ATLAS SUSY Results - University of Oxford Department of Physics

Report
Searching for Supersymmetry
using the Higgs boson
Andrée Robichaud-Véronneau
Oxford University
Outline

Higgs discovery and its consequences

Why Supersymmetry?

[email protected]

Search for SUSY decaying to Higgs

Summary and Outlook
The Standard Model of elementary
particles


The best description of matter and forces to date
Validated by precision measurements over a large range
of energy scales
Matter made from quarks and
leptons
4 elementary forces with their
carriers:
- Electromagnetic (g)
- Weak Nuclear (W, Z)
- Strong Nuclear (g)
- Gravity (?)
"We found a new boson”


July 4th, 2012: Announcement
of the discovery of a new
boson consistent with the
Higgs boson
Mass measured using ZZ->4l
and gg signatures:
126.0 ± 0.4 (stat.) ± 0.4 (syst.) GeV


Combination of all channels: ZZ, WW, gg, tt, bb, using 7
and 8 TeV dataset from ATLAS
Boson properties compatible with the Standard Model
Higgs
Phys. Lett. B 716 (2012) 1-29
Nobel prize winners!
2013 Nobel prize in Physics awarded to Prof.
Higgs and Englert "for the theoretical discovery of a
mechanism that contributes to our understanding of the origin of
mass of subatomic particles, and which recently was confirmed
through the discovery of the predicted fundamental particle, by
the ATLAS and CMS experiments at CERN's Large Hadron
Collider”
Is that the whole story?*

Not quite. We still have a few
unanswered questions:
 Matter/Antimatter imbalance
 What is Dark Matter?
 Hierarchy problem
 ...
*Respecting Hincliffe's rule
"To infinity... and beyond!” ©
SUPERSymmetry




Introducing a new symmetry
of spacetime and fields
Heavier superpartners with
spin-½ compared to the SM
MSSM: 105 parameters to be
determined!
Introducing R-parity (aka matter parity)
 SM particles (+1), SUSY particles (-1)
 Phenomenology centered around the Lightest
Supersymmetric Particle (LSP)
 If conserved, protects against proton decay
How can SUSY help?

In many ways:
 Provides a dark matter candidate (LSP)
 Cancel Higgs mass corrections using
sparticle loop
 Unifies all forces
Now, how do we go about
to look for it?
Large Hadron Collider
Using the largest, coolest machine in the world!




Proton-proton
collider at 8
TeV (soon 14)
High luminosity
(~1034 cm-2s-1)
4 interaction
points – 7
experiments
ATLAS


Hermetic multipurpose
particle detector

Inner tracking

Calorimetry

Muon detection
High precision and
granularity (~100 million
channels)
Allow to measure passage of
charged particles, leptons,
photons, muons and jets
LHC performance
N= σ L


Good data-taking efficiency for the whole dataset and
excellent work from the LHC team!
Multiple interactions for each proton bunch crossing →
pile-up
ATLAS reconstruction
ATLAS performance
Excellent muon reconstruction
efficiency over large range of
momentum and pseudorapidity
Electron reconstruction
efficiency greatly improved from
2011 (red) to 2012 (blue)
ATLAS performance



Jets can be tagged
for heavy flavour,
such as b or c
quarks
Correction factor
(data/MC) to btagging efficiency
Excellent agreement
of data and
simulation over
large energy ranges
SUSY search strategy in ATLAS




Top and bottom
(charm) squarks
Electroweak
production
Cross section

Strong production
Various scenarios of symmetry breaking, violation of Rparity or exotic long-lived particles considered
We look in every corner!
Higgs-aware SUSY




Knowledge of the mass of the SM Higgs
provides constraints in the SUSY models
It also gives information on the couplings of
the SM Higgs to sparticles
All 3 main production types can be probed
using Higgs in their signatures
We'll focus here on the electroweak
production
The SUSY Higgses

MSSM: Contains 5 Higgses, one of which is
the SM Higgs (h0)
SUSY Electroweak production



R-parity conserving models → Production of
sparticle in pair
Electroweak production means sleptons,
charginos and neutralinos, the SUSY partners
of the weak bosons of the SM
Order by index in mass → decreasing cross
section with increasing mass
Chargino-neutralino production
ATLAS-CONF-2013-093
Chargino-neutralino production


Considering the case of lowest mass states allowing
the production of a Higgs boson (Dm[χ02- χ01] > 130
GeV)
Favoured in certains area of the MSSM
parameter space
Choosing h0 → bb, since it has the
highest branching ratio.
The lepton in the W decay helps to
reduce QCD background
The LSPs generate large amount
of missing energy
ATLAS-CONF-2013-093
Signal simplified model



Simplified models
used to generate
signal points
m
∞
Settings BR to 100%
(for non-SM
processes)
Adjusting parameters
to obtain one single
process (3 params for
electroweak
production: M1, M2, m
M2
M1
60 GeV
0 GeV
ATLAS-CONF-2013-093
Signal grid



Simplified models
used to generate
signal points
Each red dot
represent a model
Using degenerate
masses between χ±1
and χ02.

Scanning χ02 mass.
ATLAS-CONF-2013-093
SM Backgrounds


Many SM process have similar signatures that the
one we are looking for in our signal

tt: WbWb with one W decaying to ln

tt+V: Smaller cross section

Single top: Mainly Wt mode

W/Z+jets: Contribution from jets mistag

Diboson: W(ln)W(qq) mostly

W/Z+H: SM process, not missing energy
Modelled using Monte Carlo simulation
ATLAS-CONF-2013-093
Event selection

Using ATLAS recommendations for physics objects reconstruction

Define baseline objects

Jets with pT > 20 GeV

Leptons (e or m) with pT > 10 GeV

Apply cleaning cut for detector defects

Reject overlapping objects (e, m, jets) in the same detector area

Extra overlap removal between e and m


DRe-m < 0.1, DRm-m < 0.05
Events are triggered by single lepton requirements

Electrons: EF_e24vhi_ medium1 || EF_e60_medium1

Muons: EF_mu24i_tight || EF_mu36_tight
ATLAS-CONF-2013-093
Event selection


From the baseline object, signal objects are selected

Leptons are isolated, with pT > 25 GeV

Central jets with pT > 25 GeV, |h| < 2.4

Forward jets with pT > 30 GeV, 2.4 < |h| < 4.5
Preselection

2 highest pT central jets

1 baseline && 1 signal lepton

Missing transverse energy (ETmiss) > 100 GeV

Nsignal_jets < 4
ATLAS-CONF-2013-093
Event selection

Targetted signal cuts

0, 1 or 2 jets to be tagged as coming from a b quark
(among the 2 highest pT jets)

mjj > 50 GeV (for the 2 highest pT jets)

Contransverse mass (mCT) > 160 GeV

Transverse mass (mT) at varying thresholds for
background estimation and signal measurement
ATLAS-CONF-2013-093
Signal region optimisation
Optimise analysis
selection cuts
based on the
mass splitting
regions
ATLAS-CONF-2013-093
Signal region optimisation



Two signal regions: SRA at low mass splittings, SRB for
high mass splittings
SRA (SRB): mT >100 (130) GeV (on top of previous mCT)
and ETmiss cuts).
Optimised for 105 < mbb < 135 GeV
SRA
ZN= √
2 erf − 1(1− 2pvalue )
SRB
ATLAS-CONF-2013-093
Signal predicted yields


SRA has high yields in low mass splitting regions due to
cross section and high a x e in the high mass splitting region
SRB consistently has high yields and a x e in high mass
splitting region
ATLAS-CONF-2013-093
Background kinematics



Distributions
scaled using
background fit
results
ETmiss cut
applied, all
other three
variables
untouched
Main
background
contribution
from tt before
selections cuts
ATLAS-CONF-2013-093
Background estimation

Strategy:





Reducible background: estimate from data
Irreducible background: validate MC simulation with
data
Use control regions (close kinematically to data, but
designed to target background processes) to obtain scale
factors to fit MC simulation to data
Use validation regions to validate fit (obtain good
agreement between data and simulation using fit results
above)
Apply normalisation to signal regions to get background
estimate
ATLAS-CONF-2013-093
Control and validation regions



Cut above
applied to the
entire plane
mbb binning for
all regions: 5075, 75-105,
105-135, 135165, > 165 GeV
*: signal bin not
considered in
backgroundonly fit
ATLAS-CONF-2013-093
Systematic uncertainties

Lepton (electron or muon) energy scale, resolution, identification and trigger

Jet energy scale and resolution, JVF

ETmiss resolution

Btagging calibration

Luminosity

Pile-up

Generator uncertainties

ISR/FSR

Parton shower

Scale uncertainties

Background s uncertainty

Signal s uncertainty
ATLAS-CONF-2013-093
Profile Likelikood Fit
Model dependent fit
Background only fit



Using only control regions
without Higgs bin

Obtain normalisation
factor for two main
background, tt and W+jets

Used for model
independent limits
Using all bins of control
and signal regions
Obtain normalisation
factor for two main
backgrounds and the
signal strength for each
signal point on the grid
ATLAS-CONF-2013-093
Data/MC comparison
ATLAS-CONF-2013-093
Data/MC comparison
ATLAS-CONF-2013-093
Data/MC comparison
Data and SM expectations in excellent agreement → No SUSY (yet)
ATLAS-CONF-2013-093
Signal region yields
SRA (Higgs bin)
Observed
SRB (Higgs bin)
4
2
tt
2.8 ± 2.8
1.0 ± 0.7
W+jets
0.7 ± 0.4
0.3 ± 0.2
Single top t-channel
0.26 +0.27-0.26
0
Single top Wt-mode
1.4 ± 1.3
0.6 ± 0.4
Z+jets
0.01 +0.02-0.01
0.00 +0.01-0.00
Diboson
0.01 +0.05-0.01
0.05 +0.07-0.05
WH
0.18 ± 0.10
0.12 ± 0.07
tt + V
0.01 ± 0.01
0.11 ± 0.06
Total
5.2 ± 3.0
2.0 ± 0.7
(130,0) GeV
6.5
0.2
(225,0) GeV
1.9
4.1
Background estimate
Signal prediction
ATLAS-CONF-2013-093
Results interpretation

No SUSY found. What do we do next?



This is precious information! It should be used to
“quantify our ignorance”
The same way a discovery like the Higgs boson add
additional constraints on theories, using this
information, we can rule out mass range for specific
models → feedback to phenomenologists
Perform likelihood fit using signal and control regions (all
bins)
ATLAS-CONF-2013-093
Model independent limits


Limits on new (non-SM) physics processes that would
have been observed if existed
Estimated using asymptotic formula and pseudoexperiments (”toys”) - results consistents
SRA
SRB
Observed s95vis (Asymptotic)
0.29 fb
0.22 fb
Expected S95exp (Asymptotic)
6.7 +3.1-1.9
4.6 +2.5-1.5
Observed s95vis (Pseudo-experiments)
0.31 fb
0.22 fb
Expected S95exp (Pseudo-experiments)
6.8 +2.7-1.4
4.4 +1.8-0.8
ATLAS-CONF-2013-093
Exclusion contour



Contour
interpolated
from individual
values of CLs of
each model
Small grey
numbers: cross
sections
excluded fpr
each point
Compute limits
using -1s line
ATLAS-CONF-2013-093
Exclusion limits in 1D
Where do we stand?
Where do we stand?
χ±1χ02→ W± (l±n) χ01 h0(bb) χ01
Summary and Outlook


The ATLAS experiment, together with the
LHC, had a very successful first run!
The Higgs boson discovery has opened new
pathways to clear out, looking for SUSY



Completing the spectrum of available decays
In our search for new physics at the TeV
scale, no excess has been observed over the
SM background so far
Looking forward to see what 14 TeV
collisions will reveal!
Backup Slides
pMSSM
T. Rizzo
BNL
13 Sep. 2012

similar documents