PROMPT PHOTONS IN PHOTOPRODUCTION: extended analysis
Additional Material
Deltaphi - PYTHIA - influence of ISR
The FGH model describes well all variables, both for all x-gamma, and high and low x-gamma.
The LMZ model, which has no explicit resolved component for the quark content of the photon, describes, within its large errors, all variables, with the exception of eta-jet and the associated delta-eta.
Note that the version of the LMZ model used here differs from that used in our previous publication by the inclusion of a gluon-quark subprocess ( see (4) in 1307.3644v2) that represents the gluon content of the resolved photon. There has also been a redefinition of x-gamma. As previously, the quark content of the resolved photon is given by the order alpha**2alpha_s subprocesses ( see (5),(6) updated LMZ model - Talk , 1307.3644 ).
See below, theory section, for more details.
Andrii's reweighting studies (19/02/2014)
reweight delta-phi Conclude no significant effects on other variables. Difference between PYTHIA and data is probably physics ( FGH NLO describes data well, PYTHIA LL + parton shower does not).
________________________________________________________________________________________________________
DVCS MC study: DVCS - E,gamma,P, DVCS - E,gamma,gamma,P. See plot 4 for the influence of the soft gamma on the transverse energy unbalance, dvcs17 plots
Andrii's presentation at ZAF (14/08/2013) with trigger and updated DVCS energy balance studies; new calorimeter systematics
DVCS - summary of analysis Justifies use of 1.1 factor for ZUFO-energy of photons. See sections 12 - 14 for the original php MC analysis.
________________________________________________________________________________________________________Summary of theory
FGH
QCD order 1) gamma q --> gamma q LO alpha_em**2 2) gamma q --> gamma q g NLO alpha_em**2 alpha_s gamma g --> gamma q qbar NLO alpha_em**2 alpha_s 3) gamma g --> gamma g NNLO alpha_em**2 alph_s**2 (box) FGH has virtual terms for (2) and fragmentation terms take account of collinear singularities. There are both direct processes (1),(2),(3) and resolved processes via the photon structure function. Uses collinear approximation. Thus FGH is LO + NLO + NNLO.LMZ
There is no direct process (1) and no resolved photon. The model is based on kt-factorisation of the proton pdf and (2) + (3). There is no fragmentation term. There appear (in contrast to FGH ) to be no virtual terms. Singularities are avoided by a 1 GeV cut-off in mass. Thus LMZ is NLO + NNLO with no LO and no resolved processes via a photon PDF - for the version of LMZ theory used in the first paper. For this second paper LMZ have added a resolved term for the photon that describes the gluon content of the photon; thus the gluon(from photon) plus quark(from proton) ( subprocess (4) in 1307.3644 )is included. Note that the quark content of the resolved photon is excluded. For more details, see 1307.3644 . Note that the final state contains a photon plus three partons. An extra parton has been added to take account of initial state radiation from a parton cascade originating from the proton (See P7, second paragraph, of 1307.3644). The effect of this is to increase the cross section primarily at low x-gamma and to change the jet-based variables from those expected from 2 -> 3 processes. For an illustration of the parton cascade, see here. For LMZ, the absence of the alpha_s independent LO term, and the dependence of the model on only the NLO and NNLO terms, leads to a larger dependence on the renormalisation scale than is found for FGH. _______________________________________________________________________________________________________
PROMPT PHOTONS IN PHOTOPRODUCTION: extended analysis - initial studies In this analysis the kinematic selections are identical those of the previous PHP analysis (A1). The main objective is to study the direct and resolved photonic interactions in detail through study of x-gamma regions.
(3) Correction factors for x-gamma regions
For IOS work: correction factors are from a 50:50 mix of direct and resolved; standard isolation criteria have been used. Black ( cyan or magenta) IOS (AI) analysis. FGH theory recalculated using running alpha_em, cteq6, and theory uncertainty from variation of renormalisation scale by a factor 2. Fit numbers from Section 57 (A1); only FGH theory changed. Theory is hadronisation-corrected using stand-alone PYTHIA (ie not ZEUS version). Fragmentation corrections are not used since they are ill-determined ( see Section 62 A1).
For simplicity, my [0.0;0.7(0.8)] and [0.8;1.0] analyses use acceptance correction ( and hadronisation) factors from resolved and direct MC acceptance corrections, respectively. No account has been taken of the restricted ranges and the 'contamination' of the alternative process. In spite of this there is good agreement with Andrii's results: see 'Comparison of cross sections' below.
Andrii's work below uses the selection criteria defined in our recent paper - Photoproduction of isolated photons, inclusively and with a jet, at HERA.
Update 3/12/13. The [0.;0.8] and[0.8;1.0] ranges in x-gamma now use a 30%-70% and 70%-30% mix of direct and resolved acceptance correction factors, respectively.
Cross sections: photon + jet
Andrii's most recent figures - comparison with theory (29/01/2014)
Comparison of cross sections ( Andrii, 25/10/2013)
xp-phi: comparison with LMZ - Andrii, 25/10/13
xg [0.;0.7]comparison with LMZ - Andrii
xg [0.8;1.0]comparison with LMZ - Andrii
Control plots - Andrii (24/10/13)
Systematic - gamma - Andrii(8/11/13)
Systematic - gamma - 08 Andrii(29/11/13)
Andrii's cross sections 0.-0.7, 0.8-1.0
Andrii's cross sections 0.-0.7,0.0-0.8, 0.8-1.0
IOS results. Cross sections: photon + jet
Fig 1.1, photon cross section
Fig 1.1a, photon, xg [0.0;0.7] Improved hadronisation correction factors. photon cross section
Fig 1.1b, photon, xg [0.8;1.0] . Improved hadronisation correction factors. photon cross section
Fig 1.1d, photon, xg [0.0;0.8] . new correction factors. photon cross section
Fig 1.2, jet, jet cross section
Fig 1.2a, jet, xg [0.0;0.7] . jet cross section
Fig 1.2b, jet, xg [0.8;1.0] . jet cross section
Fig 1.2d, jet, xg [0.0;0.8] . new correction factors. jet cross section
Fig 1.3, x-gamma
Fig 1.3a, x-gamma, log plot
Fig 1.4, xp (updated FGH 7/10/13), cross section
Fig 1.4a, xp , xg [0.0;0.7]. cross section
Fig 1.4b, xp , xg [0.8;1.0] cross section
Fig 1.4d, xp , xg [0.0;0.8]. new correction factors cross section
Fig 1.5 (2 plots) , phi gamma-jet (updated FGH 7/10/13), cross section
Fig 1.5a (2 plots) , phi gamma-jet, xg [0.0;0.7]. cross section
Fig 1.5b (2 plots) , phi gamma-jet, xg [0.8;1.0] cross section
Fig 1.5d (2 plots) , phi gamma-jet, xg [0.0;0.8]. new correction factors. cross section
Fits to xp and phi.
Fig 2.1, xp, all x-gamma.
Fig 2.2, xp, [0.0;0.7] x-gamma.
Fig 2.2a, xp, [0.0;0.8] x-gamma.
Fig 2.3, xp, [0.8;1.0] x-gamma.
Fig 2.4, phi, all x-gamma.
Fig 2.5, phi, [0.0;0.7] x-gamma.
Fig 2.5a, phi, [0.0;0.8] x-gamma.
Fig 2.6, phi, [0.8;1.0] x-gamma.
Fits to Et, eta, x-gamma for photon + jet events
Fig 3.1, x-gamma, Et, eta for gamma, jet
Fig 3.2, Et, eta, [0.0;0.7] x-gamma.
Fig 3.3, Et, eta, [0.8;1.0] x-gamma.
Fig 3.4, Et, eta, [0.0;0.8] x-gamma. gamma cross section For cross-check that 3.4 + 1.1br = 1.1.
Check that sum of gamma cross sections from two xg fit regions = cross section from overall fit. See Figs 1.1, 1.1br, 3.4 . xg range Et eta 0,0.8 9.83 9.73 0.8,1 12.64 12.18 sum 22.47 21.91 all xg 22.41 21.67 Note approximation used here. Correction factors are pure direct ( pure resolved) for x-gamma gt 0.8 (x-gamma lt 0.8).
The FGH code has been run with/without photon isolation.
Fig 2.1.1 dd,rd,df,rf differential cross sections for Et-gamma, eta-gamma.
The Figures and cross sections below show that, as expected, isolation significantly changes the direct and resolved fragmentation cross sections. There is a decrease in the dd cross section when isolation is removed.
ISOLATED cross-section dd 11.2156 cross-section rd 10.6045 cross-section df 1.11003 cross-section rf 1.15548 total cross-section isolated 24.0856 pb. cross-section dd 8.32963 cross-section rd 10.6045 cross-section df 1.11003 cross-section rf 1.15548 total cross-section no box isolated 21.1996 pb. (/pawresiso/plotcinbox.kumac) NO ISOLATION cross-section dd 10.3597 cross-section rd 10.9103 cross-section df 5.43021 cross-section rf 4.85456 total cross-section no-isolation 31.5548 pb.
To get the correct cross sections for experiment and theory in the our two x-gamma bins, 0-.8 and .8-1, it is necessary to determine the relative amounts of direct and resolved processes in each bin. Unfortunately these are not known experimentally.
To get estimates of these direct/resolved fractions it is necessary to use models. Two are available, FGH and PYTHIA ( LMZ does not make a separation between direct and resolved). The table below gives the fraction of the direct process in the x-gamma ranges
x-gamma range FGH PYTHIA
0.0 - 0.8 0.15 0.17
0.8 - 1.0 0.76 0.84
Note the larger contamination of the direct by resolved in
the > 0.8 region for FGH compared with PYTHIA.
This is due to the FGH x-gamma distribution being flatter
at large x-gamma than is found for PYTHIA ( see Fig 58.2 for PYTHIA).
From Peter:
email 12 December 2013 13:02
xgamma range Frac of resolved Frac of direct
<0.7 0.847 0.153
<0.8 0.769 0.231
>0.8 0.131 0.869