Status of ZEUS

1
Status of ZEUS

• Rainer Mankel
• DESY Hamburg
• --for the ZEUS collaboration--
• PRC Public Session 26-May-2005

2
Outline

• Data-taking detector status
• Data processing simulation
• Physics highlights
• heavy flavours
• QCD fits ?s determination
• diffractive final states

3
HERA-II Luminosity

• HERA-II finally at cruising speed
• almost reached integrated luminosity of HERA-I
• Instantaneous luminosity in ep running surpassed
  ep

4
ZEUS Data-Taking
nb-1

• Series of daily gated luminosity records (?1.2
  pb-1)
• Remarkable improvement in ZEUS data-taking
  efficiency
• in general above 80
• MVD on 95 of time
• 62 pb-1 of new ep data on tape

5
Detector Status

• Calorimeter stable with low number of bad
  channels
• MVD signal noise are stable
• CTD low trip rate. No signs of premature ageing
• Increased use of Global Track Trigger at global
  SLT level

• Efficiency improvements
• comprehensive monitoring via web statistics pages
• pre-luminosity test trigger runs
• complete rebuild of TLT farm hardware software

• Detector is in good shape

6
Straw Tube Tracker (STT)
add insulation cooling
FDET/STT

• STT presently off because of insufficient
  cooling, which otherwise interferes with stable
  operation of solenoid
• Preparing repair that does not break beam vacuum
• additional cooling insulation
• Decision within next weeks

GO magnet
7
Data Processing

• Central reconstruction analysis facility for
  ZEUS
• Presently farm of 150 Intel Xeon processors
• gt600 000 analysis batch jobs in one year
• Up to 22000 jobs per week
• Average data throughput 10 TB/day

8
ZEUS Reconstruction Performance

• Daily production (events) during reprocessing of
  2004 e data

• Peak production 3 M events/day
• at HERA-I times, peak rate was 500 k events/day
• reconstruction of HERA-II events is 6x more
  CPU-intensive than with HERA-I
• Sizable increase in processing performance
• Fast turnaround

Events reconstructed per day
3 M events/day
3 M
2 M
1 M
17-Oct
9-Nov
Date
9
HERA-II MC Production

• Assessment of Monte Carlo simulation needs in
  2003 revealed opening of a huge gap
• Reasons
• HERA-II luminosity
• additional demand from new detector components
• considerable increase in simulation demand during
  HERA-II
• Solution Computing Grid Technology

10
Solution ZEUS Grid

• Built on top of LHC Computing Grid software
• Transparent integration of traditional funnel
  sites and grid sites
• Reuse of existing ZEUS-specific middleware
• ZEUS Grid Toolkit
• Brings the grid to the end user

DESY
CE
RLS
RB
UI
Padova
Classical funnel sites
Tel Aviv
Zeus Integrated MC Production System
KEK
Rutherford
11
ZEUS Grid-Based MC Production Totals

• by time of last PRC 600,000 events

• first non-LCG2 site Univ. Wisconsin (joined last
  month)

• more than 200 million events produced
• 29 grid sites in 7 countries on 2 continents
• gt60 of MC produced on grid

12
Grid ZEUS Other Experiments

• In January, ZEUS production amounted to 16 of
  total CPU power in all UK grid sites
• running at very low priority (opportunistic
  mode)

13
Physics Highlights

• Many essential results still coming from HERA-I
  data
• 8 papers published since last PRC
• HERA-II data are under analysis
• still less statistics than HERA-I, but increasing
• also first results from fresh e data

14
First Impact of New e-p Data D
D ? D0? ? (K-?)?

• About equal sample sizes of e and e- data in
  HERA-II
• Already 2x as much integrated e-p luminosity
  compared to HERA-I
• Time to revisit an old puzzle

ep
40 pb-1
e-p
33 pb-1
m(D0?) - m(D0)
15
D in New Data (contd)

• There is no obvious mechanism that should lead to
  different D cross sections for ep and e-p
  scattering
• Still, an indication of such an asymmetry was
  seen in the HERA-I data
• r(e-p)/ r(ep)1.67?0.21 (Q2gt40 GeV2)
• statistical fluctuation, or mysterious anomaly?

Charm production by boson-gluon fusion
16
D in New Data (contd)

• HERA-II data with increased e-p statistics do not
  show this anomaly any more

17
Total bb Cross Section
HERA-I Data

• Traditional analysis tags single B hadrons (e.g.
  with ?)
• example ptrel study
• sensitive to fast B hadrons
• strong cuts into phase space to suppress
  background
• Tagging of both Bs
• makes use of the correlation
• allows low ptµ thresholds
• is sensitive to B hadrons that are practically at
  rest
• measures total bb cross section

double tag
18
Di-Muons Data vs MC

• Same-charge combinations used to normalize
  light-flavor background

• Good overall description with MC
• bb contribution 2000 events, purity 43

19
bb Cross Section from Di-Muon Events (contd)

• NLO QCD predictionsPHP 5.8 nb
• (FMNR,CTEQ5M)DIS 1.0 nb
• (HVQDIS,CTEQ5F4)
• ? 6.8 nb
• NLO prediction lower than the data
• applied first time at HERA

this analysis
3.0 1.7
Dµ PHP
Dµ DIS
20
?s Determination From NLO QCD Fits
Gluon distribution relative uncertainty in bins
of Q2

• Recently, parton distribution functions have been
  obtained by a rigorous fit to combined inclusive
  DIS jet data
• uncertainty of the mid-to-high x gluon
  distribution in inclusive DIS is very effectively
  constrained by the jet cross sections
• sufficient sensitivity to determine PDFs entirely
  in one experiment

10
21
?s Determination From NLO QCD Fits (contd)

• This method can also be used to determine ?s(MZ)
  as a free parameter in the fit

?s(MZ) 0.1183? 0.0028 (exp.) ? 0.0008 (model)
? 0.0050 (th.)

• in agreement with world average ?s(MZ)0.1182 ?
  0.0027
• first precise determination of ?s from HERA data
  alone

world average (Bethke 2004)
22
Strong Coupling Constant
C. Glasman, DIS2005

• Comprehensive average of ?s(MZ) measurements at
  HERA
• conservative assumptions on correlations of
  theoretical uncertainties (higher order, PDFs,
  hadronisation)
• ?s(MZ)0.1186?0.0011 (exp)?0.0050 (th)
• experimental uncertainty 0.9, theoretical 4
• Need for NNLO calculations

23
NC Cross Section at High x

• High-x region is hard to measure
• for current jet at very low angle, ?had is either
  badly resolved, or the jet is not detected at all
• x range ends at 0.65
• limits determination of PDF at large x

24
NC Cross Section at High x (contd)
e
One jet

• New approach beyond classical double angle
  method
• use jet finder determine x from Ejet and ?jet
• x ? Ejet ? good resolution
• if no jet distinguishable, classify event into
  integrated bin (xmaxltxlt1)

Zero jet
e
25
NC at High x Data vs. Theory

• Good agreement with PDFs
• This method applied first time at HERA

26
Rapidity Gaps at High Q2

• Interest in diffraction at high Q2 extends far
  beyond HERA
• clean final states at LHC, e.g. for Higgs?
• Test factorization of diffractive component of
  proton
• At last PRC, first results for charged currents
  were shown
• statistics did not allow to measure Q2 dependence

Large rapidity gap

• Look for rapidity gaps in neutral current events
• Compare with charged current analysis

27
Rapidity Gaps in NC Events
Rapidity gap
Forward Plug Calorimeter (FPC)
with FPC veto (at beam pipe)

• Normal DIS MC (Ariadne) clearly insufficient at
  low ?max
• Need AriadneRAPGAP (diffractive MC) to describe
  the data

5
28
Rapidity Gaps in NC Q2 Dependence
High -Q2 NC x Plt0.05

• Sizable diffractive contribution to NC cross
  section
• drops with rising Q2
• still 2 at Q21500 GeV2
• NC and CC compatible

29
Exclusive Electroproduction of Vector Mesons
(new ?)

• Very clean final state, large kinematical range
• At HERA, observe transition from soft to hard
  regime
• Present set of measurements ?, ? (new), J/?

• Differential cross sections show high sensitivity
  to the gluon density functions
• challenge to understand diffractive scattering
  quantitatively

30
? Electroproduction Signature

• Reconstructed electron 2 charged tracks
• Otherwise detector completely empty
• Scattered proton escapes near beam line
• 3642 ? signal events

31
VM Electroproduction W Dependence
J/?

• Highly sensitive not only to different models,
  but also to different gluon densities within the
  same model

32
? Electroproduction Q2 Dependence

• Clean separation of ?L and ?T components
  challenges models
• good qualitative description, in particular ?T
• better overall agreement with FS04 model
• Excellent sector for further optimizations of
  pQCD calculations

33
Summary

• Encouraging data rates from HERA-II
• Detector data chain in good shape
• MC mass production on computing grid
• Wealth of new physics results

34

• The End

35

• Backup Slides

36
More Pieces to Pentaquark Puzzle
?(1520) both in forward backward hemisphere
? only in forward hemisphere

• ? signal mainly from forward pseudo rapidity
  region
• unlike regular baryons ?(1520) and ?c
• predominantly at medium Q2
• similar to ?c
• no sign of decuplet partners seen in ?? and
  ?? (? NA49)

37
Di-Muon Mass Spectra (Data vs MC)

• Good description with MC
• bb contribution 2000 events

38
bb Cross Section a Di-Muon Event
39
bb from Di-Muons Normalization of BG-MC

• cc normalize to Dmu analysis
• Bethe-Heitler, elastic charmonium normalize to
  data under isolation cut
• Light flavor use like sign spectrum (minus bb MC)

40
NC Cross Section at High x (contd)

• Considerable extension of kinematical region
  towards high x, in particular for moderate Q2
  (zero jet case)
• Improved x resolution (one jet case)
• Only 99-00 e data used so far (65 pb-1)

41
? Electroproduction W Spectrum

• ?W?
• Test scaling of ? with (Q2MV2)

• J/? is always hard
• ? hardness increases with Q2
• ? shows intermediate behavior

42
J/? Electroproduction Theory vs Exp
shown by T. Teubner at DIS2005