Date post: | 19-Dec-2015 |
Category: |
Documents |
View: | 215 times |
Download: | 0 times |
Physics in CollisionJ. Huston June 1999
Review of Parton Distributions and Implications for the Tevatron
and LHC(Partons in Collision at Physics in
Collision)J. Huston
Michigan State Universitythanks to James Stirling, Lenny Apanasevich and Michael Botje, Heidi Schellman and Ursula Bassler for figures
See also http://www.pa.msu.edu/~huston/lhc/lhc_pdfnote.ps
Physics in CollisionJ. Huston June 1999
Determination of parton distribution functions (pdf’s)
Calculation of production cross sections at the Tevatron and LHC relies upon knowledge of pdf’s in relevant kinematic range
pdf’s are determined by global analyses of data from DIS, DY and jet and direct production
Two major groups that provide semi-regular updates to parton distributions when new data/theory becomes available
CTEQ->CTEQ4(5) MRS->MRST98 GRV(not really global
analysis; concentrate on x) Giele-Keller-Kosower (no
pdf’s yet; error analysis)
Title: Cteq5mQ5.SP -- /users/m2a/1result/Cteq5/CandB/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
“Evolution” (in time) of pdf’s
u valence quark distribution •u sea quark distribution; influence of HERA data clearly seen
HERAdataincluded
Physics in CollisionJ. Huston June 1999
Evolution (in Q2) is the great equalizer
Physics in CollisionJ. Huston June 1999
Gluon Distribution-log x
Physics in CollisionJ. Huston June 1999
Gluon Distribution-linear x
Physics in CollisionJ. Huston June 1999
Comparison of LO and NLO pdf’s
LO fits conducted separate from NLO fits; many processes have large K-factors (NLO/LO); resulting LO pdf’s reflect this
Physics in CollisionJ. Huston June 1999
Comparison of LO and NLO pdf’s
Physics in CollisionJ. Huston June 1999
Comparison of LO pdf’s Many LHC comparisons have
used CTEQ2L (default in PYTHIA), a pdf that is “several generations old”
…again differences are smaller at highQ2
light Higgs region
Physics in CollisionJ. Huston June 1999
Comparison of LO pdf’s
Physics in CollisionJ. Huston June 1999
Comparison of gluons-linear
Physics in CollisionJ. Huston June 1999
Impact of new data
In the last few years, improved and new experimental data have become available in many processes; these data have been incorporated into the new CTEQ and MRST analyses
DIS: NMC and CCFR have published final analyses; H1 and ZEUS have published more extensive and precise data on F2
Lepton-pair production (p/d)asymmetry: E866 has measured ratio of lepton-pair production in pp and pd collisions over the x range of (0.03-0.35)
Lepton charge asymmetry in W production: CDF has improved accuracy and extended the y range
Inclusive large pT jet production: CDF and D0 have recently finished final analyses of Run 1b inclusive jet cross section, including full information on correlated systematic errors; provides crucial constraints on gluon distribution in CTEQ5 analysis
Physics in CollisionJ. Huston June 1999
The latest in pdf’s CTEQ5M: main pdf set;
performed in MSbar scheme CTEQ5D: fit performed in DIS
scheme CTEQ5L: fit performed using
leading order matrix elements CTEQ5HJ: in MSbar scheme but
with increased emphasis on high ET jet points
CTEQ5HQ: uses systematic generalization of MSbar scheme to include heavy-quark partons
CTEQ5F3: uses a fixed 3-flavor scheme where charm and bottom quarks are treated as heavy particles and not partons
MRST1: main pdf set performed in MSbar scheme with nominal s(MZ) and kT smearing values
MRST2: smaller kT corrections
MRST3: larger kT smearing corrections
MRST4: as in MRST1 but with a lower value of s(MZ)
MRST5: as in MRST1 but with a higher value of s(MZ)
MRSTDIS(1-5): DIS versions of MRST(1-5)
MRSTLO(1-5): LO versions of MRST(1-5)
MRSTHT(1-5): HT versions of MRST(1-5)
Physics in CollisionJ. Huston June 1999
Evolution and the uncertainty in s
pdf’s determined at a given x and Q2 “feed down” to lower x values at higher Q2
accuracy of extrapolation depends both on accuracy of original measurement and uncertainty on s
@ large x, DGLAP equation for F2
can be approximated as ∂F2/∂logQ2 ~s(Q2)PqqXF2
Effect on evolution of error on s for F2 shown on right
Extrapolation uncertainty of ±5% in F2 at high Q2 from uncertainty in s
Physics in CollisionJ. Huston June 1999
Higher orders in evolution
There is a relatively large effect going from LO to NLO.
Should be smaller going from NLO to NNLO.
Necessary for LHC?
Title: d99a.epsCreator: HIGZ Version 1.23/09Preview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
Heavy quark pdf’s
In processes where heavy quarks play important role (charm production at HERA), standard schemes using zero-mass heavy quarks partons may be inadequate. Also of interest is b quark pdf’s for Higgs production at LHC.
Thus, CTEQ has produced CTEQ5HQ set using ACOT scheme which gives a more accurate formulation of charm quark physics, valid from Q=mc to Q>>mc.
PDF’s defined in (mass-independent) MS scheme, matched with hard-scattering cross sections using on-mass shell heavy quarks.
In practice, only makes a difference for DIS structure functions.
CTEQ5HQ gives slightly better overall fit than CTEQ5M
Mixing CTEQ5HQ pdf’s and MS cross sections increases 2 by 600
Title: 5M-5HQ.SP -- /users/wkt/1result/cteq5/Round4/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
CTEQ and MRST heavy flavor pdf’s
MRST uses similar (Thorne-Roberts) scheme for treating massive quarks; again important primarily for DIS
Title: ChmC5hqMrs.SP -- /users/m2a/1hep/0fits/Round4/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Title: MRS-5HQ.SP -- /users/w2a/1result/cteq5/Round4/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Differences can be explained by:•slightly different choices of charm mass•differences in procedure for treating charm quark masses in Wilson coefficients
Phenomenology is the same if appropriate ME’sare used.
Physics in CollisionJ. Huston June 1999
Uncertainties on pdf’s
of quark distributions (q + qbar) is well-determined over wide range of x and Q2
Quark distributions primarily determined from DIS and DY data sets which have large statistics and systematic errors in few percent range (±3% for 10-4<x<0.75)
Individual quark flavors, though may have uncertainties larger than that on the sum; important, for example, for W asymmetry
information on dbar and ubar comes at small x from HERA and at medium x from fixed target DY production on H2 and D2 targets
Note dbar≠ubar
strange quark sea determined from dimuon production in DIS (CCFR)
d/u at large x comes from FT DY production on H2 and D2 and lepton asymmetry in W production
Title: DbrUbrC5.SP -- /users/m2a/1result/Cteq5/CandB/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Bodek and Yang have argued that D2 dataneed to be corrected for nuclear binding effects
which would lead to larger d/u ratio at large x
Physics in CollisionJ. Huston June 1999
Nuclear corrections to D2 and the d/u ratio
Bodek and Yang: if nuclear corrections areapplied to D2, then d/u->0.2 (rather than 0) as x->1. Result is d quark distribution increases.
Impact on high x CC at HERA
Impact on jet production at Tevatron
Physics in CollisionJ. Huston June 1999
NMC and W asymmetry
NMC data and CDF W asymmetry can be well-fit without using nuclear corrections for D2 data
Title: CDFwAsym.SP -- /users/m2a/1result/Cteq5/CandB/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Title: NmcRatio.SP -- /Net/cteq06/users/wkt/1hep/0fits/Round5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
No model of nuclear corrections is used inthe CTEQ5 fits (i.e. D2 cross section istreated as incoherent sum of p and n ones.
Physics in CollisionJ. Huston June 1999
d/u uncertainty
M. Bottje study; hep-ph/9905518
Title: ../lqdis/lqdis_ratios.epsCreator: HIGZ Version 1.23/09Preview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Title: ../lqdis/lqdis_dou.epsCreator: HIGZ Version 1.23/09Preview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
With present data, can’t say one way oranother. Higher statistics should provide definitive answer.
Physics in CollisionJ. Huston June 1999
d/u
Previously, driving force for d/u was one data point (from NA51) for both MRS and CTEQ.
E866 covers a much wider kinematicrange.
Title: E866a.SP -- /users/wkt/h2a/0fits/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
Gluon Uncertainty Gluon distribution is least well known
(but one of most important for physics processes at the LHC)
Momentum fraction carried by quarks is very well known from DIS data; at Qo=1.6 GeV
momentum fraction carried by quarks is 58%±2%
thus momentum fraction carried by gluons is 42%±2%
->if gluon increases in one range, it must decrease in another
X bin Momentum fraction (Q=5 Gev)
10-4 to 10-3 0.6%
10-3 to 0.01 3%
0.01 to 0.1 16%
0.1 to 0.2 10%
0.2 to 0.3 6%
0.3 to 0.5 5%
0.5 to 1.0 1%
Momentum shifted to lower x as Q2 is increased
Physics in CollisionJ. Huston June 1999
CTEQ Study CTEQ study; vary gluon
parameters in a global analysis and then look for incompatibilities with data
Use only DIS and DY data sets where theoretical and experimental systematic errors are under good control
Use standard parameterization for gluon distribution
AoXA1(1-x)A2(1+A3xA4)
Vary A1,A2,A3,A4 each time refitting other quark, gluon parameters
Fairly tight constraints on the gluon distribution except at high x
Physics in CollisionJ. Huston June 1999
CTEQ gluon study More important to know uncertainties
on gluon-quark and gluon-gluon luminosity functions at appropriate kinematic region (in =x1x2=s_hat/s
Define:
dL/d = ∫g(x,Q2)q(/x,Q2)dx/x
Define: dL/d = ∫g(x,Q2)g(/x,Q2)dx/x
Uncertainties√ range gluon-gluon gluon-quark<0.1 +/-10% +/-10%0.1-0.2 +/-20% +/-10%0.2-0.3 +/-30% +/-15%0.3-0.4 +/-60% +/-20%
Physics in CollisionJ. Huston June 1999
Giele-Keller-Kosower Study
Goal is “honest error estimates”; as mentioned before, spread of predictions using different pdf sets is not a proper error estimate.
Honest error estimate requires evaluation of errors on pdf’s due to measurement errors and method for propagating these errors to observables.
Their solution: use functional integration. Construct a probability functional Prob(f,o|data) that the parton distribution f along with o(=s(mZ
2) provide a description of the data.
Data selection: because of worries about
nuclear corrections do not use any data on nuclear targets
Because of worries about scale dependence, don’t use prompt photon data.
Use only data sets that have published correlated systematic errors
In fits so far, only H1 ep (ZEUS rejected), BCDMS H2 and CDF W asymmetry data used.
->a lot of information ‘thrown away’ (my phrasing)
“…promises an end to the tyranny of the Global Fitters”
Physics in CollisionJ. Huston June 1999
Z vs W
H1, BCDMS H2 alone Add CDF W asymmetry also
CDF error ellipse
Physics in CollisionJ. Huston June 1999
Direct Photons and kT
NLO QCD inadequate to explain size of observed kT in DY, W/Z, and diphoton distributions; full resummation calculations needed
May be similar effect in direct ; no rigorous resummation calculation available for direct
•Soft gluon radiation causes deviations fromNLO QCD at low ET at Tevatron•<kT> increases as log of s
•1 GeV/c for fixed target•3-4 GeV/c for Tevatron collider•6-7 GeV/c for LHC (low mass states)
don’t expectphoton-jet balancingat low ET
Physics in CollisionJ. Huston June 1999
New Photon Result from CDF (1b)
Physics in CollisionJ. Huston June 1999
Diphoton Measurements at CDF
2 aspects:•QCD measurements of •exotic searches with diphotons,e.g. Higgs->: looser cuts to maximizeefficiency
Require:•ET
1,2 > 12 GeV/c•Isolation energy in cone of 0.4 < 1 GeV/c
saturated by MB energy forN.B. backgrounds come from jets withzo (=Eo/Ejet) > Eo/(Eo+1)
•zmin~0.95 for ET=20 GeV/c
•fragmentation functions not welldetermined here, especially notwith gluons and especially notin Monte Carlos
Note that distributions that are functionsat LO are not well-described at NLO
->need resummed predictions
Physics in CollisionJ. Huston June 1999
Direct Photons and kT
Effects of kT more severe at fixed target energies
Theoretical uncertainties too large to use direct photons for determination of gluon distribution (->CTEQ conclusion (jets ‘determine the gluon’); MRST uses direct photons with kT)
Physics in CollisionJ. Huston June 1999
CTEQ5 and direct photons
So, CTEQ5 has no direct photon data in the fit..but
...both WA70 and E706 are well-fit with CTEQ5 pdf’s
WA70 with no kT
Title: Wa70a.SP -- /users/w2a/1result/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Title: E706.SP -- /users/w2a/1result/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
E706 with the experimentally measured values of kT
Physics in CollisionJ. Huston June 1999
CTEQ5 and MRST gluons
Difference in approach to direct photon cross sections (and the gluon distribution) leads to the most striking differences between CTEQ5 and MRST pdf’s (most striking difference in any contemporary pair of CTEQ/MRS pdf sets).
Title: GluCtqMrsQ5.SP -- /users/w06/h2a/0fits/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
Influence of Jets @LO, jet cross section is proportional
to s2g(x,Q)g(x’,Q) and
s2g(x,Q)q(x’,Q)
• flexibility in gluon allows for increase in
theoretical cross section at high ET
700 GeV/c
1.4TeV/c
2.1 TeV/c
2.8TeV/c
@LHC assuming xT universality
Physics in CollisionJ. Huston June 1999
Differential Dijet Production Differential dijet production
directly probes larger x and Q2 range than inclusive cross section
Physics in CollisionJ. Huston June 1999
Differential Dijet Production
Physics in CollisionJ. Huston June 1999
Dijet Mass Cross Section
Physics in CollisionJ. Huston June 1999
Role of LHC in pdf determination
ATLAS/CMS measurements of DY (including W/Z), direct photon, jet, top production,etc will be useful in determining pdf’s relevant for LHC
can try to extract parton-parton luminosities directly from cross sections (Dittmar et al)
can input data into global fitting analyses
DY production will provide information on quark (and anti-quark) distributions while direct photon, jet and top production will provide, in addition, information on the gluon distribution
For example, direct photon production.
Physics in CollisionJ. Huston June 1999
Jet Production at the LHC Jet production at the LHC has a
similar sensitivity to pdf’s as at the Tevatron
Physics in CollisionJ. Huston June 1999
Diphoton Production at the LHC
cross section at 14 TeV
gg
Physics in CollisionJ. Huston June 1999
W/Z + top cross sections at the LHC
Physics in CollisionJ. Huston June 1999
W/Z production at the LHC
W+/W-/Z rapidity distributions provideinformation on quark and antiquark distributions
Physics in CollisionJ. Huston June 1999
CTEQ5M/5HJ and Tevatron Jets
Title: D0-CTEQ5.SP -- /users/w2a/1result/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Title: CDF-CTEQ5.SP -- /users/w2a/1result/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
CTEQ5/MRST comparison
Title: D0-CTEQ-MRS.SP -- /users/w2a/1result/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Title: CdfCtqMrs.SP -- /users/w2a/1result/cteq5/plotsCreator: SciPlotPreview: This EPS picture was not saved with a preview (TIFF or PICT) included in itComment: This EPS picture will print to a postscript printer but not to other types of printers
Physics in CollisionJ. Huston June 1999
s from inclusive jet production large correlation between s and
gluon distribution makes indepen-
dent measurement of s difficult