Forward Jets and Particles in ep collisions and parton dynamics

• Parton dynamics at high energies

• Central region

• Forward Jets

• Forward 0

• Conclusions

On behalf of H1 and ZEUS Collaborations

Jacek Turnau, Institute of Nuclear Physics, Cracowep (27.5 + 820 G

eV) collisions

QCD Evolution at low x

At low x scattered parton usually descends from long cascade of parton branchings.

g 2T

1 2xp~ ~

E k

gnggg ...321

Tk

pxg

gngg xxx ...21

TnTT kkk ...21

n ...21

DGLAP

BFKLBFKL

CCFMCCFM

Comparison of the data to MC models with different QCD dynamics

No kt ordering in initial state radiationkt ordered initial state radiation

RAPGAP DIR RAPGAP RES at scale Q2

+ pt2 (jets) or Q2 + 4pt

2 (0) CASCADE 1.0

Resolved photon CCFM evolution equation

QET QET QET

DGLAP description gradually deteriorates when going from backward to forward direction

Forward region: • huge NLO correction• large deviations at small Q2 & ET

Moral: NNLO may be important in forward region

Inclusive Jets in DIS

5< Q2 <100 GeV2, 0.2 < y <0.6

incl. kT algorithm in Breit frame

Forward jets : forward region under special scrutiny

Large xjet/xbj to enhance phase space for BFKL evolution

22 QETjet to suppress DGLAP evolution

22 GeV 755 :DIS Q

Forward jet (incl. kt algo.)

0.25.0

035.0

0.200.7

2

2

Q

p

x

tjet

jet

jet

•DGLAP direct : too low

•CCFM: too high

•DGLAP (DIR+RES): OK

similar pattern of agreement/disageement for xjet, pTjet

(1997)Forward Jets

222TpQ

Forward jets are experimentally difficult:

• Interference with proton remnant

•Hadronic corrections strongly model dependent at small x

Forward jets in ZEUS detector

Forward jets forward particles (0)

jet

0

)( 0fragmentationby Lund model or convolution with FF

)( jet

There is another interesting aspect of bringing particle into the game - later

algorithms

DGLAP direct: too low

Best description:

direct + resolved at scale 2 = Q2 +4pT

2

CCFM too low at small x

Mod. LO BFKL tuned to H1 1997 data + recent FF describes the data

Forward 0 cross section : x dependence

Similar pattern of agreement/disagreement for other distributions

Evolution scheme

Renor. & factor.scale

Fragmentation scheme

Fragmen-tation scale

Forward jets

Forward 0

DGLAP (dir+res) RAPGAP

2=Q2+p2t

Up to 2=Q2+4p2

t

JETSET 7.4 (Lund model)

String inv. mass

OK OK at upper limit of renor. scale

CCFM CASCADE

2= Q2+4mQ2 PYTHIA 6.2

(Lund model)String inv. mass

too high too low

Mod. LO BFKL

2=kTjet2 LO KKP FF z_pi* Q2 OK OK

Something wrong either in mod. LO BFKL (KMO) or CASCADE

Overview of description of jet/particle x-sections

Transverse energy flow associated with forward 0

In hadronic CMS:

• 0 close to proton (most forward)

0 towards to photon (less forward)

SUMMARY AND CONCLUSIONS

Forward jets at HERA: after ~10 years description still difficult.

NLO DGLAP not enough to describe forward jet data

DGLAP direct + resolved describes the fwd jet & 0 data

Energy flow pattern slightly favors DGLAP direct + resolved in comaprison with other schemes

Mod. LO BFKL tuned to jet data describes 0 data

CCFM slightly overshoots jet data and underestimates 0 data

There seems to be contradiction between last two points : something must be wrong...

Forward jets and BFKL

220

2 GeV 5.0),4/( kkTs22

02 GeV 5.0),( kkTs

Modified LO BFKL calculation

Kwiecinski, Martin Outhwaite hep-ph/9903439

Normalization very sensitive to infrared cut-off k0 and scale for S

H1 1997 jet data corrected to parton level with CASCADE

gx

Tjetjet kx ,

),( 2Qxg

),( Tjetjeta

a kxf

),(,,)( 22

224 jTjjT

jjT

s kxFQkx

x

kjet

Modified LO BFKL calculationKwiecinski, Martin Outhwaite hep-ph/9903439

Forward Jets