Dijets in DIS and PHP at ZEUS - wiki.bnl.gov fileDIS2011,NewportNews,USA,April11-15,2011 OlegKuprash–1/22 Dijets in DIS and PHP at ZEUS Oleg Kuprash DESY / Taras Shevchenko National

DIS 2011, Newport News, USA, April 11-15, 2011 Oleg Kuprash – 1 / 22

Dijets in DIS and PHP at ZEUS

Oleg KuprashDESY / Taras Shevchenko National University of Kyiv

DIS 2011, Newport News

DIS: Eur.Phys.J.C70:965-982,2010PHP: ZEUS-prel-10-014

On behalf of the ZEUS collaboration

HERA collider


pP X

e′e

q

l l′

Electrons: 27.5 GeVProtons: 920 GeV√s = 318 GeV

Kinematics:Q2 = −q2 = −(l − l′)2

y = PqP l

xBj =Q2

2Pq

Q2 = xBjys

Q2 ≤ 1 GeV2: photoproduction (PHP)

Q2 ≥ 1 GeV2: DIS


Dijets in DIS

Dijet processes in DIS


αSq(p1)

g(p2)

e(l) e(l′)

p(P )

X

γ/Z0(q)jet 1

jet 2

q(p1)

q(p2)

e(l) e(l′)

p(P )

X

jet 1

jet 2αS

γ/Z0(q)

• Processes of leading order in αS :• QCD Compton scattering (quark-induced) • Boson-gluon fusion (gluon-induced)

• Jets in Neutral Current DIS:- two hard scales: virtuality of the exchanged boson and transverse energy of the jets- General tests of pQCD (hard matrix element, factorisation, perturbative expansion, PDFuniversality)- High-precision measurements of strong coupling αS

- Sensitive to parton distribution functions in proton σ =∑

n αnS

∑

a=q,q,g fa/p ⊗ σ(n)a

Dijet processes in DIS: observables


)2 (GeV2Q125 6750 13375 20000

Eve

nts

1

10

210

310

410

(GeV)jetT,BE

10 20 30 40

Eve

nts

210

310

410 )−1ZEUS (374 pb

LEPTO

ARIADNE

(GeV)jjM20 40 60 80

Eve

nts

210

310

410

Bjx0 0.05 0.1

Eve

nts

210

310

410

)ξ(10

log−2 −1.5 −1 −0.5 0

Eve

nts

1

10

210

310

410

*η0 0.5 1 1.5 2

Eve

nts

10

210

310

ZEUS• Q2 - virtuality of the exchanged boson• xBj - for the partonmodel process is the fraction of the protonmomentum carried by the struck parton

• ¯E

jetT,B = 1

2

(

E1T + E2

T

)

- mean transverse energy• Mjj =

√

(p1 + p2)2 - invariant dijet mass(in LO Mjj is identical to the centre-of-massenergy of the parton-boson system)

• ξ = xBj

(

1 +M2

jj

Q2

)

- in LO is fraction of the protonmomentum carried by the initial parton•η′ = 1

2 |η1 − η2| - difference in pseudorapidityis invariant under longitudinal boosts

Dijets in DIS: samples and selections


)2 (GeV2Q125 6750 13375 20000

Eve

nts

1

10

210

310

410

(GeV)jetT,BE

10 20 30 40

Eve

nts

210

310

410 )−1ZEUS (374 pb

LEPTO

ARIADNE

(GeV)jjM20 40 60 80

Eve

nts

210

310

410

Bjx0 0.05 0.1

Eve

nts

210

310

410

)ξ(10

log−2 −1.5 −1 −0.5 0

Eve

nts

1

10

210

310

410

*η0 0.5 1 1.5 2

Eve

nts

10

210

310

ZEUS• 1998-200 and 2004-2007: 203 pb−1

of electron data + 171 pb−1 of positron data• Event phase space:125 < Q2 < 20 000 GeV2

0.2 < y < 0.6• Jets phase space:−1 < η

jetLAB < 2.5

EjetT,B(1,2) > 8 GeV, Mjj > 20 GeV

• Jet selections:- kT clusteralgorithm in the longitudinally invariantinclusive mode running on calorimeter cells- Jetsearch in the η − φ plane of the Breit frame

Dijets in DIS: data corrections and uncertainties


• Correction to the hadron level• QED corrections

)2 (GeV2Q1000 10000

rela

tive

unce

rtai

nty

−0.1

0

0.1

)2 (GeV2Q1000 10000

rela

tive

unce

rtai

nty

−0.1

0

0.1

21

)2syst

δ+2ES

δtotal systematic uncertainty: (

ESδcorrelated systematic uncertainty:

statδstatistical uncertainty:

ZEUS

Uncertainties:• on theabsolute energy scaleof the jets: ±1% forE

jetT,LAB > 10 GeV and

±3% for lower EjetT,LAB

•on the absolute energyscale of the electroncandidate: ±2%• very high precision of measurements

Dijets in DIS: NLO calculations


rela

tive

diffe

renc

e

−0.1

0

0.1

rela

tive

diffe

renc

e

−0.1

0

0.1

−0.1

0

0.1

−0.1

0

0.1

−2 −1.5 −1 −0.5 0

−0.1

0

0.1

−0.1

0

0.1

)ξ(10

log−2 −1.5 −1 −0.5 0

)ξ(10

log

2 < 250 GeV2125 < Q 2 < 500 GeV2250 < Q

2 < 1000 GeV2500 < Q 2 < 2000 GeV21000 < Q

2 < 5000 GeV22000 < Q 2 < 20000 GeV25000 < Q

scaleR

µPDF uncertainty (CTEQ6.6)MSTW 2008ZEUS−JETSZEUS−S

• NLO (O(α2S)) calculations were

obtained using NLOJET++ / DISENT

• CTEQ6.6parameterisations of the proton PDFs

• predictions corrected to the hadronlevel (corrections differ from unity by 5%)

• Sources of uncertainty

in the theoretical predictions:

- due to terms beyond NLO:±6% at low Q2 and ±3% at high Q2

- dueto uncertainty on αS : mostly below ±3%- uncertainty of the modellingof the parton shower: less than 2%- due to proton PDFs:±4% at low Q2 and ±2% ath high Q2

Dijets in DIS: results - Single-differential cross sections


)2 (

pb/G

eV2

/dQ

σd

−410

−310

−210

−110

)−1ZEUS (374 pb

0Z C⊗hadr C⊗NLO

2jetT,BE+2=Q

R2µ

2=Q2R

µ2jet

T,BE=2R

µ

)2 (GeV2Q1000 10000

rel.

diff.

to N

LO

−0.2

0

0.2 jet energy scale uncertainty

NLO uncertainty

(pb

)B

j/d

xσd

210

310

Bjx0.02 0.04 0.06 0.08 0.1re

l. di

ff. to

NLO

−0.2

0

0.2

(pb

/GeV

)je

tT

,BE

/dσd

−110

1

10

)−1ZEUS (374 pb

0Z C⊗hadr C⊗NLO

2jetT,BE+2=Q

R2µ

2=Q2R

µ2jet

T,BE=2R

µ

(GeV)jetT,BE

10 20 30 40 50 60rel.

diff.

to N

LO

−0.2

0


NLO uncertainty

• Kinematic region: - 125 < Q2 < 20 000GeV2, 0.2 < y < 0.6−1 < ηjetLAB < 2.5, Ejet

T,B(1,2) > 8 GeV, Mjj > 20 GeV

• Good description of data by NLO QCD in the whole measured range

Dijets in DIS: results - Single-differential cross sections


(pb

/GeV

)jj

/dM

σd

−110

1

(GeV)jjM20 40 60 80 100 120re

l. di

ff. to

NLO

−0.2

0

0.2

* (p

b)η

/dσd10

210

)−1ZEUS (374 pb

0Z C⊗hadr C⊗NLO

2jetT,BE+2=Q

R2µ

2=Q2R

µ2jet

T,BE=2R

µ

*η0 0.5 1 1.5 2re

l. di

ff. to

NLO

−0.2

0


NLO uncertainty

) (p

b)ξ(

10/d

log

σd

0.05

0.1

0.15

310×

)ξ(10

log−2 −1.5 −1 −0.5 0re

l. di

ff. to

NLO

−0.2

0

0.2

η∗ = 12|η1 − η2|, ξ = xBj

(

1 +M2

jj

Q2

)

• Good description of data by NLO QCD in the whole measured range

Dijets in DIS: gluon induced events


gluo

n fr

actio

n (C

TE

Q6.

6)

0

0.2

0.4

0.6

0.8

gluo

n fr

actio

n (C

TE

Q6.

6)

0

0.2

0.4

0.6

0.8

0

0.2

0.4

0.6

0.8

0

0.2

0.4

0.6

0.8

−2 −1.5 −1 −0.5 00

0.2

0.4

0.6

0.8

0

0.2

0.4

0.6

0.8

)ξ(10

log−2 −1.5 −1 −0.5 0

)ξ(10

log

2 < 250 GeV2125 < Q 2 < 500 GeV2250 < Q

2 < 1000 GeV2500 < Q 2 < 2000 GeV21000 < Q

2 < 5000 GeV22000 < Q 2 < 20000 GeV25000 < Q

gluo

n fr

actio

n (C

TE

Q6.

6)

0

0.2

0.4

0.6

0.8

gluo

n fr

actio

n (C

TE

Q6.

6)

0

0.2

0.4

0.6

0.8

0

0.2

0.4

0.6

0.8

0

0.2

0.4

0.6

0.8

20 40 600

0.2

0.4

0.6

0.8

0

0.2

0.4

0.6

0.8

(GeV)jetT,BE

20 40 60

(GeV)jetT,BE

2 < 250 GeV2125 < Q 2 < 500 GeV2250 < Q

2 < 1000 GeV2500 < Q 2 < 2000 GeV21000 < Q

2 < 5000 GeV22000 < Q 2 < 20000 GeV25000 < Q

• gluon fraction ranges from about 75% at 125 < Q2 < 250 GeV2 and small ξ to about5% at the highest Q2 above 5000 GeV2

• lower Q2-region is not statistically limited → precise input for the PDF fits can beexpected


Dijets in PHP

Dijet processes in PHP


αS

q(p1)

q(p2)

e(l) e(l′)

p(P )

Xp

jet 1

jet 2

αS

Xγ

• Processesof leading order in αS are the same as in DIS• Processes of higher orders in αS :- direct (point-like exchanged boson, same as in DIS)- resolved(represents ’hadronic structure of the photon’)• Jets in PHP:- one hard scale: jet transverse energy- Generaltests of pQCD (hard matrix element, factorisation,perturbative expansion, PDF universality)- High-precision measurements of strong coupling αS

- dijet cross sections sensitive to both proton and photon parton distribution functionsCross section for resolved photoproduction in LO:

dσγpres =

∑

i,j

∫

xγ

∫

xpdxγdxpfi/γfj/pdσij

xγ =Ejet1

Te−ηjet1

+Ejet2

Te−ηjet2

2yEexp =

Ejet1

Teη

jet1+Ejet2

Teη

jet2

Ep

Dijet processes in PHP: observables


αS

q(p1)

q(p2)

e(l) e(l′)

p(P )

Xp

jet 1

jet 2

αS

Xγ

Direct Resolvedγq → qg qg → qg

γq → qq qq → qq

qq → gg

gg → qq (on the figure)gg → gg

qq → qq

qq → q′q′

qq′ → qq′

EjetT - average jet transverse energy

ηjet - average jet pseudorapidityxobsγ - fraction of the incoming photon

momentum taken by the dijet systemMjj - invariant dijet mass| cos θ∗| - scatteringangle in the dijet centre-of-mass system

Dijets in PHP: samples and selections


• data collected during 2005-2006: 189 pb−1 of electron data

• Event phase space: - Q2 < 1GeV2

- γp centre-of-mass energy 142 < Wγp < 293 GeV, or equally 0.2 < y < 0.85

• Jets phase space:

−1 < ηjetLAB < 2.5 E

jetT,LAB 1,2 > 17 GeV

Ejet

T,LAB 1(2) > 21 (17) GeV or Mjj > 60 GeV

| cos θ∗| < 0.8• Jet selections:

- kT cluster algorythm in the longitudinally invariant inclusive mode running on calorimeter cells

- Jet search was performed in the η − φ plane of the laboratory frame

Data corrections:

• Correction to the hadronic level

• No need in polarisation corrections

Uncertainties:

• the absolute energy scale of the calorimetric jets in simulated events was varied by its

uncertainty of ±1%• on the CAL energy uncertainty on Wγp was estimated by varying yJB by ±1% in simulated

events

• total systematic uncertainty in the cross sections was typically below ±5%

Dijets in PHP: NLO calculations


ZEUS

-0.20

0.2 terms beyond NLO

theo

reti

cal r

elat

ive

unce

rtai

nty

-0.20

0.2 αs(MZ)

-0.20

0.2 hadronisation

-0.20

0.2 pPDFs

-0.20

0.2

20 30 40 50 60 70 80

γPDFs

E––

jet

T (GeV)

• NLO (O(α2S))

calculations were obtained using theprogram by Klasen, Kleinwort and Kramer• µR = µF = (Ejet

T )max

• ZEUS-Sparameterisations of the proton PDFs• GRV-HOparameterisations of the photon PDFs• predictionscorrected to the hadron level (correctionsdiffer from unity by less then 5%)• Sources of uncertainty

in the theoretical predictions:

- due to terms beyond NLO: up to 20%- due to uncertainty on αS : below ±5%- uncertainty of the modellingof the parton shower: less than 2%- due to proton PDFs: less than 5%- due to photon PDFs estimated by using an alternative set of parameterisations, AFG04:less than 5%

Dijets in PHP: results - Single-differential cross sections


ZEUS102

10

1

10-1

10-2

10-3

ZEUS (prel.) 189 pb-1

NLO hadr: p/γ PDFs⊗(Klasen et al.)

ZEUS-S/GRV-HO

MSTW08/GRV-HOZEUS-S/AFG04

ZEUS-S/CJK

Ejet1

T > 21 GeV, Ejet2

T > 17 GeV

-1 < η jet < 2.5

Q2 < 1 GeV2

0.2 < y < 0.85

jet energy scale uncertainty

dσ/

dE–– je

t

T

(pb/

GeV

)

-0.5

0

0.5

20 30 40 50 60 70 80

E––

jet

T (GeV)

rel.

diff

. to

NL

OZEUS

0

250

500

750

1000



ZEUS-S/GRV-HO

MSTW08/GRV-HOZEUS-S/AFG04

ZEUS-S/CJK

Ejet1

T > 21 GeV, Ejet2

T > 17 GeV

-1 < η jet < 2.5

Q2 < 1 GeV2

0.2 < y < 0.85

jet energy scaleuncertainty

dσ/

dxγ

ob

s (p

b)

-0.5

0

0.5

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

xγ obs

rel.

diff

. to

NL

O• Kinematic region: Q2 < 1GeV2, 0.2 < y < 0.85−1 < η

jetLAB < 2.5, Ejet

T,LAB 1(2) > 21 (17) GeV

• Good description in shape and normalisation

xγ =Ejet1

Te−ηjet1

+Ejet2

Te−ηjet2

2yEecross section is expected to be most sensitive to the photon

PDFs, especially at low xγ , where resolved processes are dominant

Dijets in PHP: results - Single-differential cross sections


ZEUS10

1

10-1

10-2

10-3



ZEUS-S/GRV-HO

Ejet

T > 17 GeV -1 < η jet < 2.5 Q2 < 1 GeV2

0.2 < y < 0.85 |cosθ*| < 0.8

jet energy scale uncertainty

dσ/

dMjj (

pb/G

eV)

-0.5

0

0.5

60 80 100 120 140 160

Mjj (GeV)

rel.

diff

. to

NL

OZEUS

0

100

200

300ZEUS (prel.) 189 pb-1


ZEUS-S/GRV-HO

Ejet

T > 17 GeV -1 < η jet < 2.5

Q2 < 1 GeV2

0.2 < y < 0.85 Mjj > 60 GeV

jet energy scaleuncertainty

dσ/

d|co

sθ* | (

pb)

-0.5

0

0.5

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8

|cosθ*|

rel.

diff

. to

NL

O• Kinematic region: Q2 < 1GeV2, 0.2 < y < 0.85−1 < η

jetLAB < 2.5, Ejet

T,LAB 1,2 > 17 GeV, | cos θ∗| < 0.8, Mjj > 60 GeV2

• Good description in shape and normalisation• Demonstates validity of the description of the dynamics of dijet production by pQCD atO(α2

S)

Conclusions


• Dijet measurements in both PHP and DIS were performed at HERA using

ZEUS detector

• The measurements of dijets in DIS have very small statistical and

systematical uncertainties

• ... and the description of the data by the predictions of NLO QCD is very

good

• → DIS dijet data will provide useful precision information for the

determination of the strong coupling constant and the exctraction of the

proton PDFs

• Precise measurements of dijets in PHP was done

• Given cross sections are sensitive to the parton densities in the proton and

photon

• ... and the description of the data by the predictions of NLO QCD is very

good

• → Measurements allow for improving the determination of the photon and

proton PDFs in future QCD fits


Backup slides

Dijets: Infrared Sensitivity


• In some regions of the dijet phase space the cancellation between soft and collinearsingularities in NLO theoretical predictions is incomplite• Real emissions of hard gluons are suppressed in regions in which the azimuthal difference∆φjj of the two jets is close to π, where ∆E

jjT and Mjj is close to threshold

• But if jets slightly decorrelated,soft gluon emission is allowedbut kinematically constrained• ... some of the virtualdivergences are left uncancelledand theory calculations becomesensitive to the soft gluon emission• in order to make theoryinfrared insensitive asymmetric ET

cut can by applied or cut on Mjj

• In DIS dijet analysis cut onMjj > 20 GeV has been applied

• In PHP dijet analyses asymmetric cut on Ejet1(2)T,LAB > 21 (17) GeV has been applied for

cross sections as functions of EjetT , ηjet, xobs

γ and cut on Mjj > 60 GeV for Mjj and| cos θ∗|

Correlation between Mjj and η′


M2jj = 2 · Ejet1

T · Ejet2T · [cosh(η1 − η2)− cos∆φjj ]

Dijets in DIS and PHP at ZEUS - wiki.bnl.gov fileDIS2011,NewportNews,USA,April11-15,2011 OlegKuprash–1/22 Dijets in DIS and PHP at ZEUS Oleg Kuprash DESY / Taras Shevchenko National

Documents