[Frost] Analysis review comments

[Michael Dugger]

Igor,

The committee is made up of Andy Sandorfi (chair), John Price, and Zhiwen 
Zhao.

-Michael

On Thu, 9 Aug 2012, Igor Strakovsky wrote:

> Hi Mike,
>
> Whois in the committee?
>
> igor
>
> On Thu, 9 Aug 2012 13:20:49 -0400 (EDT), Michael Dugger <dugger at jlab.org> 
> wrote:
>
>> 
>> Hi,
>> 
>> We have received comments from the Analysis Review Committee regarding the 
>> helicity asymmetry for eta photoproduction.
>> 
>> As discussed during today's FROST meeting, some of the comments we received 
>> might come up again for others during an analysis review of g9a data.
>> 
>> At the bottom of this email I have copied those committee comments that are 
>> not specific to the eta analysis. I have placed, what I feel is the most 
>> important committee comment first. This comment deals with the fact that 
>> carbon is not helium or oxygen, and is a question I expect all analysis 
>> reviews will want a good answer for. Any help in addressing the comment 
>> regarding the differences between carbon and helium and oxygen is greatly 
>> appreciated. In fact, I feel it would be good that the FROST group develop 
>> a standard answer to this.
>> 
>> Thanks for your time.
>> 
>> Sincerely,
>> Michael
>> 
>> ---------------------------------
>> Committee comment->
>> 
>> 
>> Assumption A. Nuclei have Fermi momenta and excited states. These
>> additional degrees of freedom can change the momentum balance and
>> result in negative tails on an M^_x distributions calculated with
>> free nucleon kinematics. The shape of the M^2_x distributions, and
>> in particular the fraction of events in the negative tail M^2_x <
>> -0.4 GeV^2, is assumed to be the same for events originating on
>> any nucleus. In particular, region (1) contains 3He and 4He (as
>> coolant in the FROST mixing chamber), 12C and 16O (in the
>> Butanol), 19F, 35Cl and 37Cl (in the pCTFE target cup), while
>> region (2) contains 12C. We would expect the general shapes of
>> M^f_x distributions from quasi-free production on these nuclei to
>> be similar, particularly near the quasi-free peak, but the tails
>> depend on how the 2-body kinematics is miss-matched by Fermi
>> motion and the spectra of nuclear excited states in the recoiling
>> product nuclei. It would be amazing if the tails were identical!
>> We can assume the above equation for the moment, but there needs
>> to be some reasonable (justifiable) estimate for the systematic
>> error associated with this assumption.
>> 
>> Response->
>> 
>> ---------------------------------
>> Committee comment->
>> 
>> p1, Section II:
>> There are 2 different conventions in use for defining the net
>> entrance channel helicity which differ by a sign, either the
>> scalar sum of the particle helicities in the center of mass, or
>> what amounts to the net projection of the total angular momentum
>> along the beam axis. Judging from Section III.F, the authors are
>> using the latter. It would help to avoid confusion if the first
>> equation were augmented to include something like,
>> 
>> E = [H_{1/2} - H_{3/2}]/[H_{1/2} - H_{3/2}]
>>   = [\sigma_{1/2} - \sigma_{3/2}]/[\sigma_{1/2} - \sigma_{3/2}]
>> 
>> where the A and P designations refer to antiparallel and parallel
>> beam and target spin alignments. With this there can be no
>> confusion. There are also different definitions of E  in terms
>> of observable quantities, which differ by signs. The authors have
>> chosen a particular one. A translation table relating different
>> schemes in use in the literature is given in arXiv-1108.5411.
>> Since the assignment of helicity states is referenced to \pi^+
>> production in III.F, it would be helpful to specify the
>> convention.
>> 
>> Response->
>> 
>> ---------------------------------
>> Committee comment->
>> 
>> p1, Setion III:
>> The text states that "energy and momentum corrections" are
>> applied to the proton. Technically, correcting one of these
>> automatically corrects the other. Typically, CLAS analyses
>> correct the "momentum" of the detected particles and the "energy"
>> of the photon beam. The energy of the proton can't really be
>> "corrected" , since its energy is never measured. Rather, because
>> it looses energy passing through the material within CLAS, the
>> measured momentum isnt correct and must be adjusted. Further,
>> note that the kinematic fitting routine does not "correct" the
>> energy; rather, it applies the known measurement uncertainties to
>> determine the most likely "actual" value of a measured quantity,
>> using what amounts to an economic algorithm.
>> 
>> Response->
>> We have removed the mention of energy corrections.
>> 
>> 
>> ---------------------------------
>> 
>> Committee comment->
>> 
>> Section III.A:
>> It would be useful to remind the reader at the outset of the
>> composition of the target region involved in the analysis. In
>> particular, instead of a reference to "FROST", it would have been
>> useful to clearly state the composition as 5 cm of Butanol
>> (C4H9OH) centered at a z-vertex position of _____, __cm of Carbon
>> centered at z = ____cm, and ___cm of CH2 centered at z = _____cm,
>> etc.
>> 
>> Response->
>> We have included the requested text that the 5.28 cm butanol (C4H9OH) 
>> target is centered at a z-vertex position of 0, 0.15 cm of carbon is
>> centered at z = 6.25 cm, and 0.35 cm of CH2 is centered at z = 16.1 cm.
>> 
>> ---------------------------------
>> Committee comment->
>> 
>> p2, Section III.B:
>> What was the range of the live-time during the runs? There are
>> typically some runs with abnormally low live-time, due to
>> beam-steering, etc, but these seem to have been included.
>> 
>> Response->
>> 
>> 
>> ---------------------------------
>> Committee comment->
>> 
>> p3, Section III.C:
>> This is a single-hit analysis whenever more than one tagger
>> channel fires within an RF bucket (1 ns), the event is discarded.
>> A more general way to treat such multi-tag events is to update
>> each of the histograms associated with those energies for which
>> tagging channels fired within the true timing peak; multiple
>> over-counting is then corrected at the end by an accidental
>> subtraction. This is a little more work, although not all that
>> much, and potentially improves statistics. The improvement might
>> be negligible if the fraction of multiple tags is small. Since the
>> statistical errors on the extracted asymmetries are appreciable
>> for some bins, this should be addressed  perhaps with a plot of
>> the multiplicity in the tagger, or at least a statement about the
>> relative fractions
>> 
>> Response->
>> As noted, there is two ways of dealing with multiple tagger hits
>> for an idividual event: Either throw away hits that are ambiguous,
>> or loop over each possible photon. We have always chosen
>> to throw out ambiguous photon events so as to keep the signal as
>> clean as possible (Differential cross sections using g1 data for
>> eta, etaPrime, pi0, pi+, as well as for beam asymmetries using
>> g8b data for eta, etaPrime, pi+, pi0).
>> 
>> ---------------------------------
>> Committee comment->
>> 
>> p5, Section III.F
>> The discussion of identifying beam helicity states is confusing.
>> The text states that "the helicity 3/2 state was assigned a
>> negative sign" based on \pi^+ production. I presume this means
>> that the E asymmetry for single \pi^+ production was assumed to
>> be negative near 900, which it would for
>> E = (\sigma_A-\sigma_P)/(\sigma_A+\sigma_P). That fixes the beam
>> helicity states and those assignments were use in the
>> (\gamma,\eta) analysis. Please clarify the wording.
>> 
>> Response->
>> 
>> ---------------------------------
>> 
>> _______________________________________________
>> Frost mailing list
>> Frost at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/frost
>
> Igor Strakovsky, SAID CNS The George Washington University
> Tel: 703-726-8344(NV),202-994-4742(FB),Skype: igors1945_2
> Fax: 202-994-3001(FB),Emails: igor at va.gwu.edu, igor at jlab.org
>