University of Pennsylvania

Search for Higgs in H→WW→lνlν

John Alison

AT

L-P

HY

S-P

UB

-20

10

-01

5

08

/1

1/

20

10

ATLAS NOTE

November 8, 2010

ATLAS Sensitivity Prospects for Higgs Boson Production at the LHC

Running at 7, 8 or 9 TeV

The ATLAS Collaboration

Abstract

Projections for the ATLAS sensitivity to the Standard Model Higgs boson from the LHC

running at centre-of-mass energies of 7, 8 and 9 TeV are reported. These studies extend our

previous 7 TeV results by adding extra analysis channels and using more accurate cross-

sections calculations. The VH, H → bb̄ and H → τ+τ− improve the sensitivity at low

mass, while new analyses of H → ZZ → l+

l−νν̄ and H → ZZ → l

+l−

bb̄ give major

improvements above 200 GeV. There are also a range of possible integrated luminosity

scenarios considered. The projections are largely derived using the ratio of cross-sections at

different centre-of-mass energies.

The ATLAS experiment, with 1 fb−1

at 7 TeV and using rather conservative assumptions,

should have sensitivity such that Higgs boson masses between 129 and 460 GeV have at least

a 50% chance of being excluded at 95% CL. If 2 fb−1

at 8 TeV is available with analysis

improvements the sensitive region ranges from 114 to over 500 GeV.on behalf of the ATLAS Collaboration

Higgs Physics

2.2 Fit results 8

[GeV]HM50 100 150 200 250 300

2 !"

0

1

2

3

4

5

6

7

8

9

10

LEP

95%

CL

Teva

tron

95%

CL

#1

#2

#3

Theory uncertaintyFit including theory errorsFit excluding theory errors

[GeV]HM100 150 200 250 300

2 !"

0

2

4

6

8

10

12

14

16

18

20

LEP

95%

CL

Teva

tron

95%

CL

#1

#2

#3

#4

Theory uncertaintyFit including theory errorsFit excluding theory errors

neglects correlations

Figure 4: Indirect determination of the Higgs boson mass: !!2 as a function of MH for the standard fit(top) and the complete fit (bottom). The solid (dashed) lines give the results when including (ignoring)theoretical errors. Note that we have modified the presentation of the theoretical uncertainties here withrespect to our earlier results [1]. Before, the minimum !2

min of the fit including theoretical errors was usedfor both curves to obtain the o"set-corrected!!2. We now individually subtract each case so that both !!2

curves touch zero. In spite of the di"erent appearance, the theoretical errors used in the fit are unchangedand the numerical results, which always include theoretical uncertainties, are una"ected.

2.2 Fit results 7

meas!) / meas - Ofit

(O-3 -2 -1 0 1 2 3

tmbmcm

W"

WM)2

Z(M(5)

had#$

b0Rc0RbAcA

0,bFBA

0,cFBA

)FB

(Qlepteff%2sin

(SLD)lA(LEP)lA

0,lFBAlep0R

0had!

Z"

ZM

0.3-0.00.00.2

-1.2-0.1-0.80.10.6-0.12.50.9

-0.7-2.00.2

-0.8-1.0-1.70.20.1

[GeV]HM20 40 60 80 100 120 140 160

tmbmcm

W"

WM)2

Z(M(5)

had#$

b0Rc0RbAcA

0,bFBA

0,cFBA

)FB

(Qlepteff%2sin

(SLD)lA(LEP)lA

0,lFBAlep0R

0had!

Z"

ZM

- 74 +209141 - 25 + 32 99 - 24 + 31 96 - 24 + 30 96 - 34 + 75116 - 43 + 62 44 - 24 + 31 96 - 24 + 31 96 - 24 + 31 96 - 24 + 31 96 - 19 + 40 61 - 24 + 30 94 - 24 + 30 94 - 32 + 40122 - 24 + 31 92 - 25 + 32100 - 25 + 31 99 - 25 + 31 97 - 25 + 32 97 - 26 + 49 61

Figure 2: Comparing fit results with direct measurements: pull values for the complete fit (left), and resultsfor MH from the standard fit excluding the respective measurements from the fit (right).

[GeV]HM6 10 20 210 210&2 310

Standard fit

WM

0,bFBA

(SLD) lA

(LEP) lA

-24 +31 96

-45 +41 68

-330 +494562

-19 +60 37

-71 +263117

Figure 3: Determination of MH excluding all the sensitive observables from the standard fit except the onegiven. Note that the results shown are not independent. The information in this figure is complementaryto that of the right hand plot of Fig. 2.

Theory vs Experiment

Best Fit for Higgs Mass

Standard Model Remarkably Accurate Description Data. One Remaining Piece: Higgs Boson. Data predict m(H) below ~200 GeV

arXiv:1109.0975

arXiv:1109.0975

Searching for the Higgs

H→WW→lνlν Strongest sensitivity over broad range of m(H) Critical in the region between LEP and SM EWK exclusion

[GeV]HM100 120 140 160 180 200

Bran

chin

g ra

tios

-310

-210

-110

1bb

cc

gg

Z

WW

ZZ

LHC

HIG

GS

XS W

G 2

010

arXiv:1101.0593

]2 [GeV/cHm100 110 120 130 140 150 160 170 180 190 200

Rec

onst

ruct

able

Cro

ss-s

ectio

n [p

b]-310

-210

-110

1

)TEh

l(q qH q q qq

ll WW H gg

llll ZZ H gg H gg

Hww AnalysisEn

tries

/ 5

GeV

-210-1101

10210310410510610710810910 Data stat) SM (sys

WW WZ/ZZ/W

t t Single Top Z+jets W+jets

H [150 GeV]

ATLAS Preliminary-1 L dt = 1.70 fb = 7 TeV, s

eeWWH

[GeV]eeM0 50 100 150 200 250

Dat

a / M

C

0.51

1.52

Entri

es /

5 G

eV

-210-1101

10210310410510610710810910 Data stat) SM (sys

WW WZ/ZZ/W

t t Single Top Z+jets W+jets

H [150 GeV]

ATLAS Preliminary-1 L dt = 1.70 fb = 7 TeV, s

eeWWH

[GeV]eeM0 50 100 150 200 250

Data

/ M

C

0.51

1.52

Entri

es /

5 G

eV

-210-1101

10210310410510610710810910 Data stat) SM (sys

WW WZ/ZZ/W

t t Single Top Z+jets W+jets

H [150 GeV]

ATLAS Preliminary-1 L dt = 1.70 fb = 7 TeV, s

eeWWH

[GeV]eeM0 50 100 150 200 250

Dat

a / M

C

0.51

1.52

Entri

es /

10 G

eV

51015202530354045 Data stat) SM (sys

WW WZ/ZZ/W

t t Single Top Z+jets W+jets (data driven)

H [150 GeV]

ATLAS Preliminary-1 L dt = 1.70 fb = 7 TeV, s

+ 0 jetsllWWH

[GeV]TM60 80 100 120 140 160 180 200 220 240

Dat

a / M

C

0.51

1.52

Entri

es /

10 G

eV

51015202530354045 Data stat) SM (sys

WW WZ/ZZ/W

t t Single Top Z+jets W+jets (data driven)

H [150 GeV]

ATLAS Preliminary-1 L dt = 1.70 fb = 7 TeV, s

+ 0 jetsllWWH

[GeV]TM60 80 100 120 140 160 180 200 220 240

Data

/ M

C

0.51

1.52

Entri

es /

10 G

eV

51015202530354045 Data stat) SM (sys

WW WZ/ZZ/W

t t Single Top Z+jets W+jets (data driven)

H [150 GeV]

ATLAS Preliminary-1 L dt = 1.70 fb = 7 TeV, s

+ 0 jetsllWWH

[GeV]TM60 80 100 120 140 160 180 200 220 240

Dat

a / M

C

0.51

1.52

Challenging AnalysisMany backgrounds No clear mass peak

Dilepton Selection

After Hww Selection

All channels

Missing EnergyJet Veto Exploit Higgs properties Spin-0. Expected Mass Range

http://cdsweb.cern.ch/record/1383837

[GeV]Hm120 140 160 180 200 220 240 260 280 300

SM/

95%

CL

Lim

it on

-110

1

10

210ObservedExpected 1 ! 2 !

ATLAS Preliminary

-1 Ldt = 1.7 fb = 7 TeVs

ll(*)WWHCLs Limits

Resultshttp://cdsweb.cern.ch/record/1383837

2.2 Fit results 8

[GeV]HM50 100 150 200 250 300

2 !"

0

1

2

3

4

5

6

7

8

9

10

LEP

95%

CL

Teva

tron

95%

CL

#1

#2

#3

Theory uncertaintyFit including theory errorsFit excluding theory errors

[GeV]HM100 150 200 250 300

2 !"

0

2

4

6

8

10

12

14

16

18

20

LEP

95%

CL

Teva

tron

95%

CL

#1

#2

#3

#4

Theory uncertaintyFit including theory errorsFit excluding theory errors

neglects correlations

Figure 4: Indirect determination of the Higgs boson mass: !!2 as a function of MH for the standard fit(top) and the complete fit (bottom). The solid (dashed) lines give the results when including (ignoring)theoretical errors. Note that we have modified the presentation of the theoretical uncertainties here withrespect to our earlier results [1]. Before, the minimum !2

min of the fit including theoretical errors was usedfor both curves to obtain the o"set-corrected!!2. We now individually subtract each case so that both !!2

curves touch zero. In spite of the di"erent appearance, the theoretical errors used in the fit are unchangedand the numerical results, which always include theoretical uncertainties, are una"ected.

ExpectedExcluded

ATLAS Hww

arXiv:1109.0975

Results

The Future is Now.Figure 13 (left) shows the effect of increasing the integrated luminosity from 0.5 fb

−1to 5 fb

−1at

7 TeV. The lower expected exclusion limit reduces by about 7 GeV with each doubling of the integrated

luminosity, meeting the LEP bound for 5 fb−1

. The 2 fb−1

line can be taken as indicative of what might

be achieved by combining the results of 1 fb−1

analysed by each of ATLAS and CMS.

[GeV]Hm100 110 120 130 140 150 160 170 180 190 200

SM!/

!95

% C

L Li

mit

on

-110

1

10 -10.5fb-11fb-12fb-15fbTeVatronLEP

ATLAS Preliminary (Simulation)

=7 TeVsProjections,

[GeV]Hm100 150 200 250 300 350 400 450 500 550 600

SM!/

!Se

nsiti

vity

to

-110

1

10-1 obs., 1 fb!3-1 obs., 2 fb!3-1 obs., 5 fb!3

TeVatron 95% CL exc.LEP 95% CL exc.

ATLAS Preliminary (Simulation)Median sensitivity

=7 TeV Projectionss

Figure 13: Combined sensitivity for different integrated luminosity scenarios. Left shows the 95% CL

median sensitivities and right gives the 3σ sensitivity. Public limits from LEP [39] and the Tevatron [38]

are shown for comparison.

The 3σ sensitivity is illustrated in Fig. 13 (right). There is a region, extending from Higgs boson

masses of 139 to 180 GeV, where 1 fb−1

would be expected to lead to evidence at this level. If 2 fb−1

becomes available then this region extends down to 131 GeV, and in addition there is close to 50% chance

of 3σ evidence for a Higgs boson with mass between 200 and 430 GeV.

The effect of raising the collision energy to 8 or 9 TeV is shown in Fig. 14, where results at 1 and

2 fb−1

are contrasted. The expected excluded region for 1 fb−1

at 8 TeV is 127 to 525 GeV. There is

a small but significant increase in sensitivity at low mass, while there is a particularly clear gain at the

upper mass limit, because of the increased phase-space for the production of heavy objects. This effect

can be seen in the bottom plot.

The aggressive analyses which have been mentioned with each channel summarise the potential to

make more sensitive searches if various conditions can be met. This has been included by applying

scaling factors to the sensitivity of some of the channels (15% for H → ZZ, 50% for H → γγ, and 30%

for each of H → τ+τ− and H → bb̄.) The H → WW channel has had the systematics decreased to

10% as described in Ref. [1]. The expected limits in this optimistic scenario using 2 fb−1

at 8 TeV are

also shown in Fig. 14, where it can be seen that under these assumptions there is at least 50% exclusion

probability from the LEP bound to well over 500 GeV.

The amount of integrated luminosity at 8 or 9 TeV which gives the same median sensitivity as 1 fb−1

at 7 TeV as a function of Higgs boson mass is shown in Fig 15. This is estimated by first fitting the

gain in sensitivity from 1 to 2 fb−1

to a function of the form σ ∝ 1/Lα. The exponent α would be

0.5 in a background dominated regime with no systematics. In fact it varies from 0.6 to 0.3, with the

minimum at 170 GeV where the systematic errors are important. This expression is used to find the

integrated luminosity at 8 or 9 TeV which matches the sensitivity seen for 1 fb−1

at 7 TeV. For masses

below 200 GeV approximately 750 to 800 pb−1

at 8 TeV or 600 to 650 pb−1

at 9 TeV is as powerful as

1 fb−1

at 7 TeV. The required integrated luminosity is lower for higher mass Higgs bosons.

26

Presented Here

In the works

Its a great time to be doing Higgs Physics!

Figure 13 (left) shows the effect of increasing the integrated luminosity from 0.5 fb−1

to 5 fb−1

at

7 TeV. The lower expected exclusion limit reduces by about 7 GeV with each doubling of the integrated

luminosity, meeting the LEP bound for 5 fb−1

. The 2 fb−1

line can be taken as indicative of what might

be achieved by combining the results of 1 fb−1

analysed by each of ATLAS and CMS.

[GeV]Hm100 110 120 130 140 150 160 170 180 190 200

SM!/

!95

% C

L Li

mit

on

-110

1

10 -10.5fb-11fb-12fb-15fbTeVatronLEP

ATLAS Preliminary (Simulation)

=7 TeVsProjections,

[GeV]Hm100 150 200 250 300 350 400 450 500 550 600

SM!/

!Se

nsiti

vity

to

-110

1

10-1 obs., 1 fb!3-1 obs., 2 fb!3-1 obs., 5 fb!3

TeVatron 95% CL exc.LEP 95% CL exc.

ATLAS Preliminary (Simulation)Median sensitivity

=7 TeV Projectionss

Figure 13: Combined sensitivity for different integrated luminosity scenarios. Left shows the 95% CL

median sensitivities and right gives the 3σ sensitivity. Public limits from LEP [39] and the Tevatron [38]

are shown for comparison.

The 3σ sensitivity is illustrated in Fig. 13 (right). There is a region, extending from Higgs boson

masses of 139 to 180 GeV, where 1 fb−1

would be expected to lead to evidence at this level. If 2 fb−1

becomes available then this region extends down to 131 GeV, and in addition there is close to 50% chance

of 3σ evidence for a Higgs boson with mass between 200 and 430 GeV.

The effect of raising the collision energy to 8 or 9 TeV is shown in Fig. 14, where results at 1 and

2 fb−1

are contrasted. The expected excluded region for 1 fb−1

at 8 TeV is 127 to 525 GeV. There is

a small but significant increase in sensitivity at low mass, while there is a particularly clear gain at the

upper mass limit, because of the increased phase-space for the production of heavy objects. This effect

can be seen in the bottom plot.

The aggressive analyses which have been mentioned with each channel summarise the potential to

make more sensitive searches if various conditions can be met. This has been included by applying

scaling factors to the sensitivity of some of the channels (15% for H → ZZ, 50% for H → γγ, and 30%

for each of H → τ+τ− and H → bb̄.) The H → WW channel has had the systematics decreased to

10% as described in Ref. [1]. The expected limits in this optimistic scenario using 2 fb−1

at 8 TeV are

also shown in Fig. 14, where it can be seen that under these assumptions there is at least 50% exclusion

probability from the LEP bound to well over 500 GeV.

The amount of integrated luminosity at 8 or 9 TeV which gives the same median sensitivity as 1 fb−1

at 7 TeV as a function of Higgs boson mass is shown in Fig 15. This is estimated by first fitting the

gain in sensitivity from 1 to 2 fb−1

to a function of the form σ ∝ 1/Lα. The exponent α would be

0.5 in a background dominated regime with no systematics. In fact it varies from 0.6 to 0.3, with the

minimum at 170 GeV where the systematic errors are important. This expression is used to find the

integrated luminosity at 8 or 9 TeV which matches the sensitivity seen for 1 fb−1

at 7 TeV. For masses

below 200 GeV approximately 750 to 800 pb−1

at 8 TeV or 600 to 650 pb−1

at 9 TeV is as powerful as

1 fb−1

at 7 TeV. The required integrated luminosity is lower for higher mass Higgs bosons.

26

http://cdsweb.cern.ch/record/1303604

Bonus

[GeV]Hm120 140 160 180 200 220 240

SM/

95%

CL

Lim

it on

-110

1

10

ObservedExpected 1 ! 2 !

ATLAS Preliminary

-1 Ldt = 1.0-2.3 fb = 7 TeVs

CLs Limits

Combined Limit

[GeV]Hm100 200 300 400 500

-1In

tegr

ated

Lum

inos

ity, f

b

0

2

4

6

8

10

12

=7 TeVs !5 =8 TeVs !5=7 TeVs !3 =8 TeVs !3

=7 TeVs95% CL =8 TeVs95% CL

120 150

ATLAS Preliminary (Simulation)

Figure 3: The luminosity required to give exclusion, evidence or discovery sensitivity for a SM Higgswith data at

!s = 7 or 8 TeV.

5

The future is now.

The Future is Now.

Figure 13 (left) shows the effect of increasing the integrated luminosity from 0.5 fb−1

to 5 fb−1

at

7 TeV. The lower expected exclusion limit reduces by about 7 GeV with each doubling of the integrated

luminosity, meeting the LEP bound for 5 fb−1

. The 2 fb−1

line can be taken as indicative of what might

be achieved by combining the results of 1 fb−1

analysed by each of ATLAS and CMS.

[GeV]Hm100 110 120 130 140 150 160 170 180 190 200

SM!/

!95

% C

L Li

mit

on

-110

1

10 -10.5fb-11fb-12fb-15fbTeVatronLEP

ATLAS Preliminary (Simulation)

=7 TeVsProjections,

[GeV]Hm100 150 200 250 300 350 400 450 500 550 600

SM!/

!Se

nsiti

vity

to

-110

1

10-1 obs., 1 fb!3-1 obs., 2 fb!3-1 obs., 5 fb!3

TeVatron 95% CL exc.LEP 95% CL exc.

ATLAS Preliminary (Simulation)Median sensitivity

=7 TeV Projectionss

Figure 13: Combined sensitivity for different integrated luminosity scenarios. Left shows the 95% CL

median sensitivities and right gives the 3σ sensitivity. Public limits from LEP [39] and the Tevatron [38]

are shown for comparison.

The 3σ sensitivity is illustrated in Fig. 13 (right). There is a region, extending from Higgs boson

masses of 139 to 180 GeV, where 1 fb−1

would be expected to lead to evidence at this level. If 2 fb−1

becomes available then this region extends down to 131 GeV, and in addition there is close to 50% chance

of 3σ evidence for a Higgs boson with mass between 200 and 430 GeV.

The effect of raising the collision energy to 8 or 9 TeV is shown in Fig. 14, where results at 1 and

2 fb−1

are contrasted. The expected excluded region for 1 fb−1

at 8 TeV is 127 to 525 GeV. There is

a small but significant increase in sensitivity at low mass, while there is a particularly clear gain at the

upper mass limit, because of the increased phase-space for the production of heavy objects. This effect

can be seen in the bottom plot.

The aggressive analyses which have been mentioned with each channel summarise the potential to

make more sensitive searches if various conditions can be met. This has been included by applying

scaling factors to the sensitivity of some of the channels (15% for H → ZZ, 50% for H → γγ, and 30%

for each of H → τ+τ− and H → bb̄.) The H → WW channel has had the systematics decreased to

10% as described in Ref. [1]. The expected limits in this optimistic scenario using 2 fb−1

at 8 TeV are

also shown in Fig. 14, where it can be seen that under these assumptions there is at least 50% exclusion

probability from the LEP bound to well over 500 GeV.

The amount of integrated luminosity at 8 or 9 TeV which gives the same median sensitivity as 1 fb−1

at 7 TeV as a function of Higgs boson mass is shown in Fig 15. This is estimated by first fitting the

gain in sensitivity from 1 to 2 fb−1

to a function of the form σ ∝ 1/Lα. The exponent α would be

0.5 in a background dominated regime with no systematics. In fact it varies from 0.6 to 0.3, with the

minimum at 170 GeV where the systematic errors are important. This expression is used to find the

integrated luminosity at 8 or 9 TeV which matches the sensitivity seen for 1 fb−1

at 7 TeV. For masses

below 200 GeV approximately 750 to 800 pb−1

at 8 TeV or 600 to 650 pb−1

at 9 TeV is as powerful as

1 fb−1

at 7 TeV. The required integrated luminosity is lower for higher mass Higgs bosons.

26

[GeV]Hm200 300 400 500 600

SM/

95%

CL

Lim

it on

-110

1

10

ObservedExpected 1 ! 2 !

ATLAS Preliminary

-1 Ldt = 1.0-2.3 fb = 7 TeVs

CLs Limits

Full Mass Range

Hww Sensitivity

H!WW!l"l"

Introduction

Aim - Document analysis in paper. Will be basis of future results. - Venue to address analysis aspects put off in conference rush

3

H!WW!l"l" is one of our most important channels - Strongest SM sensitivity over a broad range of m(H) - Critical in mass region between LEP and SM EWK exclusion.