Yet another statistically insignificant hint of New Physics

or

ZZ cross section measurement with4.6 fb-1 of 7 TeV ATLAS data

arxiv:1211.6096, submitted to JHEP

Alexandru Dafinca, David Hall, Roger Naranjo, Edgar Valencia, Ivonne A. Maldonado Cervantes, Gonzalo Diaz, Jose La Rosa, Margot Delgado de la Flor, Alejandro de la Puente, Estefania Coluccio Leskow, Gilberto Chavez Meza, Vanessa Theodoro,

Gibraham Napoles, Andrés Melo

CLASHEP-13, Arequipa, Peru 2

Motivation

Tests electroweak physicsBackground to Higgs search

Set limits on anomalous triple gauge couplings (aTGC)

neutral TGCs not in SM

ATLAS diboson x-sections measured high at 7 TeV

17/03/2013

CLASHEP-13, Arequipa, Peru 3

The ATLAS detectorInner detector• within solenoid magnet• pixels for vertex location• silicon detector tracks charged particles• |η| < 2.5

EM & hadronic calorimeters• measures energy deposited• |η| < 4.9

Muon spectrometer• toroidal magnet• tracks muons• |η| < 2.7 η = - ln tan(θ/2)

Level Rate

Initial 40 MHz

L1 100 kHz

L2 3 kHz

L3 200 Hz

3-level trigger system

17/03/2013

CLASHEP-13, Arequipa, Peru 4

Analysis overviewpp → ZZZZ (*) → ℓℓℓℓ ZZ → ℓℓνν

Both on-shellZZ → ℓℓℓℓ One off-shellZZ * → ℓℓℓℓChannel Classification mll

ZZ → ℓℓℓℓ On-shell 66 – 116 GeV

Off-shell > 20 GeVZZ → ℓℓνν 76 – 106 GeV

17/03/2013

Identification of Z bosons

Resolve ambiguous cases by selecting the combination that minimises the sum of |mll - mZ| for the two pairs

CLASHEP-13, Arequipa, Peru 5

ZZ(*) → ℓℓℓℓ event selection• pass electron or muon trigger

• exactly 4 isolated leptons with pT > 7 GeV

– e+e-e+e-, μ+μ-μ+μ-, e+e-μ+μ-

• require at least one muon with pT > 20 GeV or

electron with pT > 25 GeV

• each same-flavour, opposite-charge dilepton system forms a Z candidate

• events sorted into ZZ and ZZ* categories

17/03/2013

CLASHEP-13, Arequipa, Peru 6

ZZ → ℓℓνν event selection

17/03/2013

• pass electron or muon trigger

• exactly 2 isolated leptons with pT > 20 GeV: e+e-, μ+μ-

• Z candidate in mass window

• veto event if jet pT > 25 GeV (reduce top background)

• discriminating MET variables to reduce Z+jets bkg:

CLASHEP-13, Arequipa, Peru 7

How to get an inclusive σ from data

17/03/2013

LC

NN=σ

ZZ

bkgobsfidZZ

BFLCA

NN=σ

ZZZZ

bkgobstotalZZ

totalMCσ fid

MCσ recoMCσMC

Inclusive

Data

Fiducial region(truth level)

Signal region(reco. level)

ZZAmeasure ZZCmeasure

obsNfidZZσ

totalZZσ

• σfid – cross section measured within a restricted (kinematic and geometric) phase space

• σtotal – cross section extrapolated to total phase space

CLASHEP-13, Arequipa, Peru 8

Systematic uncertainties

17/03/2013

But how do we obtain Nbkg?

CLASHEP-13, Arequipa, Peru 9

• give rise to “lepton-like jets” (lepton candidates failing isolation criteria)

• background: ℓℓℓj and ℓℓjj with mis-ID jet(s)

• define

• measure f based on lepton-like jets in a sample of Z+jets in data/MC in bins of pT

• total number of expected background events:

Backgrounds to ZZ(*) → ℓℓℓℓ

17/03/2013

Leptons from HF decay can pass ID cuts

Other mis-ID sources:• leptons from π0, K decay• jets misidentified as leptons

CLASHEP-13, Arequipa, Peru 10

Backgrounds to ZZ(*) → ℓℓℓℓ

17/03/2013

Expected number of background events < 1 for ZZ → ℓℓℓℓ

CLASHEP-13, Arequipa, Peru 11

Backgrounds to ZZ → ℓℓνν• tt, WW, Wt and Z → ττ

– two true isolated leptons with missing pT

• extrapolate from a control sample of eμ events

• WZ– 3 lepton final state, but one lepton is not identified

• estimated using MC• validated using a control region of 3 high-pT leptons

• Z+jets– may have missing pT due to mis-measurement of jet pT

• estimated using high-pT γ+jets events

• Z+jets, W+jets, Zγ, Wγ– jet or γ mis-identified as lepton

• use the “matrix method” to estimate the fraction of events in the signal region containing at least one fake lepton

17/03/2013

CLASHEP-13, Arequipa, Peru 1217/03/2013

Backgrounds to ZZ → ℓℓνν

Distributions in signal region using background estimations in previous slide

CLASHEP-13, Arequipa, Peru 13

Cross section results

Measured (fb) NLO prediction (fb)

17/03/2013

fid'' llllZZ

)lumi(0.1)syst()stat(4.25 2.10.1

3.30.3

)theory()stat(1.09.20 1.1

9.0

fid

* '' llllZZ )lumi(2.1)syst()stat(8.29 7.1

5.18.35.3

)theory()stat(1.06.25 3.1

1.1

fid

llZZ )lumi(5.0)syst()stat(7.12 7.17.1

1.39.2

)theory()stat(1.05.12 0.1

1.1

totalZZ pb )lumi(3.0)syst()stat(7.07.6 4.0

3.0 pb )theory(89.5 22.0

18.0

• maximum likelihood fitting method• treat systematics as nuisance parameters• single Gaussian across channels for correlated uncertainties

CLASHEP-13, Arequipa, Peru 14

aTGC limits

17/03/2013

CP conserving

CP violating

• aTGCs described by 4 parameters fVi (with fV

i = 0 in SM)

• aTGCs produce longer tail in pTZ spectrum

• fit pTZ distribution

• limits set using maximum profile likelihood ratio

95% CL

CLASHEP-13, Arequipa, Peru 15

Conclusions• total and fiducial ZZ cross sections measured with full 2011

dataset• used ZZ(*) → ℓℓℓℓ and ZZ → ℓℓνν decay channels• statistically consistent with SM NLO prediction, although

measured value is slightly higher– analysis recently performed in 4ℓ channel with 20 fb-1 of

8TeV data: ATLAS-CONF-2013-020

17/03/2013

CLASHEP-13, Arequipa, Peru 1617/03/2013

CLASHEP-13, Arequipa, Peru 17

We thank our sponsors

17/03/2013

Backup slides