[Clascomment] OPT-IN: A study of the P11(1440) and D13(1520) resonances from CLAS data on ep X eXX pi+ pi X p

[GUIDAL Michel (57321)]

Hi Paul,

On Thu, 3 May 2012, Paul Stoler wrote:

> Hi Kijun:
> Really nice comprehensive study. I am impressed. Two points:
> 1. On the t distributions, for small t I do not understand why GK obtain a dominance of  sigma_t over sigma_l, since the pion pole should dominate there. How does this connect with Hermes?
> 2. Before submission you need to have someone carefully proof read and edit the grammar and usage.
> Anyway, congratulations on a job well done.
> Regards,
> Paul

I think that there is a bit of confusion on which paper you are adressing.
It seems that you are talking about Kijun et al.'s paper on "exclusive 
pi+ above the resonance region" and NOT about Victor's et al. paper on 
"double pi production in the resonance region".
If so, let me first add Sebastian, Michel and Mauri to the list
of recipients of this email and then, respond to your question:

> 1. On the t distributions, for small t I do not understand why GK obtain
> a dominance of  sigma_t over sigma_l, since the pion pole should 
> dominate there. How does this connect with Hermes?

I wrote recently to P. Kroll and here is his answer:

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

>From kroll at physik.uni-wuppertal.de Fri May 11 08:09:14 2012
Date: Fri, 11 May 2012 08:08:49 +0200
From: kroll at physik.uni-wuppertal.de
To: "GUIDAL Michel (57321)" <guidal at ipno.in2p3.fr>
Cc: kroll at physik.uni-wuppertal.de, goloskkv at theor.jinr.ru
Subject: Re: question

Dear Michel,

indeed sigmaL is quite small for Jlab kinematics in contrast to
the situation at HERMES where we have a nice fit to the data. The long.
cross section dominates at small t. Well, we are not happy with this
situation but we always clearly said that our GPD parametrization
is optimized for small skewness. This means that we fitted the GPD
parameters only to data at small skewness. Of course we can use our GPDs
to evaluate observables for electroproduction at large skewness in order
to examine trends and magnitudes.

For pi+ production at large skewness in particular we understand what is
problematic. It is indeed the pion pole contribution and in particular the
pion-nucleon form factor. In our transversity paper you find its
parametrization in Eq. (23). Theoretically (23) looks fine. If the 
exchanged pion goes on shell for t-> m^2_pi the form factor becomes unity.
The phenomenologically problem with (11) is that for forward scattering
t->t_0=-4m^2\xi^2/(1-\xi^2) the term is larger than the parameter 
Lambda_N^2. Neglecting the latter one sees the pion pole contribution to
the cross section behaves as ~1/\xi^4. This is it what makes small the
pion-pole contribution. For HERMES kinematics this is different since t_0
becomes comparable or even smaller that Lambda_N^2 and, hence, the
skewness dependence of the cross section at small t is much softer.

We have a solution for this problem -- replacing in (23) t by t'.
Then our results agree quite well with the sigmaL and sigmaT data from the
F-pi collaboration. They also compare much better with your data.
The price to pay for this solution is that one has to view the form factor
(23) as an effective parametrization. One may also reggeize the pion-pole
contribution. In this situation the form factor (23) is part of the Regge
residue and there is problem with a t' dependence.

Best regards, Peter

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

Peters' answer is quite clear but I am not sure how to take this into 
account in our article. I think that Peter's message is, in a few words, 
that we should take the GK calculation as an extrapolation of a model
well suited for high energies (HERMES & H1/ZEUS) towards low energy (i.e. 
JLab6), without any re-fitting of any parameter, and that some care should 
therefore be taken when making conclusions.
I think that we already try to adress this concern in the present version
of the article, p.15, l4 to 12:
..."This is remarkable, as one should note that the GK model was optimized 
for higher-energy kinematics (HERMES) and that no further adjustment of 
the parameters was done for the present CLAS kinematics. We should also 
note that the GK model is applicable only for small values of the ratio 
t/Q2. Outside this regime, additional higher-twist contributions that
are not taken into account in the GK handbag formalism approach are
expected..."
So, let me ask our ad-hoc committee, Sebastian, Michel and Maurizio,
if, at the light of Peter's answer (who has actually our article in hands
and who is not particularly requesting any correction to our text), 
further precisions/changes/modifications are required, we will follow your 
advice.
Amities,

Michel

***
Michel GUIDAL
Institut de Physique Nucleaire
Bat 100 - M052
91406 ORSAY Cedex
Tel: (33) 01 69 15 73 21
Fax: (33) 01 69 15 64 70
E-mail: guidal at ipno.in2p3.fr
***