit on !/ !SM

1 CLs limits 500

"

Expected, CL

300

400

600

10 102
10

ATLAS

mH [GeV]

1 1 100 10-1
-2 S.M. Expect. Observed

Tevatron exclusion LEP exclusion

400 500 600 s=7 TeV : The expected (dashed) and observed (solid) cross-section limits for the individual search chanm [GeV] H# WW (l$l$) % L dt ~ expected, malized to the Standard Model Higgs boson cross section, as functions of the Higgs boson mass. Fig. 10, except that limits calculated using the CL Hprocedure$qq) added. As35 pb when the # WW (l are Fig. 11. Same as 10 are its Limits sults use Cranmer (NYU) the prole likelihood technique with 95%EPS-HEPusing theATLAS Higgslimits uctuatene dottedmethods converge, but downwardusing CL ,HandZZthe application PCL 95% CL constraint CL limits 2011 - CLS construction. The up, both line marks the results obtained uctuations# less pronouncedofwith CL due to 9 lar coverage. the power Kyle Combination
H-1
s s sb

200

300 H#" "

solid line. The limits are -3 10 calculated at the masses marked with symbols. The lines between the points are to guide the

Systematics

Center for Cosmology and Particle Physics

The channels utilize three main strategies for addressing systematic uncertainties: data-driven techniques: eliminates dominant impact of uncertainty leaving uncertainties associated with extrapolation from control to signal region

implicit parametrization: parametrized functions flexible enough to describe effect of uncertainty on distribution (eg. exponential + Crystal ball in H) explicit parametrization: variational histograms obtained from modifying July 22, 2011 13 : 48 DRAFT simulated samples according to variations the source of uncertainty (eg. HZZ)

10

Table 2: Main correlated log-normal distributions Uncertainties in normalization described bysystematic uncertainties used in the analysis. These relative uncertainties (%)

The combination requires subject to theoretical uncertainties. Prescription agreed upon by LHC-HCG & Higgs () cross-section working group. H ZZ () H WW
"(pp ! H+X) [pb]
pp !
LHC HIGGS XS WG 2010

correspond to the overall effect on the per-event signal efciency of the 1 variation of the source of systematic between rates of different channels, thus it is relationsuncertainty. Some systematic errors, such as energy scale in the H search, are included but are not apparent in this table as they do not affect event rates. H 3.7 +11.6 10.4

H bb 3.7 2.3
+1.5 1.6 +2.1 1.5 +1.1 2.0 +21 17

H (N NLO

10

+NN

s= 7 TeV
LL Q CD +N LO EW )

Luminosity e/ efciency e/ energy scale e/ resolution efciency resolution Jet/MET energy scale Jet resolution MET b-tag efciency Theory

3.7 1.4
+0.1 0.4 +0.0 0.5 +0.6 0.6 +4.2 4.5 +4.6 7.9

qq 3.7 0.3 +15 18 +8.2 9.0 +0.9 0.8

3.7 1.9 1.2 -

3.7 1.2
+0.8 1.1

qq 3.7 1.1 0.6

pp !

pp

pp ZH

qqH

(NNL

O QC

(N

NL

(N

D+N LO E

NL D

W)

pp

QC

QC O

ttH

10-1

(N

+N L

+N )

LO

QC

EW

LO

EW

D)

5.8 2.5
+5.5 6.1 +37 33

+0.8 0.7 +5.9 4.0

10-2 100 200 300 400 500 1000 MH [GeV]

+15.0 20.0

+15.0 20.0

+15.0 20.0

+15.0 20.0

+6.6 4.2 +4.3 4.4 +15.0 20.0

+3.7 10.4 +2.1 0.0

+15.0 20.0

KyleFig. 41: The SM Higgs production cross section at Cranmer (NYU)

s = 7 TeV.

EPS-HEP 2011 - ATLAS Higgs Combination

10

Statistical Procedure

Center for Cosmology and Particle Physics

The full complexity of individual channels likelihood functions are packaged using RooFit/RooStats workspaces, and a combined probability model is formed by identifying nuisance parameters associated to common systematic effects the common parameter of interest is a cross-section scale factor: = / SM The profile likelihood ratio is used as a test statistic: () = Ls+b (, )/Ls+b (, )

nuisance parameters are profiled based on the data one-sided variants of the test statistic are used for upper-limits and discovery

The distribution of the test statistic is obtained in two ways:

Ensemble tests with toy Monte Carlo using a fully frequentist procedure

randomize auxiliary measurements instead of randomizing nuisance parameters

Cowan, Cranmer, Gross, Vitells Eur.Phys.J. C71 (2011)

Using asymptotic distribution of likelihood ratio

used for primary result

Primary result based CLs, conservatism introduced to protect against downward fluctuations

results based on power-constrained CLs+b (PCL) in backup Additional comparisons with Bayesian procedure with a uniform prior on
EPS-HEP 2011 - ATLAS Higgs Combination 11

Kyle Cranmer (NYU)

Limits in the low mass range

Impressive sensitivity to a Higgs boson in the mass range 135-200 GeV Excess in H WW ll leads to weaker-thanexpected limits near MH = 130-160 GeV
SM

Center for Cosmology and Particle Physics

ATLAS Preliminary Observed CLs Expected 1 2 Ldt = 1.0-1.2 fb s = 7 TeV

-1

95% CL Limit on

/ 10 1
110 120 130 140 150 160 170 180 190 200

mH [GeV]
155<mH<190 GeV/c2 excluded at 95% CL

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

12

Limits in the low mass range

Impressive sensitivity to a Higgs boson in the mass range 135-200 GeV Excess in H WW ll leads to weaker-thanexpected limits near MH = 130-160 GeV Extends exclusion range significantly beyond Tevatron from ~175 to 190 Not yet competitive near LEP limit
SM

ATLAS Preliminary

when the background model grossly overpredicts the data or the data exhibit a large background-like uctuation limit at the -1 background uctuation level.). Within Figure 9, 95% C.L. power-constrained limits can be Center for at the points for which 1-CLs+b exceeds 95%. The expected range of exclusion is 40% larger using PCL tha Cosmology and Bayesian and CLs limits quoted here. We continue our convention of quoting Bayesian and CLs limits however Particle Physics In summary, we combine CDF and D0 results on SM Higgs boson searches, based on luminosities up to 8.2 Compared to our previous combination, more data have been added to the existing channels, additional cha have been included, and analyses have been further optimized to gain sensitivity. We use the recommendation PDF4LHC working group for the central value of the parton distribution functions and uncertainties [32]. We u

95% CL Limit on

95% CL Limit/SM

10

Observed CLs -1 Ldt ! 1.0-1.2 fb Tevatron Expected Run II Preliminary, L= 8.2 fb-1 10 1 Expected Tevatron Observed s = 7 TeV 1" Expected Exclusion 2 2" Expected

SM=1

March 7, 2011

180 190 200 2 mH (GeV/c ) 110 120 130 140 150 160 170 180 190 200

130

140

150

160

170

FIG. 7: Observed and expected (median, for the background-only hypothesis) 95% C.L. upper limits on the ratios to t cross section, as functions of the Higgs boson mass for the combined CDF and D0 analyses. The limits are expresse multiple of the SM prediction for test masses (every 5 GeV/c2 ) for which both experiments have performed dedicated se in dierent channels. The points are joined by straight lines for better readability. The bands indicate the 68% an probability regions where the limits can uctuate, in the absence of signal. The limits displayed in this gure are obtaine the Bayesian calculation. 2

mH [GeV]

155<mH<190 GeV/c excluded at 95% CL

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

13

Background-only p-values at low-mass

Center for Cosmology and Particle Physics

Broad WW excess is modulated by local fluctuations in and 4l local significance, no look-elsewhere effect correction applied
p0 1 ATLAS Preliminary

Test statistic defined such that downward fluctuation gives p0=50% Largest excess has approximately 2.8 significance

10-1 10-2 10-3 10 10

-4

-5

Observed Expected 2 3 120 140

Ldt = 1.0-1.2 fb s = 7 TeV 160 180

-1

10-6

200 mH [GeV]

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

14

Background-only p-values at low-mass

Center for Cosmology and Particle Physics

Broad WW excess is modulated by local fluctuations in and 4l local significance, no look-elsewhere effect correction applied
p0 1 ATLAS Preliminary

Test statistic defined such that downward fluctuation gives p0=50% Largest excess has approximately 2.8 significance

10-1 10-2 10-3 10 10

-4

-5

Observed Expected 2 3 120 140

Ldt = 1.0-1.2 fb s = 7 TeV 160 180

-1

10-6
deficit excess
Kyle Cranmer (NYU)

200 mH [GeV]

4l candidate

Broad WW ll excess
14

EPS-HEP 2011 - ATLAS Higgs Combination

Limits full mass range

Additional High-mass channels extend the HZZll exclusion Noticeable excess around 250 GeV from H ZZ4l candidates
SM

Center for Cosmology and Particle Physics

ATLAS Preliminary Observed CLs Expected 1 2 Ldt = 1.0-1.2 fb s = 7 TeV

-1

95% CL Limit on

/
10 1

CLs

ATLAS Preliminary

10-1 10-2 10-3 10-4 100 200 300 400 Observed Expected 90% CL 95% CL Ldt = 1.0-1.2 fb s = 7 TeV 500
-1

200

300

400

500 600 mH [GeV]

600 mH [GeV]

155<MH<190 and 295 <MH< 450 GeV/c2 excluded at @ 95% CL

15

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

p-values full mass range

Center for Cosmology and Particle Physics

Approximately 8% chance of background-only fluctuation this large anywhere in range

p0 1 ATLAS Preliminary

No combined excess beyond 3 is observed

10-1 10-2

Asymptotic approximation in good agreement with ensemble tests

10-3 10-4 10-5 100

HZZ 4l candidates

Obs. Asymp. -1 Expected Ldt = 1.0-1.2 fb 2 s = 7 TeV 3 Obs. Ensemble 200 300 400 500 600 mH [GeV]

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

16

Cross-checks

Center for Cosmology and Particle Physics

The full limit procedure was performed with toys to confirm the asymptotic distributions of the profile likelihood ratio

Toy Monte Carlo is significantly more computationally intensive and sensitive to fit failures etc. Asymptotic results are robust

In addition, a Bayesian procedure, which is known to reproduce the CLs limit in simple problems, also yielded consistent results.
SM

95% CL Limit on /

8 6 4 2 0

Observed CLs Expected CLs Observed Bayesian Expected Bayesian Observed Ensemble Tests Expected Ensemble Tests

95% CL Limit on /

10

ATLAS Preliminary Ldt = 1.0-1.2 fb s = 7 TeV

-1

SM

10 8 6 4 2 0

ATLAS Preliminary
Observed CLs Expected CLs Observed Bayesian Expected Bayesian Observed Ensemble Tests Expected Ensemble Tests

Ldt = 1.0-1.2 fb s = 7 TeV

-1

120

140

160

180

200 mH [GeV]

100

200

300

400

500

600 mH [GeV]
17

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

Conclusions

Center for Cosmology and Particle Physics

Thanks to the excellent LHC operations, ATLAS has collected more than 1 fb-1 of 7 TeV data leading to substantial gains in sensitivity to the Standard Model Higgs

In the low-mass range (120 140 GeV) an excess of events with a significance of approximately 2.8 is observed.

ATLAS has extended the 95% CL excluded region around 2MW to 155<MH<190 GeV and excluded a new range from 295<MH<450 GeV We congratulate the LHC for terrific performance and look forward to more successful running in 2011! We also look forward to the results from CMS and the upcoming ATLAS+CMS Higgs combination

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

18

Center for Cosmology and Particle Physics

Backup

Kyle Cranmer (NYU)

EPS-HEP 2011 - ATLAS Higgs Combination

19

95% CL Limit on /

Results with PCL

ATLAS Preliminary Alternate Test Observed PCL Observed -1 Ldt = 1.0-1.2 fb Expected 10 1 s = 7 TeV While CLs is well a established technique + 2

Center for Cosmology and Particle Physics

in our field, it is considered a nonstandard procedure by statistician mixing notions of power and coverage
1

it intentionally over-covers to protect against setting limits beyond the experiments sensitivity dueJuly 22, 2011 08 : 57 fluctuations DRAFT to downward

An alternative approach (PCL) is based on purely frequentist CLs+b together with a power-constraint at500 experiments sensitivity achieves the same the 600 10-1 200 300 400 protection without mixing the mH [GeV] of coverage and power notions
95% CL Limit on /SM
95% CL Limit on /SM
ATLAS Preliminary Observed PCL Observed Expected 1 + 2 Alternate Test

10

Ldt = 1.0-1.2 fb
s = 7 TeV

-1

10

ATLAS Preliminary Observed PCL Observed Expected 1 + 2

Alternate Test

Ldt = 1.0-1.2 fb s = 7 TeV

-1

10-1 110 120 130 140 150 160 170 180 190 200 mH [GeV]
Kyle Cranmer (NYU)

10-1

200

300

400

500 600 mH [GeV]

20

EPS-HEP 2011 - ATLAS Higgs Combination

M