ww scattering studies at a future linear collider · plans for a future linear collider ww...

< ↑ ↓ >

WW scattering studies at a Future Linear Collider

A presentation given at the University of Edinburgh - High Energy Physics Group.

Andres F. Osorio-Oliveros

[email protected]

High Energy Physics Group

The University of Manchester

Andres Osorio - Edinburgh, 14/10/2005 – p.1

< ↑ ↓ >

Outline

Motivations- Physics (why WW scattering?)- Why a FLC?

Plans for a Future Linear Collider

WW Scattering Analysis

Z and W reconstruction

Results

Sensitivities

Conclusions


< ↑ ↓ >

Motivations

The mechanism describing how Naturegives mass to particles remains one of theopen questions in particle physics today.

The Higgs Mechanism is the answer in theStandard Model

If there is no a Higgs, new physics isneeded at the TeV scale to restore unitarity

It is in this context, that the strong scatteringof WLWL bosons provides a window tolook for information about the underlyingsymmetry.

WL

WL

WL

WL


< ↑ ↓ >

Motivations

The mechanism describing how Naturegives mass to particles remains one of theopen questions in particle physics today.

The Higgs Mechanism is the answer in theStandard Model

If there is no a Higgs, new physics isneeded at the TeV scale to restore unitarity

It is in this context, that the strong scatteringof WLWL bosons provides a window tolook for information about the underlyingsymmetry.

WL

WL

WL

WL

The EW can be described by the EW ChiralLagrangian.

This is an effective theory which:- has operators of higher dimensions- introduces anomalous couplings

In particular there are two 4D operators:L4 = α4

16π2 tr(VµVν)tr(V µV ν)

L5 = α5

16π2 tr(VµV µ)tr(VνV ν)

The coefficients α4 and α5 are related tothe scale of the new physics ( in the SMthese parameters are 0)


< ↑ ↓ >

Motivations

From WW scattering studies at LHC(Butterworth,et.al-Phys.Rev.D65:096014):- EW Chiral Lagrangian- Unitarisation protocols⊲ Prediction of resonances depending onthe values of the α4 and α5 parameters

As an example:�4 = 0:000 , �5 = 0:003 (arXiv hep-ph: 0201098)

To sum up:

What’s the sensitivity to the α4 and α5 thatcan be reached at a Future Linear Collider?Can I improve previous analysis on thesubject?(Ref: Chierici, et al - LC-PHSM-2001-038)

Given that these parameters can bemeasured, what can we learn about newphysics at higher energies?( LHC - LC complementarity )


< ↑ ↓ >

Motivations

The LC scenario


< ↑ ↓ >

Motivations

The LC scenario The LHC scenario


< ↑ ↓ >

A Future Linear Collider

A global collaboration and project (ILC home page http://www.linearcollider.org)

electron sources(HEP and x-ray laser)

linea

rac

cele

rato

rlin

ear

acce

lera

tor

x-ray laser

electron-positron collisionhigh energy physics experiments

positron source

aux. positron and2nd electron source

damping ring

damping ring

positronpreaccelerator

e-

e+

e-

H.Weise 3/2000

33 k

m


< ↑ ↓ >


Signal consists of the following processes:

e+e− → νν̄W+W− → νν̄qq̄qq̄

e+e− → νν̄ZZ → νν̄qq̄qq̄

e-

e+

νe

νe

qq

qq

W-

W+

W-/Z

W+/Z

Scenario: α4 = α5 = 0.0 (SM) Higgs → ∞

Backgrounds:e+e− → νν̄qq̄qq̄ (non-res.)e+e− → e+e−W+W−

e+e− → e+νW−Z

e+e− → W+W−(ZZ)

e+e− → qq̄


< ↑ ↓ >


Signal consists of the following processes:

e+e− → νν̄W+W− → νν̄qq̄qq̄

e+e− → νν̄ZZ → νν̄qq̄qq̄

e-

e+

νe

νe

qq

qq

W-

W+

W-/Z

W+/Z

Scenario: α4 = α5 = 0.0 (SM) Higgs → ∞

Backgrounds:e+e− → νν̄qq̄qq̄ (non-res.)e+e− → e+e−W+W−

e+e− → e+νW−Z

e+e− → W+W−(ZZ)

e+e− → qq̄

Framework:

WHiZard 1.29 is the main event generator:- 6 fermions final states- all possible quark final states weregenerated- need to apply cuts to separate signalsfrom other processes- the anomalous α4 and α5 quarticcouplings are included- beam polarisation


< ↑ ↓ >


We followed TESLA project specifications:(TeV Energy Super-conductingLinear Collider Accelerator)- C.M.E.:

√s = 800 GeV

- Polarised beams:0.80 e−, 0.40 e+ (PoWER Group)- Luminosity: L = 1000 fb−1

> L(1034 cm−2s−1)=5.8

Both ISR and FSR are turned On

Summary of the cross sections obtainedfrom our study:Type Generated pro ess: e

+e−

→ Cross Se t. [fb℄ Generator6 fermions W+W−νν̄ → qq̄qq̄νν̄ 9.21 WhizardZZνν̄ → qq̄qq̄νν̄ 4.05 Whizard

qq̄qq̄νν̄ (ba kgrounds) 6.55 WhizardWZeν → qq̄qq̄eν 38.50 Whizard

e+e−W+W−

→ e+e−qq̄qq̄ 234.80 Pythia*4 fermions W+W−

→ qq̄qq̄ 1948.10 Pythia*ZZ → qq̄qq̄ 142.90 Pythia*

e+e−tt̄ 1.30 Pythia*2 fermions tt̄ → X 136.90 Pythia *qq̄ 4464.60 Pythia** Pythia does not in lude a e+e− polarised beams option during integration


< ↑ ↓ >


We followed TESLA project specifications:(TeV Energy Super-conductingLinear Collider Accelerator)- C.M.E.:

√s = 800 GeV

- Polarised beams:0.80 e−, 0.40 e+ (PoWER Group)- Luminosity: L = 1000 fb−1

> L(1034 cm−2s−1)=5.8

Both ISR and FSR are turned On

Summary of the cross sections obtainedfrom our study:Type Generated pro ess: e

+e−

→ Cross Se t. [fb℄ Generator6 fermions W+W−νν̄ → qq̄qq̄νν̄ 9.21 WhizardZZνν̄ → qq̄qq̄νν̄ 4.05 Whizard

qq̄qq̄νν̄ (ba kgrounds) 6.55 WhizardWZeν → qq̄qq̄eν 38.50 Whizard

e+e−W+W−

→ e+e−qq̄qq̄ 234.80 Pythia*4 fermions W+W−

→ qq̄qq̄ 1948.10 Pythia*ZZ → qq̄qq̄ 142.90 Pythia*

e+e−tt̄ 1.30 Pythia*2 fermions tt̄ → X 136.90 Pythia *qq̄ 4464.60 Pythia** Pythia does not in lude a e+e− polarised beams option during integration

We used the fast Detector SimulationSIMDET (v.4.01):- Tracking system: CCD vertex detector(1.5cm) + Forward Tracker- Magnetic field: 4 T

Calorimetry:- ECal resolution: ∆E/E = 0.2/

√E

- HCal resolution: ∆E/E = 0.5/√

E

3D cells - good granularity -> Energy Flowconcept.


< ↑ ↓ >


Main problem is to reconstruct Z and W pairs

Objects from the detector simulation areforced into 4 jets using the KT jet algorithm- exclusive mode, E recomb. scheme

If succeed, we then have 3 possiblecombinations

Use a 1C Kinematic Fit to find the best pairoptionQ(~x, ~λ) = (~x − ~x0)V −1(~x − ~x0) + 2~λ~f(~x)

Where: ~f(~x) : constraints~x : jet parameters (Ptot,θ,φ)~λ : Lagrange multipliers~V : error matrixError matrix: resolution functionsσP tot(pq), σθ(pq), σφ(pq, θq)

1c : Mjet1jet2 = Mjet3jet4


< ↑ ↓ >








Jet configurations

η

-5 -4 -3 -2 -1 0 1 2 3 4 5

φ

-4

-3

-2

-1

0

1

2

3

4

Jet Contents Jet oneJet twoJet threeJet fourReco. Jetsphotonspionsklong

Jet Contents

η

-5 -4 -3 -2 -1 0 1 2 3 4 5

φ

-4

-3

-2

-1

0

1

2

3

4

Jet Contents Jet oneJet twoJet threeJet fourReco. Jetsphotonspionsklong

Jet Contents


< ↑ ↓ >








Kinematic Fit results

p_Chi2

0 0.2 0.4 0.6 0.8 1

Eve

nts

0

200

400

600

800

1000

1200

p_Chi2

0 0.2 0.4 0.6 0.8 1

Eve

nts

0

500

1000

1500

2000

2500


< ↑ ↓ >


Recoil Mass distribution

(GeV)recoil

M0 100 200 300 400 500 600 700 800 900

Eve

nts

1

10

210

310

410

510

610

(signal)ννZZ

(signal)ννWW

(background)ννqqqq

6fermions (background)



Transverse momentum distribution

(GeV)TP

0 50 100 150 200 250 300 350 400

Eve

nts

1

10

210

310

410

510

610

710

(signal)ννZZ

(signal)ννWW





more...

General selection cuts- Recoil mass: Mrecoil ≥ 200 GeV- PT ≥ 20 GeV- Etrans ≥ 150 GeV- Scaled ycut parameter4.0 ≤ ln

p

ycut1,2 ∗ s ≤ 7.2

- Total missingmomentum| cos θPmiss

| < 0.99

- Most energetic track | cos θPEmax| < 0.99

- Charged tracks in each jet nTracks ≥ 2

- Probability (χ2) > 0.005- Ask for energy around highest track ≥ 2

GeV 5 deg cone

ZZ selection85 < M1C < 100 GeV

WW selection75 < M1C < 85 GeV


< ↑ ↓ >

Results

Cut flow summaryCut �ow Signals Ba kgrounds

νν̄WW νν̄ZZ qq̄qq̄ WZνe Othernone 9210 4050 6550 38500 6928600

Mrecoil 7128 3825 1935 49 32815

PT 6863 3664 1860 49 3450

ET 6347 3562 1688 42 2439

ycut 6141 3497 1601 42 1890

cos θPemax

6133 3493 1595 42 1886

cos θPmiss

6086 3470 1580 42 1778Charged tra ks 6086 3028 1490 21 1505

Econe 5971 2981 1468 9 1244

P (χ2) 5971 2295 1139 negl. 1047

νν̄ZZ selection

(GeV)ZM

50 60 70 80 90 100 110 120

Eve

nts

0

100

200

300

400

500

600

700

800

900

(signal)ννZZ

(signal)ννWW






< ↑ ↓ >

Results

Cut flow summaryCut �ow Signals Ba kgrounds

νν̄WW νν̄ZZ qq̄qq̄ WZνe Othernone 9210 4050 6550 38500 6928600

Mrecoil 7128 3825 1935 49 32815

PT 6863 3664 1860 49 3450

ET 6347 3562 1688 42 2439

ycut 6141 3497 1601 42 1890

cos θPemax

6133 3493 1595 42 1886

cos θPmiss

6086 3470 1580 42 1778Charged tra ks 6086 3028 1490 21 1505

Econe 5971 2981 1468 9 1244

P (χ2) 5971 2295 1139 negl. 1047

νν̄ZZ selection

(GeV)ZM

50 60 70 80 90 100 110 120

Eve

nts

0

100

200

300

400

500

600

700

800

900

(signal)ννZZ

(signal)ννWW





νν̄WW selection

(GeV)WM

50 60 70 80 90 100 110 120

Eve

nts

0

500

1000

1500

2000

2500

(signal)ννZZ

(signal)ννWW





Armentero-Podolanski plot

α-1 -0.5 0 0.5 1

0

10

20

30

40

50

60

70

80 (signal)ννZZ

(signal)ννWW


2fBackgs (background)




< ↑ ↓ >

Sensitivity study

Now I have Z and W candidates and a working analysis system.I combined the pairs of Z and W and make some distributions:

MWW and |cos θ∗| distribution for α4 = α5 = 0.0

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

200

400

600

800

1000

1200

1400

1600

Signal+Backgrounds

Backgrounds

= 0.00004α= 0.00005α

-1L = 1000 fb

) .*θCos(

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1E

ven

ts0

50

100

150

200

250

300Signal+Backgrounds

Backgrounds

= 0.00004α= 0.00005α

-1L = 1000 fb

What happens when α4 and α5 are different from zero?

Expected rate change in the end results and therefore deviations from SM predictions


< ↑ ↓ >

Sensitivity study

I generated 20 scenarios, varying one of the a.c. at a time (10 for α5 = 0.0 and 10 forα4 = 0.0)


< ↑ ↓ >

Sensitivity study


Looked at the cross section for each and compare it to SM


< ↑ ↓ >

Sensitivity study



Coupling value

-0.25 -0.2 -0.15 -0.1 -0.05 -0 0.05 0.1 0.15 0.2 0.25

SM

σ/σ

-0.5

0

0.5

1

1.5

2

2.5

3

4α

5α

ννZZ

Coupling value

-0.25 -0.2 -0.15 -0.1 -0.05 -0 0.05 0.1 0.15 0.2 0.25

SM

σ/σ-0.5

0

0.5

1

1.5

2

2.5

3

4α

5α

ννWW


< ↑ ↓ >

Sensitivity study



Coupling value

-0.25 -0.2 -0.15 -0.1 -0.05 -0 0.05 0.1 0.15 0.2 0.25

SM

σ/σ

-0.5

0

0.5

1

1.5

2

2.5

3

4α

5α

ννZZ

Coupling value

-0.25 -0.2 -0.15 -0.1 -0.05 -0 0.05 0.1 0.15 0.2 0.25

SM

σ/σ-0.5

0

0.5

1

1.5

2

2.5

3

4α

5α

ννWW

We expect better sensitivity to α5 than α4 in particular when looking at the νν̄WW channel


< ↑ ↓ >

Sensitivity study

Generated 12 extra scenarios with both α5 and α4 diff. from zero (each scenario had 400,000events, enough statistics)


< ↑ ↓ >

Sensitivity study


Rerun the whole analysis


< ↑ ↓ >

Sensitivity study



Examples of the new MWW distributions:


< ↑ ↓ >

Sensitivity study




m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.12674α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.09504α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.06334α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.03174α= 0.00005α

-1L = 1000 fb


< ↑ ↓ >

Sensitivity study




m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.12674α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.09504α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.06334α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.03174α= 0.00005α

-1L = 1000 fb

In order to study the sensitivity I used a Binned Maximum Log Likelihood applied to bothMWW and |cos θ∗| distributions for each scenario


< ↑ ↓ >

Sensitivity study




m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.12674α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.09504α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.06334α= 0.00005α

-1L = 1000 fb

m_WW [GeV]

0 100 200 300 400 500 600 700 800 900

Eve

nts

0

500

1000

1500

2000

2500

Signal+Backgrounds

Backgrounds

= -0.03174α= 0.00005α

-1L = 1000 fb

In order to study the sensitivity I used a Binned Maximum Log Likelihood applied to bothMWW and |cos θ∗| distributions for each scenario

BLLH takes as input: SM predicted signal, "observed" events, expected BKG and uncertaintiesin selection efficiencies, backgrounds and luminosity (not really needed ; = 0.001)


< ↑ ↓ >

Sensitivity study

I end up with 32 LLH values for every scenario and each distribution


< ↑ ↓ >

Sensitivity study


These are the results for those 20 scenarios mentioned before:


< ↑ ↓ >

Sensitivity study



4α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450

WWM combinedνν and WWννZZ

5α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450


4α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450

*}θCos{ combinedνν and WWννZZ

5α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450



< ↑ ↓ >

Sensitivity study



4α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450


5α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450


4α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450


5α

-0.15 -0.1 -0.05 0 0.05 0.1 0.15

-lo

g L

H

0

50

100

150

200

250

300

350

400

450


Observation: no symmetric behaviour of those LLH points with respect to the minimum (SMscenario)


< ↑ ↓ >

Sensitivity study

The combined results for those two distributions are

4α-0.15 -0.1 -0.05 0 0.05 0.1 0.15

5α

-0.15

-0.1

-0.05

0

0.05

0.1

0.15All distributions

combinedνν and WWννZZ

295.0 262.3 247.5 244.9 246.9244.9250.6 272.3 332.7 501.7

600.9

326.4

256.1

244.7

245.4

246.3

258.2

335.6

650.3

874.8256.5

368.6 246.4

258.2

263.7251.5

244.6

245.3

243.8 244.8

246.3

239.3


< ↑ ↓ >

Sensitivity study

Under construction!

The fit of those LLH values has been more complicated than expected

At the moment, I’m using S-plines to get a nice fit to those points andextract the limits that I want (and desperately need :) )

What are the estimated values?


< ↑ ↓ >

Conclusions

The hadronic decays of processes

e+e− → νν̄W+W−

e+e− → νν̄ZZ

can be exploited to find new physics provided LHC does not find a Higgs boson

If our ignorance about new physics is parametrised in term of the α4 and α5 anomalouscouplings, they could be measure at the ILC

I achieved to write an analysis framework to study hadronic and leptonic decays⊲ It is OO and it would be interesting to see what can do with semi-leptonic decays

I applied very useful analysis techniques (jet reconstruction, kinematic fit, binned loglikelihoods)

There is progress done on extracting the sensitivity to those anomalous couplings andtherefore provide 68/90 percent Confidence Levels


< ↑ ↓ >

Conclusions

If you can look into the seeds of time,And say which grain will grow and which will not

Speak to me.Macbeth, William Shakespeare


< ↑ ↓ >

Extra slide 1

Signal preparation using the following cuts in phase space:

147.0GeV ≤ m1qq̄ + m2

qq̄ ≤ 171.0 GeV : W

171.0 GeV ≤ m1qq̄ + m2

qq̄ ≤ 195.0 GeV : Z

|m1qq̄ + m2

qq̄ | ≤ 20.0 GeV

mνν̄ ≥ 100.0GeVMultipli ity: 3Resonan es: 3Log-enhan ed: 2t- hannel: 2 W+�eZ�eW�e+ (64) u (4)�d (32)�e (1)��e (2)� (8)s (16)e� (128)31

m_nunu

0 100 200 300 400 500 600 700

Eve

nts

0

5000

10000

15000

20000

25000

30000

35000

m_QuarkPairs_Sel

60 70 80 90 100 110 120

Eve

nts

60

70

80

90

100

110

120


< ↑ ↓ >

Extra slide 2

More distributions:

(GeV)TE

0 100 200 300 400 500 600 700 800 900

Eve

nts

1

10

210

310

410

510

610

(signal)ννZZ

(signal)ννWW





(GeV)cone

E0 20 40 60 80 100 120 140 160 180 200

Eve

nts

1

10

210

310

410

510

(signal)ννZZ

(signal)ννWW





)|Pmiss

θ|cos(0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Eve

nts

1

10

210

310

410

510

610

(signal)ννZZ

(signal)ννWW





)|Pemax

θ|cos(0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Eve

nts

1

10

210

310

410

510

610

(signal)ννZZ

(signal)ννWW






< ↑ ↓ >

Chierici et al results

Previous results:

Confidence Levels contours

-6

-4

-2

0

2

4

6

-6 -4 -2 0 2 4 6α4

α5

ZZνν and WWνν combined - Ref[Chierici et.al.]

Sensitivities (1 dimensional limits)1 DIM analysis 68% C.L. (TESLA 1000 fb−1 800 GeV and polarisation)α4 σ

−= −1.1 σ+ = 0.8

α5 σ−

= −0.4 σ+ = 0.3

My preliminary results⊲ using a polynomial fit

Sensitivities (2 dimensional limits)2 DIM analysis 68% C.L. (TESLA 1000 fb−1 800 GeV and polarisation)

α4 σ−

= −1.18 σ+ = 1.24

α5 σ−

= −0.95 σ+ = 0.96


ww scattering studies at a future linear collider · plans for a future linear collider ww...

Documents