[E02013] GEn prl

[Gordon D. Cates]

Hi John,

There are several reasons why I want to stay with F1d/F1u vs F2d/F2u.   
First, I have had interactions with both Craig's group as well as  
Jerry and I am not convinced that the comparison with theory there is  
as meaningful.  In both cases the F1 calculations seem to be on more  
solid ground.  While it is nice to challenge theorists to do better, I  
believe we are already doing that, and some of the details we would be  
challenging (with F2) have more to do with the groups wanting to get  
something out on F2 than on whether they are incorporating into their  
calculations meaningful underlying assumptions.   Secondly, there is  
real and interesting physics in the Q2 behavior of F1d/F1u.   It truly  
tells us about the relative spatial distribution of the d and u  
quarks, something we recognized up front and something that Marc  
Vanderhaeghen has since written about.  There is also the interesting  
connection between this and the ratio of d/u in pdf's.  I have already  
gotten some very interesting reactions showing the F1d/F1u plot.   It  
allows us to make some quite interesting comments about nucleon  
structure.

Finally, we want to submit the paper.  We needed to make a choice.    
The reviewers forced us to expand our discussion of systematics (which  
I think was fine and understandable) and even suggested where we  
should cut.  The reviewer (I believe it was reviewer A) wanted us to  
cut out the entire figure.   This seems like a reasonable compromise.

One last thought.  Since there will be a long paper, hopefully within  
the next year or so, we will have at least one more opportunity to  
publish it.

I am just speaking for myself here, but I would vote for moving  
forward with the paper as it is now.   Between your efforts, plenty of  
work from everyone else who gave feedback, and reviewers who clearly  
tried to do a thorough job, I think we have a greatly improved  
manuscript.  It was extremely difficult to accommodate the reviewer's  
comments without running over.  I shudder at the thought of starting  
on that endeavor again.

Sincerely,

Gordon


On Oct 30, 2010, at 11:17 AM, John R. Arrington wrote:

>
> Seamus,
>
> I agree with Pete that having both the F1d/F1u and F2d/F2u ratios was
> better, and it didn't save much space, since the new single-plot  
> version
> is almost as large.  If you need to remove one for space, the F2 ratio
> appears to be more useful as there is a much greater spread in the
> calculations.  With most of the calculations for the F1 ratio being in
> pretty good agreement, the new data doesn't add much here.  I still  
> think
> it would be good to show both, but it seems that the F2 ratio is more
> useful if there's only room for one.  Maybe there's some specific
> conclusion that only comes from F1, but the only specific conclusion  
> in
> the paper appears to be the comparison to lattice, but that's only a
> comparison to the previous data at low Q^2.
>
> -John
>
> On Sat, 30 Oct 2010, Seamus Riordan wrote:
>
>> Hi Pete,
>>
>> Thanks for the thoughtful attention to our results and paper.
>>
>> We have made some progress on the lowest Q2 point analysis, but due  
>> to
>> manpower considerations, as our graduate student is leaving for a few
>> months, there will not be any serious work on it again until February
>> (barring any problems with visas, etc.).  It is our hope to have
>> something for the June Hall A collaboration meeting.  As you can
>> imagine, we're very interested in completing and sharing it.
>>
>> Regarding any further changes, we don't anticipate anything
>> significant.  Issues that will be pursued for the long paper include
>> further refinements to the Monte Carlo we use for the inelastic  
>> analysis
>> as well as addition work Gordon is doing to improve the target
>> polarimetry uncertainty.  It's my hope that this paper will also  
>> induce
>> more enthusiasm in the theory community to think about (and  
>> calculate)
>> 3He FSI relevant to spin-dependent measurements.  All of these  
>> aspects
>> have appropriate systematic errors on them right now and we expect  
>> any
>> further work to simply reduce them.
>>
>> The F2d/F2u figure was removed solely to reduce space, which was
>> suggested by one of the reviewers.  This along with several parts  
>> of the
>> text, were removed only after considerable deliberation, as we felt  
>> all
>> aspects of the paper were worthwhile to the presentation of our
>> results.  Additionally, we feel adding in our fits to the paper  
>> enhances
>> the discussion of the differences.
>>
>> Thanks again for your comments and questions,
>> Bogdan, Gordan, Nilanga, and Seamus
>>
>>
>> Pete E.C. Markowitz wrote:
>>> Hi Bogdan, Gordan, Nilanga, and Seamus,
>>>
>>> Congratulations!  The reviewers obviously agreed that the results
>>> are important and that the collaboration did a good experiment.  I  
>>> think
>>> you also were able
>>> to answer their comments.
>>>
>>> As a collaborator, I do have a couple of questions but I doubt
>>> that you need to change your draft to answer them.
>>>
>>> One of the reviewers asked about overlapping data with other  
>>> experiments,
>>> and you mentioned the low Q2 point under analysis (but it not in  
>>> the paper).
>>> Could you tell us what the status of that point is?  When might  
>>> the analysis
>>> be completed on that point, and is there a preliminary number which
>>> you can share with us now?
>>>
>>> You are all doing an commendable job of pursuing the systematic
>>> uncertainties, and
>>> I applaud you for implementing the change between the May 2010  
>>> results
>>> and the
>>> present October 2010 results. The change is modest (and basically  
>>> brings
>>> us back to
>>> December 2009).  But having 3 sets of numbers in 10 months might  
>>> suggest
>>> that the
>>> numbers are possibly not quite final?  Are there any issues you  
>>> know of
>>> which could again change the
>>> results?
>>>
>>> One intriguing comment was regarding the Plaster results.  The  
>>> disagreement
>>> is slightly less with the (new) October 2010 results, but the  
>>> point that
>>> the results may not
>>> be compatible escaped me until the reviewer pointed it out.   
>>> Plaster was
>>> deuterium,
>>> and this 3He result has different systematics but the GEn values are
>>> still more than 1 sigma
>>> away from each other.  [Your new fit sort of splits the  
>>> difference, and
>>> your statement that
>>> it is about the same fit even without the new data suggests that the
>>> final point from Plaster
>>> may be inconsistent with the other previous data.]  However, any
>>> discrepancy is de-emphasized
>>> in your current paper, especially by removing the lower panel in  
>>> Fig.
>>> 2.  Although the results in
>>> the upper panel might be consistent, the results from the May 2010  
>>> lower
>>> panel of Figure 2
>>> look like they could not be consistent.  Was that why it was  
>>> removed?
>>> It obviously does
>>> not save any real space, since you blew up the top panel.
>>>
>>>    -pete
>>>
>>> On 10/23/10 4:10 PM, Bogdan Wojtsekhowski wrote:
>>>
>>>> Dear Collaborators,
>>>>
>>>> We have received a response from PRL which was quite positive.   
>>>> Two of
>>>> the referees made some constructive suggestions for changes, and we
>>>> have modified the manuscript accordingly.  Attached to this email,
>>>> please find three documents.  One, labeled GEn_update.pdf, is the
>>>> updated manuscript, with all sentences in the text that have been
>>>> changed highlighted in blue.  The second, Response_to_ed.txt,  is,
>>>> as the name suggests, the text we have prepared to send back in an
>>>> email to the editors.  Finally, we also include a document  
>>>> containing
>>>> the response from prl (From_editor.txt).
>>>>
>>>> We note that since the time we submitted our paper to prl and the
>>>> arXiv, we discovered a 3% (relative) error in our target  
>>>> polarimetry.
>>>> Half of this was a bug in a target-polarimetry code, and the other
>>>> half was due to a systematic that, in previous He-3 experiments,  
>>>> was
>>>> negligible, but because of the high performance of the GEN target,
>>>> became large enough (~1.5%) to worry about.  We have made the
>>>> appropriate adjustments to the results that appear in the  
>>>> abstract and
>>>> the tables, and have also noted this in the attached response to  
>>>> the
>>>> editors.  We also needed to increase the target polarimetry error  
>>>> from
>>>> 4% to 4.7%, which has very little effect on our overall error.
>>>>
>>>> Please send any comments over the next week.  We would like to
>>>> resubmit on Friday, October 29th.
>>>>
>>>> Best regards,
>>>>
>>>> Bogdan, Gordon, Nilanga and Seamus
>>>>
>>>
>>>
>>
>>
>> -- 
>> Dr. Seamus P. Riordan			riordan at jlab.org
>> University of Massachusetts, Amherst    Office: (757) 269-5289
>> Post-doctoral Research Associate	Pager:  (757) 584-0051
>>
>> CEBAF Center A103, Jefferson Laboratory
>> 12000 Jefferson Ave.
>> Newport News, VA 23606
>>
>> _______________________________________________
>> E02013 mailing list
>> E02013 at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/e02013
>>
> _______________________________________________
> E02013 mailing list
> E02013 at jlab.org
> https://mailman.jlab.org/mailman/listinfo/e02013

_____________________________________________________________
Gordon D. Cates,  
Jr.                                                           
Department of Physics
Professor of Physics and Radiology                                
University of Virginia
Director,  Institute for Nuclear and Particle Physics     P.O. Box  
400714
Phone: (434)  
924-4792                                                    382  
McCormick Rd.
email:  
cates at virginia.edu                                                
Charlottesville, VA, 22904
_____________________________________________________________