Moriond QCD and Hadronic Interactions,
La Thuile, Italy, 17th March 2005
First Detailed Study on the SUSY
Discovery Potential at CMS using Two
Same Sign Muons in the mSUGRA Model
David Futyan
University of California, Riverside, USA
David Futyan
UC Riverside
Principal contributors:
Alexey Drozdetskiy, Darin Acosta, Guenakh Mitselmakher
University of Florida, Gainesville, USA1
Moriond QCD and Hadronic Interactions
17th March 2005
Introduction
Inclusive search within framework of the mSUGRA model
Aim is to determine reach within mSUGRA parameter space for
potential discovery of supersymmetry soon after LHC start-up
(∫L=10fb-1)
Use same sign muon signature
First detailed look
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
2
mSUGRA parameter space
mSUGRA is a popular simplification of the MSSM (GUT sub-model)
Only 5 free parameters (m0, m1/2, tanβ, A0, sign(μ))
A0 = 0,
J.Ellis et al., hep-ph/0303043
me
David Futyan
UC Riverside
 old cosmological constraint
2
0.1 <  x h < 0.3
 WMAP cosmological constraint
0.094 <  x h 2 < 0.129
0
 require that 1 is the LSP
 excluded by b  s
 favored by g   2 at 2-σ level
Moriond QCD and Hadronic Interactions
17th March 2005
3
SUSY points considered in this study
350
6000
m0
10
300
5000
250
4,6,8,10,11,19,20
4000
17
6
200
3000
2000
2
12
1000
7
0
15
9
14 19
1350
20
0
190
1 3
0
16
4
150
100
500
1
2*
8
3
4
5
6
7
8
11
9
10
11
12
390
440
13
14
15
2500
3000
m1/2 16*
17
18
19
20
18
240
290
340
5
1000
1500
2000
Updated post-WMAP benchmark points
(hep-ph/0306219) [* - with modification]
More points
(calculated by A.Birkedal (Cornell), K.Matchev (UF))
Six of the points (high mass) are found to have negligible
cross-sections
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
m12
500
300
700
210
850
285
700
360
900
240
400
300
500
1000
1000
1500
2000
2500
250
300
m0
107
1330
149
150
181
210
2155
230
2530
330
85
1200
1620
2520
2715
3442
4192
4942
60
65
tanβ
10
10
10
10
10
10
10
10
10
20
10
35
35
35
10
10
10
10
10
10
- negligible σ
4
Signature choice
For this study the "2 Same Sign Muons" signature was chosen, based
on:
Theoretical studies for Tevatron
• hep-ph/9904282, "Supersymmetry Reach of the Tevatron via Trilepton, LikeSign Dilepton and Dilepton plus Tau Jet Signatures", K.T.Matchev, D.M.Pierce
Simple, very clean signature
High trigger efficiency
Reduced number of background events/processes in comparison to
"multi-jets only" signatures
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
5
Analysis overview
Tools
Generators: σ (LO/NLO), coupling constants, matrix elements
Generators: showering, event development
Full detector simulation
Trigger emulation
Full events reconstruction
Analysis: optimization of significance and S/B
Choice of cut variables
cuts optimization
results, error analysis (stat. & system.)
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
6
Tools
Calculation of coupling constants, cross sections for SUSY processes:
ISAJET
http://www.phy.bnl.gov/~isajet/
Calculation of NLO corrections for SUSY processes: PROSPINO
hep-ph/9611232
Event generation: CompHEP (Z/γ*bB, Single-top processes),
PYTHIA (SUSY, tt, ZZ, ZW, WW)
hep-ph/9908288 (CompHEP)
http://www.thep.lu.se/~torbjorn/Pythia.html
Full CMS detector simulation was used in this study: CMSIM, ORCA
http://cmsdoc.cern.ch/cmsim/cmsim.html
http://cmsdoc.cern.ch/cmsreco/
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
7
Cross sections, number of events: SM processes
tb
spb
N1
N2
0.212*
2,120
112
tqb
5.17*
51,700
1,798
tb
0.129*
1,290
71
tqb
3.03*
30,300
1,067
ZZ ZW WW tt
18(NLO)
180,000
256
26.2
262,000
727
70.2
702,000
39.7
Zbb
886(NLO) 232(NLO)*
8,860,000 2,320,000
142,691
12,924
All
160,000
Other processes considered at generator level only (CompHEP)
Estimate of preselected
eventscontribution
based on s.BR:
main
WWW ZWW
spb
N1
N2
spb
N1
N2
ZZW
into background
ZZZ WWWW ZWWW ZZWW ZZZW
0.129
1,290
<15
0.0979
979
<10
0.0305
305
<3
0.00994
99.4
<1
ttW
ttZ
ttWW
ttZW ttZZ
0.556
5,560
<200
0.65
6,500
<200
neg.
neg.
0.000574
neg.
0.000706
0.000442
0.000572
ZZZZ
0.0000161
- negligible
contribution
Notations:
all
butfor
, ttZ luminosity
are negligible
ttW
N1 – total number of expected
events
integral
of 10fb-1
N2 – number of events after pre-selection (two same sign muons, each with
PT>10 GeV, |h|<2.5)
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
8
Cross sections, number of events: SUSY processes
s, pb
N1
N2 (NLO)
Significance
S/B
s, pb
N1
N2 (NLO)
Significance
S/B
1
1.21
1,210
470
1.2
0.0029
11
2.27
22,700
961
2.4
0.006
2
2.43
24,300
1470
3.6
0.0092
12
2.77
27,700
2,210
4.6
0.014
3
0.161
1,610
66
0.16
0.00041
13
0.214
2,140
188
0.46
0.0012
4
83.2
832,000
14,600
35.8
0.091
SUSY point number:
5
6
0.0511
17.6
511
176,000
20.7
4,330
0.05
10.8
0.00013
0.027
7
0.0354
354
18,3
0.04
0.00011
8
5.21
52,100
1,520
3.8
0.0095
9
0.00911
91
2.71
10
31.8
318,000
11,700
28.8
0.073
14
0.00527
52.7
SUSY point number:
15
16
0.00504 0.00048
50.4
4.8
17
0.00006
0.6
18
0.000008
0.08
19
40.8
408,000
9,200
22.6
0.058
20
15.9
159,000
4,570
11.4
0.029
- excluded points
Notations:
N1 – total number of expected events for integral luminosity of 10fb-1
N2 – number of events after pre-selection (two same sign muons, PT>10 GeV)
Significance, S  2 N  N  N
(S.I.Bityukov,N.V.Krasnikov)
12
S
B
B
S/B – ratio: N / N

S
David Futyan
UC Riverside

B
Moriond QCD and Hadronic Interactions
17th March 2005
9
Example signal and background processes
Signal…
1 d
 L 
Background…
0


1


W b
pp  tt  X
pp  guL  X
t1t 
1 b
W b
 L  
Cut variables:
 Y
  
10 
Missing ET
Jet ET
Muon PT, Muon impact parameter
Plus: Muon isolation, Muon η, Jet η, Number of jets/muons, …
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
10
Event kinematics (example)
SUSY point (#3):
m0=149 GeV, m1/2=700 GeV, tanβ = 10, A0 = 0, signμ > 0
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
11
Analysis cuts
For the chosen cut variables, a set of values was chosen for
optimization:
Missing ET: 0, 100, 150, 200, 250, 400, 500 GeV
ETjet1: 0, 70, 100, 200, 300, 400 GeV
ETjet3: 0, 30, 50, 80, 100, 170, 250 GeV
PTμ1: 10, 20, 30, 60, 100, 150 GeV/c
PTμ2: 10, 15, 20, 50, 80 GeV/c
IPμmin: N/A, 0.005, 0.0015, 0.0005 cm
IPμmax: N/A, 0.1, 0.03, 0.01, 0.005 cm
For each possible combination of the above cuts:
Values of Significance, S/B and expected no. of events (NFinal) for 10 fb-1 are
calculated
Total of about 176,000 combinations
Final sets of cuts selected through an iterative optimization procedure
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
12
Illustration of cut optimization procedure
Plot Significance for all sets of cuts…
Then choose an "optimal" region: several iterations…
Finally, choose a particular set…
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
13
Analysis cuts
Two sets of cuts chosen which have high significance and S/B for all
SUSY points
Set #1:
• Missing ET > 200 GeV,
• ETjet3 > 170 GeV,
• PTμ1 > 20 GeV
Set #2:
• Missing ET > 100 GeV,
• ETjet1 > 300 GeV,
• ETjet3 > 100 GeV
These cuts are applied in addition to the "signature cut":
2 same sign muons, both with PT > 10 GeV:
Results were obtained for all SUSY points for each of the above sets of
cuts
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
14
Results: cut set #2
SET 2
N final
Signif
S/B
SM
432 ± 8.8
1
184 ± 9.3
8.06
0.43
2
560 ± 29
21.4
1.3
SET 2
7
8
N final 8.31 ± 0.39 530 ± 28
Signif
0.4
20.5
S/B
0.019
1.2
9
n/a
n/a
n/a
SET 2
N final
Signif
S/B
16
n/a
n/a
n/a
14
n/a
n/a
n/a
15
n/a
n/a
n/a
3
4
5
6
30.4 ± 1.4 1590 ± 152 9.6 ± 0.45 1030 ± 67
1.44
48.4
0.46
35
0.07
3.7
0.002
2.4
10
11
1950 ± 151 322 ± 18
56.1
13.4
4.5
0.75
17
n/a
n/a
n/a
18
n/a
n/a
n/a
12
781 ± 42
28.1
1.8
13
86.9 ± 4
4
0.2
19
20
1220 ± 106 996 ± 67
39.8
34
2.8
2.3
Monte Carlo statistical errors shown for number of events after all cuts
(N final)
All final events accepted by L1 and HLT
L1: single μ with PT>14 GeV, di-μ with PT > 3 GeV
HLT: di-μ with PT>7 GeV
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
15
Results: significance
Total number of points out of reach
(Significance < 5) for 10 fb-1
9 for cut set #1
10 for cut set #2
Remainder are potential "discovery
points" for 10 fb-1 of integrated luminosity
Significance > 5
S/B > 0.4 (excess of 40% or more over
expected number of SM events)
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Significance
SET 1
SET 2
9.05
8.06
20.8
21.4
2
1.44
25
48.4
0.77
0.46
20.6
35
0.78
0.4
15.5
20.5
n/a
n/a
31.7
56.1
12.1
13.4
27.1
28.1
6
4
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
n/a
25.6
39.8
20.6
34
16
Results: sensitive area at 10 fb-1
350
6000
m0
10
300
5000
18
250
4,6,8,10,11,19,20
4000
150
16
3000
7
2000
2
1000
9
13
15
100
14
50
20
11
0
190
1 3
0
4
19
12
0
17
200
8
6
500
240
290
340
390
440
5
1000
1500
2000
2500
m3000
1/2
Many points will be visible with ∫L<<10 fb-1
Significance for many points >> 5 for ∫L=10 fb-1
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
17
First estimate of systematic effects
To estimate the stability of the results, the effect of varying the no. of SM
and SUSY events was investigated:
Apply simultaneously:
30% decrease in no. of SUSY events
30% increase in no. of SM events
• Only one background process survives after all cuts: tt
• Precision to which cross-section is expected to be known for this process
(including theoretical systematics) is about 10%
Only one "SUSY discovery" point (#13) goes out of reach
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
18
Summary
Study performed based on the mSUGRA model
tanβ=10,20,35, sign(μ)>0, A0=0
Full detailed simulation, trigger emulation and reconstruction was used
Excess of mSUGRA events over SM processes is statistically significant
for many benchmark points for ∫L<<10fb-1
up to 600 GeV in m1/2 and at least up to 1600 GeV in m0
Results are optimistic for SUSY discovery
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
19
Prospects
Further optimization
Investigation of use of additional cut variables for S/B and Significance
optimization, e.g.:
• μ isolation
• b-tagging
• η of jets, η of muons
Possible extension of sensitivity area
Additional backgrounds
QCD multi-jet production (4b, 4c, 2b2c, Wcc, Zcc, Wbb) has been found to
contribute significantly to the background
Investigation of additional points at high m0, low m1/2 to determine extent
of sensitive region in m0
Detailed study of systematic errors
Possibility to study discovery potential for other SUSY models
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
20
Acknowledgements
Principal Contributors:
Alexey Drozdetskiy
Darin Acosta
Guenakh Mitselmakher
}
University of Florida
Thanks also to:
Salavat Abdullin (ITEP/FNAL)
Andreas Birkedal (Cornell)
Nancy Marinelli (University of Athens)
Konstantin Matchev (UF)
Luc Pape (CERN)
Albert de Roeck (CERN)
Alexander Sherstnev (MSU)
Michael Spira (PSI)
Grzegorz Wrochna (Soltan Institute for Nuclear Studies)
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
21
BACKUP SLIDES
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
22
Results, Cuts Set #1
SET 1
N final
Signif
S/B
SM
69.5±6.0
1
95.9±6.7
9.05
1.38
2
282±20
20.8
4.06
3
17.7±1.1
2
0.25
4
365±73
25
5.26
5
6.54±0.37
0.77
0.094
6
277±35
20.6
4
SET 1
N final
Signif
S/B
7
6.7±0.35
0.78
0.096
8
188±17
15.5
2.71
9
n/a
n/a
n/a
10
515±78
31.7
7.41
11
137±11
12.1
1.98
12
409±30
27.1
5.89
13
58.8±3.3
6
0.85
SET 1
N final
Signif
S/B
14
n/a
n/a
n/a
15
n/a
n/a
n/a
16
n/a
n/a
n/a
17
n/a
n/a
n/a
18
n/a
n/a
n/a
19
377±59
25.6
5.43
20
279±36
20.6
4.01
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
23
SUSY points considered in this study
6000
m0
18
5000
4,6,8,10,11,19,20
17
4000
16
3000
7
2000
2
12
1000
0
14
13
1 3
0
15
9
500
5
1000
1500
2000
2500
m3000
1/2
Updated post-WMAP benchmark points
(hep-ph/0306219) [* - with modification]
More points
(calculated by A.Birkedal (Cornell), K.Matchev (UF))
Six of the points (high mass) are found to have negligible
cross-sections
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
1
2*
3
4
5
6
7
8
9
10
11
12
13
14
15
16*
17
18
19
20
m12
500
300
700
210
850
285
700
360
900
240
400
300
500
1000
1000
1500
2000
2500
250
300
m0
107
1330
149
150
181
210
2155
230
2530
330
85
1200
1620
2520
2715
3442
4192
4942
60
65
tanβ
10
10
10
10
10
10
10
10
10
20
10
35
35
35
10
10
10
10
10
10
- negligible σ
24
Results: sensitive area at 10 fb-1
6000
m0
18
5000
4,6,8,10,11,19,20
17
4000
16
3000
7
2000
2
15
9
14
13
12
1000
1 3
0
0
500
5
1000
1500
2000
2500
m3000
1/2
Many points will be visible with ∫L<<10 fb-1
Significance for many points >> 5 for ∫L=10 fb-1
David Futyan
UC Riverside
Moriond QCD and Hadronic Interactions
17th March 2005
25