sparticle reconstruction at benchmark points

Recontres de Moriond – QCD and Hadronic Interactions

La Thuile, 28 March – April 2004

Sparticle reconstruction at benchmark Sparticle reconstruction at benchmark points points

Alessia TricomiAlessia TricomiOn behalf of ATLAS and CMS Collaboration

Dipartimento di Fisica e Astronomia and INFN Catania

QCD and Hadronic Interactions, 28 March - 4 April 2004Alessia Tricomi

OutlineOutline

In the past years several inclusive studies done to understand the detector capabilities to discover SUSY

Counting excess of events over SM expectationsNo explicit sparticle reconstruction doneSeveral different final states analysed

Etmiss + jets

No lepton + Etmiss + jets

1 l + Etmiss + jets

2 l OS + Etmiss + jets

2 l SS + Etmiss + jets

3 l + Etmiss + jets

Scan in mSUGRA (m0,m1/2) plane (also studies in p-MSSM scenario performed )Fast MC simulation used: ATLFAST, CMSJET

Today I will focus on recent studies to reconstruct strongly interacting SUSY particles

S.Abdullin, F. Charles Nucl. Phys. B547 (1999) 60S. Abdullin et al., J. Phys. G28 (2002) 469M. Dzelalija et al., Mod. Phys. Lett. A15 (2000) 465


If Supersymmetry exists, If Supersymmetry exists, LHC will probably observe itLHC will probably observe it

Cosmologically interesting Cosmologically interesting region covered with 10 fbregion covered with 10 fb-1-1

Squark-gluino production Squark-gluino production dominates the total cross dominates the total cross section at low mass scalesection at low mass scale

Squarks and Gluinos Squarks and Gluinos detectable up to 2 TeV detectable up to 2 TeV mass with 100 fbmass with 100 fb-1-1

Inclusive SUSY reach vs integrated luminosityInclusive SUSY reach vs integrated luminosity

A0=0, tan=35, >0

S.Abdullin, F. Charles Nucl. Phys. B547 (1999) 60S. Abdullin et al., J. Phys. G28 (2002) 469M. Dzelalija et al., Mod. Phys. Lett. A15 (2000) 465

Expected squark-gluino mass reach

1.5 – 1.8 TeV with 10 fb-1

2.3 – 2.5 TeV with 100 fb-1

2.6 – 3.0 Tev with 300 fb-1

2.8 – 3.2 TeV with 1000 fb-1


It is not enough to observe the excess over the Standard Model…

Fix a set of points in the parameter space

Get information on the spectrum (i.e. end-points)

Reconstruct sparticles

DISCOVERYDISCOVERY SUSY SPECTROSCOPYSUSY SPECTROSCOPYThis requires a different approach…This requires a different approach…

But …But …


Decay chainsDecay chains

2 high pt isolated leptons OS

(leptons = e,)

2 high pt non-b jets

missing Et

2 high pt isolated leptons OS

(leptons = e,)

2 high pt b jets

missing Et

g~ pp

b~02

~

qq~ g~ pp

q~02

01

~~

SM bkg: tt, Z+jet, W+jet, ZZ, WW, ZW, QCD jets

Reconstruction of sbottoms, squarks and Reconstruction of sbottoms, squarks and gluinosgluinos

p

p

g~

b~

b

b

01

~02

~~

bb~

01

~


Benchmark pointsBenchmark points

Rather low mRather low m00 and m and m1/21/2 values in order to have values in order to have high SUSY cross sectionhigh SUSY cross section

Three differentThree different tantan values (tanvalues (tan = 10, 20, = 10, 20, 35) since BR35) since BR((22

00 ll++ll-- 1100) )

depends on tandepends on tan

Proposed Post-LEP Benchmarks for Supersymmetry, M. Battaglia Proposed Post-LEP Benchmarks for Supersymmetry, M. Battaglia et alet al. (hep-ph/0106204). (hep-ph/0106204)


Point B spectraPoint B spectra

95.610 = LSP

173.81±136.2lR

361.62±196.5lL

174.420520qR

339.930559qL

361.140524.0bR

575.9tR496.0bL

392.9tL595.1g

Point B

pb .TOTSUSY 7757

10

0100

250

00

21

tan

)(

A

sign

m

m

bb~

g~ pp

b~02

01~

(17% bL, 10% bR)

(37 % bL, 25% bR)

(0.04 %) 0

1~ ~(16.4 %)

01

~ ~ (83.2 %)

2 isolated leptons, pT>15 GeV, ||<2.4

2 b-jets, pT>20 GeV, ||<2.4

Sbottom reconstruction


p

p

g~

b~

b

b

01

~

02

~ ~

10 fb-1

First step: First step: 2200 l l++ll- - 11

00

02

~ 01

~

~

~

2

χ~2~

2~

2

χ~max

M

)M)(MM(MM

01

02

)μM(μ )eM(e - SUSY events


Bkg reductionBkg reduction

SUSY

ttbar

Z+jets

SM bkg can be strongly reduced cutting on ET

miss

(GeV) )μM(e -)μM(e - )μM(μ )eM(e --

Edge = 79 2 GeVGenerated = 78.2 GeV

Edge = 79 2 GeVGenerated = 78.2 GeV


Second step: sbottom (squark) Second step: sbottom (squark) reconstructionreconstruction

p

p

g~

b~

b

b

01

~

02

~ ~

• Assuming M(Assuming M(1100) known) known

• Selecting events “in edge”Selecting events “in edge”

• Combining theCombining the 2200 obtained obtained

from the two leptons with the from the two leptons with the most energetic b-jet in the most energetic b-jet in the eventevent

At the end-point:At the end-point:

GeV 80 )M( GeV 65

p

M

M1p

01

02

~

~


Squark mass peakSquark mass peak

GeV 906σ

GeV 10536q)χ~M( 02

Result of the fit:Result of the fit:

1 fb-1

GeV 0.537)~()u~M(

GeV 8.542)~()d~

M(

L

L

L

L

cM

sM

Generated valuesGenerated values

pb 4.1)chain decayq~pp(BR pb 5.0)chain decayq~g~pp(BR

fb 60chain)decay qBR(σ

pb 2chain)decay qBR(σ

R

L

~

~


Sbottom mass peakSbottom mass peak

The peak should be considered The peak should be considered as the superposition of two as the superposition of two

peakspeaks

10 fb-1

GeV 542σ

GeV7 500b)χ~M( 02

Result of the fit:

GeV 524.0)b~

M(

GeV 496.0)b~

M(

R

L

Generated masses:

pb 1.0)chain decayb~

pp(BR pb 6.0)chain decayb

~g~pp(BR

fb 212chain)decay bBR(σ

fb 535chain)decay bBR(σ

2

1

~

~

GeV 503.9)b

~BR(σ)b

~BR(σ

)b~

BR(σ)b~

M()b~

BR(σ)b~

M()b

~(M

21

2211


p

p

g~

b~

b

b

01

~

02

~ ~

Gluino reconstructionGluino reconstruction

10 fb-1

sbottom sbottom chainchain

10 fb-1

squark squark chainchain

GeV 1.955)g~M( GeV 742

GeV7 495bb)~M( 02

GeV 575

GeV7 295qq)~M( 02

Generated Generated

value:value:

Two separated gluino mass measurements, with Two separated gluino mass measurements, with two different samplestwo different samples


estimateestimate)b~

M(-)g~M(

GeV 471

GeV 392b)~M(-bb)~M( 02

02

The reconstruction is performed assuming M(The reconstruction is performed assuming M(1100) known but …) known but …

The difference between the two masses is independent of M(The difference between the two masses is independent of M(1100))

GeV 1.71)b~

M(-)g~M(

GeV 1.99)b~

M(-)g~M(

R

L

Result of the fit:Result of the fit:

Generated value:Generated value:

Worse resolution, but model independent resultWorse resolution, but model independent result


Results @ point BResults @ point B

GeV 441σ

GeV 4502b)χ~M( 02

60 fb-1

GeV 336σ

GeV 2497b)χ~M( 02

300 fb-1

GeV 80 )M( GeV 70 GeV 80 )M( GeV 75

60 fb-1

GeV 346

GeV 4295bb)~M( 02

GeV 339

GeV 1195bb)~M( 02

300 fb-1


GeV 136σ

GeV 2532q)χ~M( 02

GeV 131σ

GeV 1536q)χ~M( 02

60 fb-1300 fb-1


Sbottom Chain Squark Chain

10 fb-1 60 fb-1 300 fb-1 10 fb-1 60 fb-1 300 fb-1

M(sbottom)

500±7 502±4 497±2 M(squark) 536±10

532±2 536±1

(sbottom) 42±5 41±4 36±3 (squark) 60±9 36±1 31±1

M(gluino) 594±7 592±4 591±3 M(gluino) 592±7 595±2 590±2

(gluino) 42±7 46±3 39±3 (gluino) 75±5 59±2 59±2

M(gl)-M(sb)

92±3 88±2 90±2 M(gl)-M(sq)

57±3 47±2 44±2

gl-sb) 17±4 20±2 23±2 gl-sq) 9±3 16±5 11±2

Squark mass peak can be reconstructed in the first few weeks (resolution ~12%)Sbottom and gluino in the first year (resolution ~6÷8%)Two independent gluino mass measurementsThe resolutions can be improved with larger statistics (~5÷6% at 300 fb-1)

Errors are of the order of 1÷2 GeV (statistical) + 2÷3 GeV (energy scale of the calorimeters)The main source of error is from the lack of knowledge on M(1

0)

M(10) assumed known for these reconstructions

No systematic errors evaluated


Reconstruction @ point GReconstruction @ point G

10 fb-1

300 fb-1

sbottom chain

squark chain

With respect to the Point B:With respect to the Point B:

Lower total SUSY cross-sectionLower total SUSY cross-section

Lower BR of useful decaysLower BR of useful decays

Higher BR’s of competitive Higher BR’s of competitive decaysdecays

Lower BR of the last decayLower BR of the last decay

16.44%) Bpoint (

%26.2)~~(BR 0

2

%)48.27 Bintpo(

%58.26)bb~

g~(BR %)7.326 Bintpo(

%92.59)qq~g~(BR

%)19.2 Bintpo(

%52.21)Wt~b~

(BR 2

B)point at allowed(not

%83.7)W~~(BR 011

B) point @ pb57 ( pb 25.8TOTSUSY


Repeating the same procedure as Repeating the same procedure as Point B:Point B:

Results @ Point GResults @ Point G

sbottosbottomm

gluingluinoo

squarsquarkk

)~

()~( bMgM

gluingluinoo

)~()~( qMgM

Sbottom Chain

300 fb-1

M(sbottom) 720±26

(sbottom) 81±18

M(gluino) 851±40

(gluino) 130±43

M(sb)-M(gl) 127±10

sb-gl) 48±11

Squark Chain

300 fb-1

M(squark) 774±9

(squark) 84±9

M(gluino) 853±11

(gluino) 126±11

M(sq)-M(gl) 82±3

sq-gl) 35±3


Reconstruction @ point IReconstruction @ point I

%25.0)~~(BR 0

2

BBGG

300 fb-1

Neutralino BR’s

B G I

BR(20→ll) 16.4

42.26 0.25

BR(20→ 83.2

275.8

898.13

IITau channel becomes predominant at large tan Tau-pair edge is not as sharp as in the e and case, but could help to cover points in which the reconstruction is problematicIt could be exploited in regions with too low leptonic BR: work in progress both in ATLAS and in CMS


Di-tau lepton edgeDi-tau lepton edge

As l identification is not possible, must rely on hadronic decays – narrow, 1-prong jets (large QCD bkg though)

_

Can typically achieve /jet ~100 for ~50-60 %

Narrow isolated jets selection :

Rjet = 0.2, Risol = 0.4

Require 0.8 GeV < Mjet < 3.6 GeV (biased against 1-prong, but improves di-tau mass resolution - less neutrino momentum)Di-tau efficiency = 41 %

Mvis = 0.66 MAdditional cuts :

min. 4 jets : ET > 100

GeV,

ET > 50 GeV

missing ET > 100 GeV,

no e, with pT > 20 GeV

1

2- 4

30 fb-130 fb-1

SM bkgSM bkg

Real from SUSYReal from SUSY

Fake from SUSYFake from SUSY

expectedexpectedvisiblevisible

ATLAS Physics TDR study (full GEANT simulation) example (“Point 6”) :

Recent results – even more encouraging…

45tan

0

)(

200

200

6int

0

0

21

A

sign

m

m

Po

10tan

100

)(

100

250

1A-SPS Point

00

21

A

sign

m

m


2. Using mass of lightest neutralino and RH sleptons can discriminate between SUSY models differing only in slepton mass.

SUSY Dark Matter

3. Lightest Neutralino LSP excellent Dark Matter candidate.– Test of compatibility between LHC

observations and signal observed in Dark Matter experiments.

Why and how to measure the Why and how to measure the 1100 mass… mass…

1. Use as starting point for other sparticle mass measurements (sbottom, gluino, squark…)

1.E-06

1.E-05

1.E-04

1.E-03

1 10 100 1000 10000

Mass GeV

-N

ucl

eon

cro

ss s

ectio

n p

b

EDELWEISSZEPLIN I

CDMS

IGEX

DAMA

10-3

10-4

10-5

10-6

Allanach et al., 2001


ATLAS

PhysicsTDR

1% error(100 fb-1)

Point 5

e+e- + +-

30 fb-1

atlfast

Physics TDR

ATLASPoint 5

ATLAS PhysicsTDR

2% error(100 fb-1)

Point 5ATLAS

PhysicsTDR

1% error(100 fb-1)

Point 5

Why and how to measure the Why and how to measure the 1100 mass… mass…

Long decay chains allow multiple measurements

Use kinematics to measure end-points (in general both maximum and minimum value due to 2-body decays)

Starting point : di-lepton OS SF mass edge

Assume 2 hardest jets are from squarks,combine each of these with leptons to from :

Mllq

qL~

~

q

~

R

Also used endpoint from Mhq from h bb -

Turns out to have enough constraints to determine all masses involved (!)

Can measure mass relations to ~1% as a function of LSP mass, determined to ~10 %

Threshold : Tllq, requiring Mll > Mll / 2max

Use dilepton signature to tag presence of 02 in event, then work

back up decay chain constructing invariant mass distributions of combinations of leptons and jets

Endpoints:

> <Mlq, Mlq

~

~

hb

b-


Systematic uncertainties not evaluatedMain sources: Th: theoretical distributions of edges Ex: detector resolution (i.e. energy

resolution, b/-tagging efficiency, bias due to cuts applied…

Mass ReconstructionMass Reconstruction

Mllq High Mlq

Low Mlq

Mll

Mllq Mhq

Combine measurements from edges from different jet/lepton combinations

Numerical solution of simultaneous edge position equations

Gives sensitivity to masses

Sparticle Expected precision (100 fb-1)

qL 3%

02 6%

lR 9%

01 12%

~

~

~

~

ATLAS Physics TDRPoint 5

01 lR

02 qL

Mass (GeV)Mass (GeV)

Mass (GeV)Mass (GeV)

~

~

~

~

Allanach et al., 2001

ATLAS ATLAS

ATLAS ATLAS

Point 5Point 5

Point 5 Point 5

Same order precision also @ point SPS1A (similar to “CMS” point B)

2tan

300

)(

100

300

5 Point

00

21

A

sign

m

m

10tan

100

)(

100

250

1A-SPS Point

00

21

A

sign

m

m

Gjelsten et al., ATLAS-PHYS-2004-07


Conclusions Conclusions LHC experiments are expected to explore SUSY in a decisive way. The plausible part of mSUGRA-MSSM parameter space will be explored in a number of characteristic signaturesStrongly interacting SUSY particles can be accessed up to 2TeV for 100 fb -1

Information on the SUSY spectrum achievable , with favourable SUSY parameters, already after the first months of data taking

Low tan region (like point B, SPS1A, point 5):

first few weeks of LHC running period: reconstruction of squark

(resolution ~12%)first year: reconstruction of sbottom and

gluino (resolutions ~6÷8%) reconstruction of gluino in the

squark chain (independent channel)

high integrated luminosity: improvement on the resolutions double fit of the sbottom peak

Intermediate tan (i.e. point G):first year: dilepton edge hardly visible no reconstruction possible in e,

channelhigh integrated luminosity: reconstruction of squarks, sbottom

(~11%) and gluino (~15%)High tan (i.e point I):

no reconstruction possible in the e- channel even with high accumulated statisticsReconstruction in the tau channel exploited. Work is going on…

New analyses are going on: New benchmark points to be analyzed Full reconstruction studies in first priority Multi-edge technique to be fully exploited Deeper insight to the tau-tau channel Systematic uncertainties to be evaluated

Lot of work before LHC start-up!!!

sparticle reconstruction at benchmark points

Documents