on behalf of the atlas collaboration
DESCRIPTION
A T L A S ________________________________________________________________________________________________ GMSB SUSY models with non-pointing photons signatures. On behalf of the ATLAS collaboration. ~. gravitino G get its mass only through gravitational interaction - PowerPoint PPT PresentationTRANSCRIPT
Damien Prieur - Moriond QCD 2005 1
A T L A S________________________________________________________________________________________________
GMSB SUSY models with non-pointing photons GMSB SUSY models with non-pointing photons signaturessignatures
On behalf of the ATLAS collaborationOn behalf of the ATLAS collaboration
Damien Prieur - Moriond QCD 2005 2
NLSP :NLSP : • neutralinoneutralino : 01 + G
• sleptonslepton : lR l + G~~
~ ~ NLSP NLSP typical decay lengthtypical decay length may be macroscopic: may be macroscopic:
54
0NLSP
GeV 100
TeV 100 B100
1 )cm(cτ
NLSPm
F54
0NLSP
GeV 100
TeV 100 B100
1 )cm(cτ
NLSPm
F
Gauge Mediated Supersymmetry Breaking (GMSB)Gauge Mediated Supersymmetry Breaking (GMSB)
gravitino G get its mass only through gravitational interaction
G is the LSP, 10-2 < mG < 104 eV
if R-parity is assumed, G is the only stable SUSY particle
all SUSY decay chains lead to the production of gravitinos
~
~
~
• Minimal GMSB model phenomenology is driven by 6 parameters
• SUSY breaking is mediated to the visible sector via gauge interactions
F, MF, Mmessmess, N, N55, , , tan, tan, sign(, sign(μ)μ)
Damien Prieur - Moriond QCD 2005 3
• We consider here cases where LSP lighest neutralino
• Measuring ccττ is a fair way to access to FF00, a free parameter of the model
• To reconstruct decay vertex :• Impact position• Photon direction • Time of arrival ttγγ
• From these informations it is possible to determine gravitino’s directiongravitino’s direction ψψ :
2
2
1
1cos
sin
)cos1(
L
Lct with
χ 01
~
Gauge Mediated Supersymmetry Breaking (GMSB)Gauge Mediated Supersymmetry Breaking (GMSB)
• missing ET
• non pointingnon pointing photons photons
• With an intermediateintermediate lifetime, we get the following event topologies :
χ 01
~
01 + G~ ~
Damien Prieur - Moriond QCD 2005 4
OutlineOutline
How to reconstruct photon direction with EM calorimeter
How to reconstruct SUSY GMSB signal and determine m and m in ATLAS
original method developed by Kawagoe, Kobayashi, Nojiri, OchiKawagoe, Kobayashi, Nojiri, Ochi (hep-ph/0309031)
using a fast simulationfast simulation of ATLAS detector
more realistic performancesrealistic performances parametrisation for non pointing photonsnon pointing photons
Estimate ATLAS sensitivity to ccττ and FF00
χ 01
~l~
• In the following:
Damien Prieur - Moriond QCD 2005 5
CalorimetersCalorimeters Electromagnetic
Hadronic
Solenoïd 2T
ATLAS detectorATLAS detector
Damien Prieur - Moriond QCD 2005 6
Electromagnetic calorimeterElectromagnetic calorimeter
Electromagnetic calorimeter - EndCapElectromagnetic calorimeter - EndCap
CryostatCryostat
Electromagnetic calorimeter - BarrelElectromagnetic calorimeter - Barrel
Damien Prieur - Moriond QCD 2005 7
MiddleMiddle
FrontFront
Z
r
0
gen
EM showerEM shower 60 GeV60 GeV
True direction: gengen
Reconstructing photon directionReconstructing photon direction
LayerLayer GranularityGranularity (x)
Pre-sampler
Front
Middle
Back
0.025 x 0.10.025 x 0.1
0.003 x 0.10.003 x 0.1
0.025 x 0.0250.025 x 0.025
0.05 x 0.0250.05 x 0.025
Damien Prieur - Moriond QCD 2005 8
MiddleMiddle
FrontFront
Z
r
0
Reconstruct barycenter :
EE
i
ii
_
for each layer : 11 , 22
11
22
Reconstructed shower direction : recrec
rec
True direction: gengen
Parametrisation of shower depth: R1(11) , R2(22)
R2(22)
R1(11)
Reconstructing photon directionReconstructing photon direction
Damien Prieur - Moriond QCD 2005 9
EM Calorimeter angular resolutionEM Calorimeter angular resolution
• EM calorimeter designed for pointing photonsdesigned for pointing photons (coming from the IP - σz=5.6 cm) resolution on polar angle polar angle θθ : σθ 6060 mrad/E(GeV) (barrel, ||<1.4) resolution direction direction : not possible because of coarse granularity
& not necessary for normal processes (σr=15 μm)
• Performance for non pointing photons studied from detailled simulationdetailled simulation (Geant3)
barrelbarrel
resolution parameterized and resolution parameterized and implemented in the fast simulation of implemented in the fast simulation of ATLAS detector ATLAS detector
used in the following studyused in the following study
systematic biassystematic bias with standard reconstruction algorithms
need specific correctionsspecific corrections to improve resolution
Damien Prieur - Moriond QCD 2005 10
Reconstruction of m and mReconstruction of m and m
• We consider the following decay chain :
χ01
~l~
• Gravitino’s energy EEGG is unknown, but we can determine its direction (coscosψψ)
• EEll , EEγγ , cosψcosψ , cosθcosθGlGl , and cosθcosθlγlγ being all measured, we can compute aa and bb parameters for each event.
• Some kinematics…
ammbl
2~
2~ 0
1 ammb
l
2~
2~ 0
1• In the (a,b)(a,b) plane, points should group along the line
)cos1(2)( ~2
~2~0
1
GG
EEppm
bam
EEmE
E
pppm
lllGl
lGl
2~
2~
~
2~
2~
01
01
)cos1(2)cos1(
)cos1(1
)(
lχ01
~ll
~γ G
~
Damien Prieur - Moriond QCD 2005 11
• GMSB point G1G1
Λ M N5 Tanβ μ
90 TeV 500 TeV 1 5 +1
• 105 SUSY events generated with HERWIGHERWIG (pp pp sparticles sparticles 8 pb) one year of LHC at low luminosity (13.9 fb-1)
• Fast simulation for all particles
χ01
~
162
324
~
~
R
L
l
l
m
m
218
117
02
01
~
~
m
m
Sparticules Masses (GeV)
456
436
04
03
~
~
m
m
Event generation and simulationEvent generation and simulation
• energy, position, direction and time of non pointing photons are smeared according to realistic resolutions from test-beam datatest-beam data or detailed simulationdetailed simulation:
• Energy : EMCAL
• Timing : EMCAL
• Position : EMCAL
• Direction : • along : EMCAL previously parametrized resolutionpreviously parametrized resolution• along φ: TRT
%7.0 MeV245%10
EEE
E
ps 100t
)GeV(
004.0
E )GeV(
mrad 5
E
mrad 1
barr
el +
end
cap
barr
el +
end
cap
Will require converted photons !!!Will require converted photons !!!
Damien Prieur - Moriond QCD 2005 12
4,3,2,1, TTTTmissTeff ppppEM
• Pre-selection cuts to limit background contribution from standard modelFull background analysis not performed
Use cuts from I. Hinchliffe/ F.E. PaigeI. Hinchliffe/ F.E. Paige (hep-ph/9812233 )
GeV 400effM
effmissT ME 1.0
Pre-selection cutsPre-selection cuts
- require 4 jets- build an effective masseffective mass :
We require :We require :
+ 2 leptons
+ 2 photons
71%71%
22%22%
efficiencyefficiency
Further require non pointing photonnon pointing photon
•We choose lifetime ccτ = τ = 100100 cm cm χ01
~
Damien Prieur - Moriond QCD 2005 13
Apply cuts to select non pointing photons candidates
Resolution on Resolution on ψψ
EEγγ > 30 GeV > 30 GeVEEγγ > 30 GeV > 30 GeV αα > 0.2 rad > 0.2 radαα > 0.2 rad > 0.2 rad
ΔΔttγγ > 1ns > 1nsΔΔttγγ > 1ns > 1nsσσψψ= 60 mrad= 60 mrad
Selection of non pointing photonsSelection of non pointing photons
ψrec-ψtrue (mrad)t (ns)
94%94%
24%24%
74%74%
(conservative)(conservative)
Damien Prieur - Moriond QCD 2005 14
• For each NP photon, we select if possible an isolated lepton with ppTT>20 GeV>20 GeV to form lγ pairs
• If several leptons available, we choose the one which minimizes mlγ
At this point, we end up with 300300 lγ pairs
ammbl
2~
2~ 0
1 ammb
l
2~
2~ 0
1
p0: 26111p1: 13635
m = 116.8116.8 GeV
m = 161.6161.6 GeV
χ01
~
l~
Reconstruction of m and mReconstruction of m and mχ01
~l~
Scatter plot - (a,b) plane
linear fit :
Need more work for improving pairing
InputInput
161.4161.4 GeV GeV
117.1 GeV
Damien Prieur - Moriond QCD 2005 15
Resolution on reconstructed masses is about 2%2% for ccτ = τ = 100100 cm cm
Input mass : 161.4161.4 GeV
Reconstructed : 159.9159.9 GeV
σ σ :: 2.1 2.1 GeVGeV
Input mass : 117.1117.1 GeV
Reconstructed : 115.5115.5 GeV
σ σ :: 1.7 1.7 GeVGeV
Reconstruction of m and mReconstruction of m and mχ01
~l~
Simulation is repeated 100100 times to estimate resolution on reconstructed sparticle masses
preliminary
Damien Prieur - Moriond QCD 2005 16
Reconstruction of NLSP decay vertexReconstruction of NLSP decay vertex
Reconstruct decay time of NLSP in lab. frame σσttDD 2 ns 2 ns
• same cuts as previously, except photon conversion not not required required
• for each NP photon, we select if possible an isolated lepton with ppTT>20 GeV>20 GeV to form lγ pairs
B
BppEE
A
AppEE ll
01
01
01
01
~~~~• We minimize
• Once we have m and m No longer need toNo longer need to : - know photon’s direction along direction along
- require a photon conversionphoton conversion
χ01
~l~
Αmmmpp lll 22~
2~~ 0
101 2
1~~
Bm
pp 2
~~2~
~
01
01
once NLSP decay position is known, we can reconstruct all the decay chain
Damien Prieur - Moriond QCD 2005 17
Reconstruction of NLSP lifetimeReconstruction of NLSP lifetime
make lifetime cc vary from 10 up to 200 cm fit of proper timeproper time ttDD// by an exponential function simulation reproduced 100 times to estimate sensitivity on fitted values in this case we need lot of statistics 10 fb-1 is not enough, use 100 fb-1 instead (1 year @ high lumi)
Systematic biais for cc > 100 cm (due to increasing amount of NLSP escaping detection) Need to know momentum distribution to correct for bias at high c cc/c/c vary from 33 to 88%
χ 01
~
8%8%
3%3%
cc
/c/c
Damien Prieur - Moriond QCD 2005 18
• A realistic angular resolutionrealistic angular resolution parameterization for non pointing photonsnon pointing photons implemented in a fast simulationfast simulation of ATLAS detector
• Resolution on reconstructed masses for NLSP and slepton masses about 2%2%
• Resolution on NLSP lifetime below 8%8% for cc between 10 and 200 cm
SummarySummary
SystematicsSystematics and acceptanceacceptance need to be carefully studied with a detailed simulationdetailed simulation of ATLAS :
- effects on direction reconstruction of pile-up, underlying events- description of EM showers- contributions from background
PR
EL
IMIN
AR
Y
%2
m
m%5
c
c%4
0
0
F
F%8
~
~
G
G
m
m
Can extend accessible cc rangerange using statistical method (but need careful studies - under investigation)
Can determine SUSY fundamental breaking scale FF00 with 4%4% precision
for cc = 100 cm
Damien Prieur - Moriond QCD 2005 19
Damien Prieur - Moriond QCD 2005 20
Readout electrode (barrel)Readout electrode (barrel)Front layerMiddle layer
Back layer=0 =0.8
=1.4
Pre samplerr
EM calorimeter granularityEM calorimeter granularity
LayerLayer GranularityGranularity (x)
Pre-sampler
Front
Middle
Back
0.025 x 0.10.025 x 0.1
0.003 x 0.10.003 x 0.1
0.025 x 0.0250.025 x 0.025
0.05 x 0.0250.05 x 0.025
Damien Prieur - Moriond QCD 2005 21
Reconstruction of NLSP lifetimeReconstruction of NLSP lifetime
• Having m and m , we are able to reconstruct the slepton decay chain
No longer need toNo longer need to : - know photon’s direction along direction along - require a photon conversionphoton conversion inside the inner detector
χ01
~l~
lχ01
~ll
~γ G
~
Αmmmpp lll 22~
2~~ 0
101 2
1~~
Bm
pp 2
~~2~
~
01
01
c
AM
c
OMt
22~
~~~ll
ppp
2~2 ~~~
pppG
- photon impact positionimpact position- photon polar anglepolar angle
we measure :
few relations :
B
BppEE
A
AppEE ll
01
01
01
01
~~~~
we minimize
CalorimeterCalorimeter
Damien Prieur - Moriond QCD 2005 22
Production of sleptonsProduction of sleptons
q
χ02
~l~ χ0
1
~
ll
pp
ppg~
G~
q
q~
χ02
~l~
l
pp
ppq~
q
χ02
~ l~
l
pp
pp