From: Armen Bunyatyan <bunar@mail.desy.de>
Date: 30 August 2011 12:41:37 CEST
To: Karin Daum <karin.daum@desy.de>
Subject: forward photons
Reply-To: Armen.Bunyatyan@desy.de


Dear Karin,
I wrote suggestion for modifications. Please have a look, then
we can discuss.3

regards, Armen


***************************************************************************** 
(A) Except for constraining the fragmentation functions, the authors claim that their results are important for understanding high energy cosmic ray data. For this reason it is required, in section 2.4, to give more details on the 'physics' which is included in the relevant MC programs such as EPOS, QGSJET 01/II and SIBYLL. Describe also the differences between them.

Answer to referee:
We add some more details on the description of cosmic ray models.
However we would like to not  go to deep into details, because with this
analysis we are unable to study the differences in details. The model
predictions have been provided by their authors and they are not
intended for the tuning to the individual experimental data.

Modification to the text: add to section 2.4, par.3

These are complex models based on Regge theory \cite{Collins:1977jy}, Gribov's Reggeon calculus \cite{Gribov:1968fc}, and pQCD. They use an unitarisation procedure to reconstruct amplitudes for
exclusive processes and to determine the total and elastic cross
sections. Central elements of these models are the production of minijets and the formation of colour strings that fragment into hadrons.
Whereas this Regge-Gribov approximation is applied to hadrons as interacting objects in the case of QGSJET and SIBYLL, it is extended to include partonic constituents in EPOS. Other important differences among models are the
approximations of saturation effects at low $x$ values and
the treatment of the hadronic remnants in collisions.


\bibitem{Collins:1977jy}
P. D. B. Collins, {\it ``An introduction to Regge theory and high-energyphysics"},
Cambridge University Press, Cambridge, 1977.

\bibitem{Gribov:1968fc}
V.~N. Gribov, % \newblock {A reggeon diagram technique},
\newblock Sov. Phys. JETP {\bf 26}, 414 (1968).
%%CITATION = SPHJA,26,4



*****************************************************************************


(B) In the results section reference is made to the LHCf paper. In this paper QGSJET and EPOS underestimate the data at high energy while SIBYLL overestimate the data. However, at lower energy QGSJET and EPOS overestimate the data which is what H1 also observes. I would have expected a more in-depth discussion of the comparison between H1 and
LHCf  rather than simply saying that the two are in different kinematic
ranges and no comparison can be made.


 Answer to referee:

  In the LHCf paper there is not much more-in-depth discussion of the
  comparison with the models. We can not discuss and interpret their results
  in many details.   We propose to add few sentences.


  Modifications to the text:

A measurement of single forward photons produced in $pp$ collisions at $7$~TeV centre-of-mass energy   at the LHC has been recently reported by the LHCf Collaboration~\cite{Adriani:2011nf}.
The energy spectra are presented in two pseudorapidity ranges
$8.81<\eta<8.99$ and $\eta>10.94$.
A direct comparison of the H1 and LHCf results is not possible
due to the different kinematic ranges of the two measurements.
The LHCf measurement also shows significant discrepancies with the
predictions of hadronic interaction models. Compared to the present measurement there are differences
in the behaviour of the models with respect to the data.
While at lower energies the EPOS and QGSJET~II-03 models overestimate the rate of forward photons by about the same amount as for the H1 data,
the SIBYLL predctions are lower then the measurement, in contrast to
the H1 observations. At high energy  the QGSJET-II-03 model shows harder energy spectrum then the LHCf data and overestimates the absolute rate,
while it predicts about the right amount of forward photons
for the H1 measurement.
The shape of SIBYLL distribution is in a good agreement with the LHCf measurement at $\eta>10.94$, but gets harder for $8.81<\eta<8.99$, similar to the H1 observations. Comprared to the LHCf data, the EPOS model shows softer photon energy spectrum  over full energy range, except the highest values, in contrast to the H1 result.


******************************************************************************

(C) In the summary section towards the end: '?All these models predict different spectra in xL and pT'. This statement is not very useful. Of course they are different but what does this mean? In the very next sentence it would be useful if you could quote the disagreement in %.


Well, we can repeat in Summary that LEPTO and CDM Monte Carlo
models overestimate the total rate of photons by about 70%,
and EPOS, QGSJET and SIBYLL models by 30-50%.
We can also repeat the sentences from the Result section that
EPOS and SIBYLL predict harder spectra, while QGSJET spectra
are slightly softer and describe the mesaured cross sections
at higher xL and pT within experimental errors.

However, for a such short paper, the result section is just a
half page above and this repetition doesn't make much sense.


******************************************************************************

Also please correct the following: (1) Introduction -  Line 3: 'understanding of proton fragmentation'. Give a references.

  change to '.. understanding of the fragmentation of proton remnant'
  Then it is obvious that aprticle production at very low angles
  provide information about remnant fragmentation.


(2) Introduction = Line 6: '..cosmic ray data'. Give references.

 OK, suggest to add references to

\bibitem{Engel:2007tv}
R.~Engel,
%\newblock {\it Implications of HERA results for very high-energy % cosmic ray physics},
\newblock Nucl. Phys. Proc. Suppl.~{\bf 75A}~(1999)~62 [astro-ph/9811225].
%%CITATION = ASTRO-PH/9811225;%%

\bibitem{Bunyatyan:2009zz}
A.~Bunyatyan {\em et~al.},
%\newblock {\it Model predictions for HERA, LHC and cosmic rays},
\newblock  {Proceedings of the Workshop on the Implications of 
HERA for LHC Physics, Geneva, Switzerland,
26-30 May 2008, p.611}, DESY-PROC-2009-02 [arXiv:0903.3861].




(3) Introduction - Second paragraph. Make clear that you are referring to this specific measurement. 'The measurement of the photon production'

This sentence is a general statement. We also refer to previous measurements.

(4) Monte Carlo section (page 8) second paragraph: 'The DJANGOH..'. change 'generates' with 'generate'.

    OK

(5) Monte Carlo Section (page 9) line 1: 'QGSJET II'->  'QGSJET II-03'

    Doesn't matter, both names are used in the references. Can be changed

(6) Results section (page 11) line 2: 'The QGSJET model predicts slightly softer' ->   Change 'slightly' to something more quantitative

  Add a sentence something like that:
  While at  lowest x_L and p_T values the QGSJET overestimates the
  measured cross sections by about 40%, at higher x_L and p_T
  the predictions are consistent with the data within experimental errors.

(7) Several of the figure captions start with very long sentences. It would be worth rewriting them so that they read better.

    OK

(8) Figure 6, caption: ?The inner error bars shows?? No inner or outer error bars are visible. Also 'shows' -> 'show'

   OK, remove 'inner'



