ATLAS NOTE
February 17, 2011
Search for the Standard Model Higgs boson in the decay channel
1
H → ZZ
(∗)→ 4 ℓ with 40 pb
−1of pp collisions at √
s = 7 TeV
2
The ATLAS collaboration
3
Abstract
4
This note presents the search for a Standard Model Higgs boson in the channelH→
5
ZZ(∗)→4ℓusing approximately 40 pb −1 of pp collison data collected by the ATLAS ex-
6
periment in 2010. No candidate events are observed in the current data. Upper limits on
7
the Standard Model Higgs production are extracted as a function of the Higgs mass in the
8
mass range 130 to 600 GeV. In addition, data-driven background estimation techniques are
9
described and the results are found to be in agreement with the Monte Carlo predictions.
10
1 Introduction
11
The decay channelH→ZZ(∗)→ℓℓℓ′ℓ′, whereℓ=e,µ, is one of the experimentally cleanest signatures
12
for the search of the Standard Model Higgs boson in the Higgs mass,MH, range between 130 and 600
13
GeV. For the high mass region – MH>180 GeV – the two on-shell Z bosons from the Higgs decay
14
allow for a selection which strongly suppresses the background leaving only the irreducible ZZ→4ℓ
15
component. At low Higgs masses, where one of the decay bosons is off-shell, contributions fromZ+jets
16
andtt¯can be significant and tighter cuts are therefore applied to reduce theses backgrounds to a level
17
safely below theZZ∗continuum. The most challenging mass region is around 160 GeV where the phase
18
space opens up to the two on-shellW boson decays and suppresses the ZZ branching ratio. The search in
19
the mass region below 130 GeV is limited by a low branching ratio and a softpTspectrum of the leptons
20
from theZ∗decays.
21
This note presents a search for four lepton events consistent with the hypothesis of Higgs boson
22
decay using the first 40 pb−1ofppcollisions at LHC at√
s=7 TeV with the ATLAS detector [1].
23
2 Data and Monte Carlo samples
24
The data used in this analysis were collected by the ALTAS detector during the 2010 LHC pprun at a
25
center-of-mass enery of 7 TeV. The total recorded data amounts to an integrated luminosity of 45 pb−1.
26
Applying beam, detector, and data quality requirements results in a total integrated luminosity of 43 pb−1
27
for the 4µ channel, 39 pb−1for the 2e2µ channel and 39 pb−1for the 4echannel.
28
MH σHO·BR MH σHO·BR MH σHO·BR
(GeV) (fb) (GeV) (fb) (GeV) (fb)
120 1.37 220 6.16 420 2.24
130 2.87 240 5.35 440 1.89
140 4.23 260 4.68 460 1.59
150 4.38 280 4.16 480 1.33
160 1.90 300 3.75 500 1.11
165 0.93 320 3.49 520 0.94
170 0.92 340 3.40 540 0.79
180 2.04 360 3.42 560 0.66
190 6.22 380 3.08 580 0.56
200 6.77 400 2.66 600 0.47
Table 1: Higgs boson production cross-sections for both gluon and vector-boson fusion processes in pp collisions at√
s=7 TeV. The cross-sections include the branching ratio ofH→4ℓ,ℓ=e,µ.
For the Monte Carlo samples, the Higgs signal is generated using PYTHIA [2], including both
29
gluon fusion and vector boson fusion production mechanisms. The PYTHIA generator is interfaced
30
to PHOTOS [3] for final-state radiation. The inclusive signal production cross-sections in ppcollisions
31
at√
s=7 TeV is estimated at the currently highest-order (HO) known, as discussed in [4]. For gluon
32
fusion the most accurate prediction comes from a NNLL QCD + NLO EW estimation, while vector bo-
33
son fusion process has been recently evaluated up to NNLO QCD, combined with full NLO EW order
34
corrections [5, 6]. The remaining total uncertainties, estimated by summing linearly the QCD scale and
35
the combined PDF +αscontributions, amount to±15-20% for gluon fusion and to less than 5% for the
36
vector boson fusion process. For Higgs masses below theWW/ZZthreshold, the interference of the final
37
state leptons leads to a significant enhancement of the branching ratio (BR) for H→4e andH →4µ.
38
Above theWW/ZZthresholds the interference is small, since the resonances occur in different regions
39
of phase space forWW and ZZ. TheBR(H →4ℓ) used in this paper have been estimated using the
40
program PROPHECY4F[7], with complete NLO QCD+EW corrections including all interference and the
41
leading two-loop heavy-Higgs contributions. The theoretical references used in the following, and listed
42
the second column of Table 1, have been then obtained asσHO(ggF + VBF)×BR(H→4ℓ),ℓ=e,µ for
43
Ref. [4].
44
For the background samples, different generators are employed. The production of the irreducible
45
ZZ(∗)→4ℓbackground receives contributions fromqq¯initiated state, which is implemented in PYTHIA,
46
and from agg initiated processes, evaluated with dedicated programs [8]. The inclusiveZ boson and
47
Zbb¯ production is modelled using ALPGEN [9], while for thett¯production MC@NLO [10, 11, 12] is
48
employed. Both ALPGEN and MC@NLO generators are interfaced to HERWIG [13] for parton shower
49
and hadronisation and JIMMY [14] for simulation of the underlying event. For the inclusiveZ boson
50
production also PYTHIA is used.
51
The simulated samples considered in this analysis are summarised in Table 2. Cross-sections from
52
the most advanced calculations are used whenever available. The generated events are used as input to a
53
full simulation of the ATLAS detector using GEANT4 [15].
54
Process Generator σ×Br
gg,qq→H Pythia See Table 1
Z/γ∗ →ℓℓ Pythia
Mℓℓ>60 GeV 0.989 nb [16, 17]
Z/γ∗bb¯→ℓℓbb¯ ALPGEN
→ℓℓbb¯+ 0p 7.95 pb
→ℓℓbb¯+ 1p 3.01 pb
→ℓℓbb¯+ 2p 0.986 pb
→ℓℓbb¯+ 3p 0.472 pb
ZZ Pythia
MZ>12 GeV
qq→ZZ 9.23 pb [18]
gg→ZZ 0.53 pb [8]
tt¯ MC@NLO 164.6 pb [19]
Table 2: Monte-Carlo programs used for modeling signal and background processes and their corre- sponding cross-sections.
3 Lepton reconstruction and identification
55
Lepton identification and reconstruction in ATLAS, of particular importance for this search, is briefly
56
described and the baseline lepton selection for the analysis is defined.
57
Electrons consist of electromagnetic clusters to which inner detector tracks are matched in a broad
58
window between the cluster position and the extrapolated track. The baseline electron identification in
59
ATLAS relies on cuts using variables that provide good separation between isolated electrons and jets.
60
These variables include calorimeter, tracker and combined calorimeter/tracker information. They can
61
be applied independently and three reference sets of cuts have been defined with increasing background
62
rejection power: loose,mediumandtight. Shower shape variables of the second calorimeter layer and
63
hadronic leakage variables are used in the loose selection. First calorimeter layer cuts, track quality re-
64
quirements and track-cluster matching are added at the level of the medium selection. The tight selection
65
adds E/p, b-layer hit requirements and the particle identification potential of the TRT. To select elec-
66
tron candidates, EM calorimeter clusters are required to pass several quality criteria and to lie outside
67
problematic calorimeter regions.
68
Muons are identified in one of three different ways: as reconstructed tracks in the muon spectrometer
69
alone – “stand-alone” –, as the fitted combination of inner detector and muon spectrometer tracks –
70
“combined” – or by matching an inner detector track of sufficient momentum with a reconstructed track
71
segment of the muon spectrometer – “segment-tagged”.
72
Throughout this paper, the medium electron and combined or segment-tagged muon selection are
73
used if not explicitly stated otherwise. The challenging low pT electron region (below 15 GeV) is not
74
yet included in this analysis. A cut at 15 GeV is thus during the event selection of the Higgs candidates.
75
4 Event selection
76
Events are collected using single lepton triggers with pT thresholds in the range 10 - 15 GeV. The
77
efficiency of these triggers with respect to the offline selection is close to 100%. Collision candidates are
78
selected by requiring a primary vertex with at least three tracks, consistent with the beam spot position.
79
Event Preselection
Electrons: Medium quality electrons withET>15 GeV and|η|<2.47 Muons: Combined or segment-tagged muons with pT>7 GeV and|η|<2.5
Four leptons required with at least two with pT>20 GeV Event Selection
Kinematic Require at least one quadruplet of leptons consisting of two pairs of same Selection flavour opposite charge leptons (Z,Z(∗)) fullfilling the following requirements.
∆R(l,l′)>0.1 for all leptons in the quadruplet.
|Mℓ1ℓ2−MZ|<∆M12; Mℓ3ℓ4>M34
Isolation Lepton Track isolation (∆R=0.30):ΣpT/pT<0.20 Lepton Calorimeter isolation (∆R=0.30) : ΣET/ET<0.30
Impact Apply impact parameter significance cut to the 2 less energetic leptons of the quadruplet.
Parameter For electrons :d0/σd0 <6 Significance For muons :d0/σd0 <3.5
ForM4l>190 GeV no requirement applied
Table 3: Summary of the event selection requirements. The two lepton pairs are denoted as Mℓ1ℓ2 and Mℓ3ℓ4. The values of the mass window∆M12are detailed in the text, while for theM34are defined through linear interpolation of the values in Table 4.
The event selection criteria, consisting of kinematic selection, isolation criteria and impact parameter
80
significance, are presented in Table 3. The candidate quadruplet is formed by selecting two opposite sign,
81
same flavour di-lepton pairs in an event. The di-lepton pair of the quadruplet closest to the nominal Z
82
boson mass is called the leading di-lepton pair, Z1, while the second di-lepton pair of the quadruplet is
83
the sub-leading,Z2. For each event there is a mass window requirement applied to the invariant mass
84
of each of the two di-lepton pairs. The cut values are chosen event-by-event using the reconstructed
85
four-lepton invariant mass. This results in a unique mass spectrum for each background regardless of the
86
hypothetised Higgs mass. TheZ1is required to be within 15 GeV from the nominal Z mass forM4ℓ<170
87
GeV and within 12 GeV forM4ℓ>180 GeV. TheZ2cut values for a set of four-lepton invariant masses
88
are shown in Table 4, where the actual cut value used is obtained by linear interpolation between these
89
mass points.
90
The track isolation discriminant is defined as the sum of the transverse momenta of tracks inside a
91
cone of ∆R=p
∆η2+∆φ2 <0.3 around the lepton over the lepton pT. Summed tracks are of good
92
quality and pass a minimumpTcut (at least four silicon hits andpT>1 GeV). Each lepton is required to
93
have a normalised track isolation less than 0.20. The calorimetric isolation discriminant is defined as the
94
sum of the calorimeter cells inside an isolation cone of 0.3 around the lepton. In case of electromagnetic
95
showers, the corresponding cells are excluded from the sum. Each lepton is required to have a normalised
96
calorimetric isolation less than 0.30. The impact parameter significance (d0/σd0) is required to be lower
97
than 3.5 for muons and 6 for electrons. The electron impact parameter is affected by Bremsstrahlung and
98
is thus broader. It is noted, that the invariant mass resolution is further improved by applying aZ mass
99
constraint.
100
MH (GeV) 120 130 140 150 160 165 180 190 >200
MZ2 (GeV) 15 20 25 30 30 35 40 50 60
Table 4: Selection requirements applied to the reconstructed massesMℓ3ℓ4 of the sub-leading di-lepton pair shown for different values of the reconstructed mass M4ℓ. The cut values for all other M4ℓ are obtained by linear interpolation.
5 Background estimation methods
101
This section describes two data driven background estimation techniques. The first estimates the ZZ
102
background using the observed Z→ℓℓproduction rate, the relative efficiencies from Monte Carlo and
103
the theoretically estimated cross-section ratio of the two processes. The second evaluates the Z plus
104
heavy quark background from a control sample which is subsequently extrapolated to the signal region
105
using high statistics Monte Carlo. Due to the low statistics of the current data, the latter method will only
106
serve as a cross-check of the Monte Carlo background expectations in the control region.
107
5.1 Estimation of theZZ(∗) background based on the measuredZ→ℓℓproduction rate
108
The prediction of the ZZ(∗) process has a number of systematic uncertainties due to: parton density
109
functions, renormalisation and factorization scales, luminosity determination, and a number of other
110
experimental considerations. The normalisation of the ZZ(∗)with respect to the production of singleZ
111
cancels out the uncertainties due to the luminosity determination and offers partial cancelation for the
112
rest of the uncertainties [20].
113
This method provides a prediction of the production rate ofZZ(∗)based on the observed rate of theZ production. The predicted number ofZZ(∗)events, is obtained by multiplying the MC prediction,NZZMC, by a scale factor, R:
R= NZData
σZεZL, (1)
whereNZDatais the observed number ofZevents,σZis the theoretical inclusiveZcross-section,εZis the
114
experimental efficiency to observeZ→ℓℓandLis the integrated luminosity. Using theZ→µµevents
115
obtained with 35/pbof integrated luminosity it isR=0.95±0.01(stat)±0.02(syst).
116
5.2 Study of the heavy quark component in theZ+jetsbackground
117
One of the important reducible backgrounds to the low mass Higgs, is the four lepton final state of the
118
ZQQ production process, where Q denotes a heavy flavour jet originating from ab orc quark. In the
119
following, “Q→ℓ” and “q→ℓ” will indicate a reconstructed lepton candidate originating from heavy
120
flavour or from a light jet, respectively. The prediction of theZ(QQ→l+l−+X)background contribu-
121
tion is affected by the theoretical uncertainty of theZQQproduction cross section and by uncertainties
122
related to lepton reconstruction withinb-jets. The estimation of this background contribution can be per-
123
formed through control samples in which theZQQ→4ℓfraction is measured. The extrapolation to the
124
signal region is then performed using high statisticsZQQ→4ℓMonte Carlo samples. The control sam-
125
ples are a compromise between highZQQ→4ℓpurity and reasonable statistics. Since the contaminating
126
sources of electrons and muons in the mentioned control region are of a somewhat different nature, two
127
separate control samples are defined for theZ(QQ→µ+µ−+X)andZ(QQ→e+e−+X)processes. The
128
various control samples and resulting background estimations are discussed in the following sections.
129
5.2.1 Z(QQ→µµ)background
130
There are two major sources of muons accompanying the Z boson and contributing to this final state:
131
muons from in-flight pion and kaon decays and punch-through hadrons (q→µ), and muons from heavy
132
quark decays (Q→µ). The contamination due to pion and kaon decays is estimated and subtracted
133
by weighting the ID track according to the probability rate for the pions and kaons to be reconstructed
134
as muons. These rates have been measured with data, using the KS0 →π+π− decays to identify the
135
pions [21], and a 10% systematic uncertainty is attributed.
136
The selection of theZboson candidates is performed following the first pair selection of theH→4ℓ
137
analysis. The selection of accompanying muons is the one used in the analysis without imposing isolation
138
and impact parameter criteria. Owing to the lack of statistics of the final state with aZ candidate and
139
two accompanying muons, the Z+µ final state is also studied. The number of observed events with
140
additional muons above the pT threshold used in the anaysis (7 GeV) is reported in Table 5. The total
141
rate of events perZdecay as well as the expected rate of muons from heavy quark decays are also shown
142
for data and Monte Carlo. Figure 1 shows the muon multiplicity normalised to theZ events after the
143
subtraction of the light quark contamination as well as the pT distribution of the additional muons which
144
are expected to originate from heavy quarks. The observed pT spectrum is well described by Monte
145
Carlo. A lower pT threshold of 3 GeV is used in this figure to make the comparison more quantitative.
146
Events Events/NZ NQ→µ /NZ NQMC→µ /NZ Z+µ 60 (0.23±0.03)×10−2 (0.15±0.03)×10−2 (0.18±0.02)×10−2 Z+µ+µ− 1 (0.04±0.04)×10−3 (0.03±0.04)×10−3 (0.02±0.003)×10−3
Table 5: Number of eventsZ+µ andZ+µ+µ−in data and rates of events per Z decay of all additional muon events as well as the ones estimated to become from heavy quark decays in data and MC. The additional muon accompanying the Z boson satisfies the requirements described in the text and a pT
threshold of 7 GeV..
5.2.2 Z(QQ→ee)background
147
This method estimates the ZQQ component from a control region defined as one Z boson (selected as in
148
the present analysis) and two extra electrons passing the medium or loose selection criteria. This control
149
muons
N -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
ZEvents/N
10-5
10-4
10-3
10-2
10-1
1
∫
L=41.74 pb-1, s = 7 TeVData muons from Q MC muons from Q
(a)
[GeV]
PT
5 10 15 20 25 30
) /2 GeV Z)x(1/N T(dN/dp
10-6
10-5
10-4
10-3
10-2
= 7 TeV s
-1, L=41.74 pb
∫
Data muons from Q MC muons from Q
(b)
Figure 1: Distributions of the additional muon multiplicity after (a) the subtraction of muons originating from light quraks. (b) The expected transverse momentum of muons reconstructed in association with theZ-boson and orginating from heavy quark decays. 3 GeV pT theshold is used.
region is contaminated by light quark components (Zqq and ZqQ). This contamination is estimated from
150
Monte Carlo by first normalizing truth matched Zqq Monte Carlo to the data in another region where the
151
light quark contribution is dominant, and then applying the control region selection to the Monte Carlo.
152
A light quark dominant region can be defined by relaxing the requirements of 3rd and 4th electrons to be
153
track-EM cluster matches which only have an anti-cut applied on one of the shower shape variables, such
154
asRη, the lateral shower leakage. This variable offers good separation between light and heavy quark
155
contributions, where heavy quark electrons peak close to anRη of 1. TheRη distribution forRη <0.7
156
is fully dominated by light quarks. One can thus normalize a Zqq MC sample to the data in the region
157
Rη <0.7 and then use this MC to predict the Zqq yields after Loose or Medium electron cuts. The
158
predicted Zqq yields are given by the following formula:
159
NZqq(Loose or Medium) =NCRZqq×NRDataη<0.7
NRMCη<0.7×εMC(Loose or Medium), (2)
whereNZqqCR is the number of Zqq events in the full control region in MC andεMCis the efficiency for Zqq
160
events passing Loose or Medium. The performance of the method is tested using Z+X MC (closure test).
161
Table 6 shows the predicted ZqX yields after application of Eq.2, for both MC and data. The systematic
162
uncertainty of these expectations are of the order of 20%.
163
As we see from Table 6, there is only one event surviving for Loose selection and no events for
164
Medium, consistent with zero ZQQ events for both selections. Thus, for the purposes of this analysis,
165
and since a consistency between data and Z+X Monte Carlo is found after Loose and Medium selection,
166
the normalized MC itself can be used to predict the Z+X yield in the Higgs signal region.
167
5.2.3 Summary ofZQQstudies
168
The estimated yields forZ(QQ→ee) andZ(QQ→µµ)are in agreement with the observed data. No
169
current evidence supports any disagreement in either the modeling or the normalization of the description
170
Electron Predicted
Quality Zqq+ZqQ Observed
Loose 2±0.3 1
Medium 0.12±0.04 0
Table 6: Predicted ZqX=NZqq+NZqQyields after Loose and Medium electron cuts for the 3rd and 4th lepton, using extrapolation from a high statistics control region (Eq.2). The first two lines demonstrate the validity of the method using Monte Carlo. The last two lines show the result of application of the method on the data where a consitency with expectations is found.
of this background in the MC. Consequently, the results for this background are solely based on Monte
171
Carlo.
172
6 Results on final selection
173
The selection criteria described in Section 4 were applied to the 2010 data and the invariant mass distri-
174
bution of the selected events after aplying the dilepton kinematic cuts are shown in Figure 2. No event
175
is found to survive the full selection. In Table Table 7, the expected signal and backfground yields for a
176
few Higgs mass hypotheses are summarized.
177
(GeV) M4l
100 150 200 250 300 350 400 450 500 550 600
Events/(100 GeV)
10-3
10-2
10-1
1
Ldt =40/pb
∫
DATA 2010 ZZ
t t Z
Figure 2: Mass distribution of four lepton events selected after the di-lepton kinematic cuts.
7 Systematic uncertainties
178
The following sources of systematic uncertainty are considered in this search.
179
Luminosity : An overall normalisation uncertainty of 11% is assumed for the luminosity. This
180
uncertainty is only applied to Monte Carlo samples where the normalisation is not obtained from the
181
data and it is assumed to be fully correlated among these samples.
182
MH Ns Nb (GeV)
150 0.038 0.010 160 0.019 0.013 180 0.026 0.058 200 0.099 0.119 220 0.091 0.108 300 0.063 0.142 400 0.059 0.065 500 0.047 0.051 600 0.009 0.077
Table 7: Expected signal and background event yields within five standard deviations of the Higgs mass distribution.
Cross-sections of the Higgs boson production : The Higgs boson production cross-sections have
183
been studied extensively by the LHC Higgs cross-section working group and the results are compiled in
184
Ref. [4]. The theoretical uncertainties on the cross-sections have been estimated to be between 15−20%
185
forgg→Hand 5% forqq→qqH. Since thegg→Hprocess dominates, a normalisation uncertainty of
186
17% is applied to the signal samples for all mass points.
187
Cross-sections of background processes : A normalisation error of 15% is assigned to the ZZ
188
contribution, which accounts for the theoretical uncertainties in the cross-section calculation and the
189
uncertainty due to the use of thegg→ZZcorrection. For theZbackground the theoretical uncertainty in
190
the cross-section calculation of 5% is used, while theZbb¯ normalisation was found to be consistent with
191
the observation in the data within 20%. The normalisation uncertainty for the tt¯background is taken
192
conservatively to be 25%.
193
Electron Reconstruction and Identification : The electron energy scale uncertainty is found to be
194
less than 1% in most of theη region of interest, while the energy resolution uncertainty is estimated to
195
vary between 0.2% and 0.4%. The reconstruction and identification efficiency uncertainty is 2.5% in the
196
ETregion relevant for electrons fromZ→eedecays.
197
Muon Reconstruction and Identification : The muon momentum scale uncertainty is estimated to
198
1%, while the uncertainty in the momentum resolution is at the few percent level . Finally, the uncertainty
199
on the identification efficiency of muons is estimated to be between 0.5% and 1% for the phase space of
200
interest.
201
Trigger : Owing to the high lepton trigger efficiency and the presence of multiple high pTleptons in
202
the final state, a trigger efficiency very close to 100% is achieved, while the corresponding uncertainties
203
are found to be negligible.
204
8 Exclusion limits on H → ZZ
(∗)→ 4 ℓ
205
The upper limits are extracted using a profile likelihood method described in Ref. [22].
206
(editor’s note: there is an effort to arrive to a common agreement within the Higgs group to use
207
profile likelihood ratios and CLs+b power constrained limits. We plan to have these results soon. The
208
present CLs limits will be kept as a reference)
209
(GeV) MH
100 200 300 400 500 600
SMσ /σ95% CL Upper Bound on
1 10 102
→ 4l
→ ZZ H
σ
± 1 σ
± 2 L dt = 40 pb-1
∫
Figure 3: The 95% confidence level upper limit on the SM Higgs production cross section expressed in multiples of the SM expectation usingCLs.
9 Summary
210
A search for the Standard Model HiggsBoson in the decay channelH→ZZ(∗)→4ℓhas been performed
211
using the first 40 pb−1ofppcollision at LHC. The lepton identification and reconstruction has been de-
212
scribed as well as the expected efficiencies, lepton energy scales and resolutions corrected in accordance
213
to the latest ATLAS performance studies and recommendations. Data driven background extraction
214
methods were established and tested on the first data. An agreement with the Monte Carlo background
215
estimation was found within the uncertainties. More statistics is needed for the analysis to fully profit
216
from the a more accurate background measurement. The analysis proceedure was explained, the sys-
217
tematic uncertainties incorporated and a limit extracted. A median limit of 28 times the Standard Model
218
cross-section is achieved for a Higgs boson mass of 200 GeV.
219
References
220
[1] ATLAS Collaboration, G. Aad et al.,The ATLAS Experiment at the CERN Large Hadron Collider,
221
JINST3(2008) S08003.
222
[2] T. Sjostrand, S. Mrenna, and P. Z. Skands,PYTHIA 6.4 Physics and Manual, JHEP05(2006) 026,
223
arXiv:hep-ph/0603175.
224
[3] P. Golonka and Z. Was,PHOTOS Monte Carlo: A Precision tool for QED corrections in Z and W
225
decays, Eur. Phys. J.C45(2006) 97–107,arXiv:hep-ph/0506026.
226
[4] LHC Higgs Cross Section Working Group, S. Dittmaier, C. Mariotti, G. Passarino,
227
R. Tanaka (Eds.), et al.,Handbook of LHC Higgs Cross Sections: 1. Inclusive Observables,
228
arXiv:1101.0593 [hep-ph].
229
[5] M. Ciccolini, A. Denner, and S. Dittmaier,Strong and electroweak corrections to the production of
230
Higgs+2jets via weak interactions at the LHC, Phys. Rev. Lett.99(2007) 161803,
231
arXiv:0707.0381v1 [hep-ph].
232
[6] M. Ciccolini, A. Denner, and S. Dittmaier,Electroweak and QCD corrections to Higgs production
233
via vector-boson fusion at the LHC, Phys. Rev. D77(2008) 013002,arXiv:0710.4749v1
234
[hep-ph].
235
[7] A. Bredenstein, A. Denner, S. Dittmaier, and M. M. Weber,Precision calculations for H→
236
WW/ZZ→4fermions with PROPHECY4f,arXiv:0708.4123 [hep-ph].
237
[8] T. Binoth and N. K. P. Mertsch,Gluon-induced QCD corrections to pp→ZZ→l anti-l l’ anti-l’,
238
arXiv:0807.0024 [hep-ph].
239
[9] M. L. Mangano, M. Moretti, F. Piccinini, R. Pittau, and A. D. Polosa,ALPGEN, a generator for
240
hard multiparton processes in hadronic collisions, JHEP07(2003) 001,
241
arXiv:hep-ph/0206293.
242
[10] S. Frixione and B. R. Webber,The MC@NLO event generator,arXiv:hep-ph/0207182.
243
[11] S. Frixione and B. R. Webber,Matching NLO QCD computations and parton shower simulations,
244
JHEP06(2002) 029,arXiv:hep-ph/0204244.
245
[12] S. Frixione, P. Nason, and B. R. Webber,Matching NLO QCD and parton showers in heavy
246
flavour production, JHEP08(2003) 007,arXiv:hep-ph/0305252.
247
[13] G. Corcella et al.,HERWIG 6.5 release note,arXiv:hep-ph/0210213.
248
[14] J. M. Butterworth, J. R. Forshaw, and M. H. Seymour,Multiparton interactions in
249
photoproduction at HERA, Z. Phys.C72(1996) 637–646,arXiv:hep-ph/9601371.
250
[15] S. Agostinelli et al.,GEANT4: A simulation toolkit, Nucl. Instrum. Meth.A506(2003) 250–303.
251
[16] K. Melnikov and F. Petriello,Electroweak gauge boson production at hadron colliders through
252
O(αs2), Phys. Rev.D74(2006) 114017,arXiv:hep-ph/0609070.
253
[17] S. Catani, L. Cieri, G. Ferrera, D. de Florian, and M. Grazzini,Vector boson production at hadron
254
colliders: a fully exclusive QCD calculation at NNLO, Phys. Rev. Lett.103(2009) 082001,
255
arXiv:0903.2120 [hep-ph].
256
[18] J. M. Campbell and R. Ellis,An Update on vector boson pair production at hadron colliders,
257
Phys.Rev.D60(1999) 113006,arXiv:hep-ph/9905386 [hep-ph].
258
[19] M. Aliev et al.,– HATHOR – HAdronic Top and Heavy quarks crOss section calculatoR,
259
arXiv:1007.1327 [hep-ph].
260
[20] J. Campbell, E. Castaneda-Miranda, Y. Fang, N. Kauer, B. Mellado, et al.,Normalizing Weak
261
Boson Pair Production at the Large Hadron Collider, Phys.Rev.D80(2009) 054023,
262
arXiv:0906.2500 [hep-ph].
263
[21] ATLAS Collaboration,Muon Reconstruction Performance, ATLAS-COM-CONF-2010-064
264
(2010) .
265
[22] ATLAS Collaboration,Higgs Boson Searches using the H→WW →ℓ+νℓ−ν¯ Decay Mode with
266
the ATLAS Detector at 7 TeV, ATLAS-COM-CONF-2010-115 (2010) .
267