RSS Analysis Meeting Minutes

RSS homepage.

09/08/2008

1. Karl :
   
   • Circulated draft version 1.8 of moments prl. Discusses updates.
   • Oscar points out that there is a new version of the Kulagin and Melnitchouk paper, although it has not yet shown up on the arXive. It appears to have been published already as PRC 77 015210 (2008). Also suggests the text of the prl be changed to make clear that the technique is good "for moments".
   • Oscar requests addition of error band to Fig. 1 g2ww. He will provide values to Karl.
   • With regard to discussion of alpha_S, Oscar suggests we mention that the SSF results are clearly inconsistent with the GLS results.
   • Oscar notes that the integration range is 1085-1910. The lowest W bin is 1085-1095 (central value of 1090).
   • Oscar has some of the systematic values needed for the table.
   • We should mention in the text that M11 is not very different from Gamma_1. Nachtmann moments mostly important for higher moments like d2.
   • We might want to include a second small data table to list our results for alpha_s,ELT and f2.
   • Oscar reminds that our analysis indicates failure of ELT. We should discuss with E,L or T to try to understand the implications. This also emphasises the need for accurate low x g2 measurements.
   • Karl will check the values in his table against the values Oscar posted for moments.
   • Goal to send out to collaboration this weekend.
2. Shige :
   • Shige circulated a draft of the RSS PRC.
   • Two sections are written.
   • Oscar notes that we should be consistent (or clear) about which versions of the R/F1 codes we are using in all publications.
   • Shige went over the PRC outline.
   • Mark suggests we drop ENGINE section, or atleast reduce it to short discussion.
   • Oscar wrote the Poltarg section already and will forward to Shige.
   • Mark will write the remainder of experiment sections, and deuteron section.

07/29/2008

1. Karl :
   
   • Circulated some new plots.
   • Karl should compare the neutron moment results he obtained using the constant D-state correction to the direct integral of Shige's neutron structure functions. Shige will provide tables to Karl.
   • Oscar evaluated the DIS part of the ELT sum rule, and finds that it is about 50% of the Resonance part.
   • The notation for d2 in the new prl has to be consistent with the first prl, (and preferably the most standard usage in the literature). Karl has to check that he is being consistent with use of d2bar, d2twiddle, etc. We published d2bar, which doesn't include DIS. Note: In all Oscar's notes/emails he refers to 2M23 as d2 (=0.0036).
   • We could reduce systematic on Gamma_1^N. Also the statistical error looks incorrect. Karl will check.
2. Shige :
   
   • Will work on long paper.
   • Discussion of error in Wally's paper (plots are correct).
   • Oscar concludes that the largest error is less than 4% from using gamma_D=0.925 instead of an x-dependent correction.

07/08/2008

1. Karl :
   
   • Circulated Draft 1.6 of prl.
   • Discussion of the draft led to the following conclusions:
     1. Oscar points out that Wally's latest paper claims that neutron D-state correction (1-3/2P_D) is good to the 1% level even at Q2=2 GeV2. This is very significant to the moments analysis and should be mentioned in the text. This constant correction also properly accounts for the effect of tensor polarization in the deuteron.
     2. The RSS fit curve should be added to Figure 1, evaluated at the data Q2, and including the fit error band.
     3. Karl should evaluate the contribution to the integrals in region 1073No need to include all the DIS data in Fig. 1., Simula curve can stay, no need to include MAID since we are not concerned here with the x-dependence. Have to mention that previous prl had an error in g2ww plot.
     4. Karl should use the fit errors when evaluating the resonance integrals instead of scaling the data errors to the fit result.
     5. Mark will send Karl the latest Deuteron F1/R fit.
2. Oscar :
   
   • Oscar and Shige have been in contact with Wally who has a better prescription for the smearing needed in the neutron extraction. Oscar is trying to evaluate if the effect of the new procedure would have any impact within our error bar. He suggests that the best way to proceed is to estimate the size of the effect and inflate our systematic accordingly.
   • Wally is working now to extrapolate his results to RSS kinematics.

06/10/2008

1. Karl :
   
   • Circulated Draft 1.5 of prl.
   • Need to define all symbols in Eq. 3. Oscar provides reference for the original formulation of the OPE expansion of Gamma_1 and Gamma_2: Phys.Lett.B323:439-445,1994 eqs. 35-38.
   • Oscar provides links for many of his previous results to include in the paper: 1 2 3 4 5
   • Need to add SLAC neutron and remove MAID curve from Fig. 2.
   • Karl is including Mark's D fit into his moments code in order to evaluate the deuteron moments and compare with Oscar.
   • Oscar notes that his DIS fit to deuteron data has better uncertainty than his neutron fit, so Karl should use the deuteron fit.
   • Oscar's value for f2 comes from integration of the fit, with data error propogated as a percentage of the central value. He concludes there is very little twist-4, atleast within our huge errors. We need to compare our results to previous publications. The DIS contribution is twice as big as our measured resonance value.
   • Oscar points out that there should be an error associated with the value of gamma_D quoted in our paper.
2. Shige :
   
   • In communication with Wally about the neutron extraction. Wally wanted to know what smearing method Shige used (=Atwood-West).
3. General :
   
   • Next meeting : Tuesday June 24 1:30.

06/03/2008

1. Shige :
   
   • Contacted Wally and he gave us MAID xg1 and xg2 at Q2=1.0 and 2.0. According to Karl, Wally seemed to be using MAID 2000 while Karl was using MAID 2003. Karl found a new version (MAID 2007) and downloaded it. He believes that the new one will not give us much impact on the MAID integrals.
   • Wally also kindly gave us his calculations of the ratio R = g1d/g1N of structure functions for the full model used in his paper (PLB 346, 165 (1995) ). It turns out the values he gave us are pretty close to what we have (differ by ~1% or so). So, the neutron results will not change much. Wally said he doesn't believe that his calculations are valid at low Q2: What we can do is to do the D-state correction in two different ways (using a constant depolarization factor and using Wally's calculations), and report the difference in neutron results as systematics.

05/08/2008

1. Karl :
   
   • Circulted draft 1.4 of the moments paper.
   • Specific suggestions made: Eqs. 1 and 5 can be simplified.
   • Text really only applies to proton for now, it needs to be generalized.
   • For Deuteron, do we include QE contribution to evaluate `nuclear' sum rules, or stick with proton, neutron, singlet and non-singlet results above pion threshold? It seems like the second choice is more realistic as we might need several more months analysis to clean up the deuteron QE results. Also, it would be difficult to make the paper general enough to cover both nucleon/nuclear sum rules.
   • Oscar points out that we have to use Wally et al's x-dependent D-state correction, and reference it in our paper. It only deviates by about 13 percent from the constant correction, and then only for our highest x point. More details on Shige's page (April 10,13 and 23 of 2007).
   • Mark points out that all the plot curves are not explained in the text. Also, each of the Table entries is not really discussed yet either. Fig. 2 should also include g2d, then it will provide our measured quantities. The introduction needs explanation of neutron extraction, and x-dependent D-state correction.
   • Oscar requests description of singlet, non-singlet, etc. be put back in the table.
   • Oscar will provide M13 calculation, and reminds Karl to include f2 in paper.
2. Shige:
   
   • Has been discussing with Wally about the apparent difference in MAID model results between his paper and RSS. He will request Wally's MAID results at Q2=1,2 GeV2 (before interpolation) so we can directly compare to our table. This should resolve whether it is just an interpolation problem or if the two models really disagree.
   • Shige is working on draft outline of long paper and will try to post it by the end of next week. Oscar suspects that we have planned to include too many things in the paper and may have to cut out some topics.

04/15/2008

1. Karl :
   
   • Circulated updated manuscript for moments paper.
   • Oscar has no major comments, but will forward some suggestions to Karl.
   • Mark suggests to include a g2p plot or possibly a plot with both g2p and g2D (g2n?). Might be good way to correct the g2ww result on our g2p plot from previous prl instead of an erratum.
   • Oscar points out that there is already a discussion of the low x part of g2p and g2ww on page 4, in his section on higher twist. Suggests to move this to the BC sum rule section.
   • In d2 section just need to refer to earlier discussion and give error values.
   • Karl points out that nachtmann moment machinery is primarily useful to enable clean higher twist extraction. Shouldn't we use them to do so to get f2?
   • Oscar has already extracted f2, and will recirculate the result. If we have enough room we can put it in the prl.
   • Karl notes that two papers with heavy theorist input extracted higher twists without using Nachtmann moments even though g2 data was available. Oscar notes that the largest effect is in the M23 (d2) moment. The M11 (Gamma_1) integral has a smaller effect. So this may be why the papers Karl reference don't use nachtmann moments. In addition, Simula and Piccione & Ridolfi have made very strong statements that the nachtmann moments must be used for clean extraction of dynamical higher twists. Finally, Oscar points out the limitations of using the scaling variable x_bj in our kinematic region: Nachtmann variable is really more appropriate for any Q2<2 GeV2.
   • Karl working to round up all results that Oscar has evaluated for the moments that are spread over several different emails, for incorporation into the paper.
   • Oscar requests additional column added for deuteron systematics.
   • Shige suggests change to abstract : ``neutron from deuteron'', instead of neutron (deuteron).
2. Shige:
   • In discussion with Wally regarding the Maid model results used in his paper which seems to place the delta at a different x (0.67 instead of 0.63) than we expect.
   • Also the highest shown model x is much less than what is seen in the data.
   • Our previous plots of MAID seem very different than what is in Wally's paper.
   • Shige working on the long paper, Oscar suggests to circulate a rough draft with placeholders for the unfinished sections.
3. General:
   • Next meeting : Thursday May 8, 1:30 pm

04/03/2008

1. Karl :
   
   • Evaluated DIS part of M11 Nachtmann moment using the Bianchi-Thomas parameterization. (See Oscar's Feb. 27 email for further details.)
   • Find's results that are within a few percent of what Oscar's g1 fit gives.
   • There remains a large discrepency in the integral that depends on g2WW. Karl believes this is because Oscar produced g2WW from his DIS g1 fit, while Karl produced it from the RSS resonance fit. He will evaluate again using the DIS fit.
   • Oscar asks Karl to also evaluate the integrals for the deuteron. Bianchi-Thomas provide a proton and a neutron so Karl will combine to form a deuteron (only correcting for the d-state). Oscar points out that there is no need (or possible method) to perform Nactmann moments for Gamma_2 which is not part of the OPE. CN moment is all there is for Gamma_2. There is no M21 which is what you'd need for Nactmann B.C. Oscar finds that the overall Nactmann moment for Gamma_1 not much different from the CN moment. This is in contrast to what Dong finds for the d2 (M23) integral in which the nactmann correction is quite large at our Q2.
   • Karl has been working on the prl draft. Incorporated about 70% of comments provided by Oscar.
   • Mark suggests discussing TMC earlier in the paper, during the introduction to Cornwall-Norton moments. Will help in explaining the motivation for using Nactmann moments.
   • Mark suggests including Gamma_2 plot, and notes that the discussion of GDH integral needs work.
   • Table should have similar results for Gamma_1 (Singlet, Non-Singlet, proton, deuteron, neutron) and Bjorken Sum Rule.
2. Oscar :
   
   • Has numerical values for everything except Gamma_2 for the deuteron. Bjorken integral seems almost completely insensitive to TMC (CN = Nactmann moments). Gamma_1 integral shows very little difference also between CN and Nactmann.
   • Shows summary table with numerical values.
   • In the paper, we will include the Nactmann values if space permits, otherwise just discuss that they are very close to the CN moments and point out why. Oscar sent email to Karl yesterday suggesting how to phrase this.
   • Oscar also extracted an effective \alpha_s from our Bjorken integral. The results have pretty good uncertainty compared to the world data.
3. Shige :
   
   • Discusses paper of Kulagin and Melnitchouk. In it, the authors form a deuteron from the MAID Proton and Neutron after applying nuclear corrections. The paper is at Q2=2 GeV2, so Shige contacted the authors and got the results at Q2=1.3 GeV2 to compare with RSS data.
   • Shige shows the comparison to RSS which is not very close. In particular, it does not seem like even the Delta lands at the same x_bj in data and calculation. Also, the maximum x_bj in the calculation is about 0.75 while the data goes to x>0.8. Shige will show to Wally to get some comment.
   • Shige is also working on the long paper.
4. General :
   
   • We will meet next on Tuesday April 15 at 1:30pm.

03/13/2008

1. Karl :
   
   • Incorporated Oscar's higher twist section into moments paper, and restructured the paper to follow the outline agreed upon last meeting. Also converted references to bibtex. Goal is still to have paper ready for the arXiv in April.
   • Oscar asks Karl to finish work on DIS part of g1 (Nachtmann moment of the B.T. function). See Oscar's Feb. 27 email for further details.
2. Oscar :
   
   • Discusses his moments technical note, and clarifies some variation in notation between authors.
   • Oscar points out that there is a typo in his note: \delta d_2^{DIS} should just be d_2^{DIS}.
   • Believes we should show BC sum rule plot with both proton and neutron in the paper.
   • Oscar is reviewing the primary reference for the ELT sum rule and will forward it to Shige for discussion in the long paper.
3. General :
   
   • We will meet next on Thursday April 3 at 3:30

02/20/2008

1. Oscar :
   
   • Suggests new structure for the moments paper:
     1. Intro : Moments, sum rules and matrix elements. Non-OPE Sum rules : BC and Bjorken. Definition/comparison of C.N. and Nacthmann moments. Target Mass Corrections.
     2. Results for First moments of g1 : Gamma_1^p, Gamma_1^D,Gamma_1^Sigma, Gamma_1^NS,Gamma_1^N,Bjorken,I_GDH.
        1. Subsection to discuss elastic contribution and why we don't need to include.
     3. Results for First moment of g2 : BC
     4. Third Moments : d2^p,d2^D,d2^Sigma,d2^NS,d2^N
        1. Subsection on Higher Twists and polarizabilities : f2
     5. Second Moments : ELT using fits only, Valence region. This section may be better suited to the long paper.
   • Karl will work on Sections 1-3. Oscar will work on Sections 4,4a. Section 5 may be moved to Shige's long paper.
   • Oscar requests Karl to evaluate DIS contribution to Nachtmann moments.
   • Oscar suggests "We present (first?) evidence for dynamic higher twists" to be added to the abstract.
   • Regarding long paper: Shige will evaluate ELT sum rule in long paper. Oscar will send us all the relevent ELT reference.
   • Shige is working on his sections now. Oscar requests that Shige compile together all the sections he has recieved so far into an initial rough draft. We agree to use bibtex for citations, and Mark will establish a CVS repository for the document so we can efficiently make changes.
   • Next Meeting : March 11

02/19/2008

1. Shige :
   
   • Working on error propagation for g1p,g2p. After discussion, it appears that Shige's work from April 02, 2007 is sufficient and nothing more is needed. This was the last piece of analysis for Shige.
2. Oscar :
   
   • Releases new version of his moments technote.
   • Points out that there does not appear to be any twist-3 visible in our g2p at low x.
   • Expects that the two different formulations of d2 (equation 17) will agree when they are evaluated at constant Q2. Oscar will verify.
   • Oscar has obtained estimate of the DIS contribution to d2.
   • Concludes that we see twist-3 effect to 3 sigma in proton and 2.5 sigma in deuteron.
   • General discussion on how to structure the moments paper. Seems best to try to put first and higher moments into one paper. If it does not fit it can be moved to second paper or to the long paper.
   • Introduction to moments paper should contain brief discussion of d2 and Nachtmann moments. For example, Eq 10 of Oscar's TN can be shown and then demonstrate how it reduces to C.N. moment.
   • Oscar requests that we all recheck our responsibilities for the long paper and send any outstanding material to Shige.

01/08/2008

1. Oscar :
   
   • Found an error on our published g2ww_p result. calculated g2WW for proton and deuteron. Both approach toward zero at low x, so this makes more sense. (see his email sent yesterday for details). Discussed how we should report this error and we prefer to mention this in our new upcoming paper rather than publishing an erratum. But Frank was not at the meeting, so we don't know his opinion.
2. Shige:
   
   • Will recalculate the smeared proton errors as his approach (apr02, 2007) looks incorrect. Will also update the neutron errors. As suggested by Oscar, dg1/g1 and dg2/g2 are calculated easily knowing the delta_F1/F1 of ~3% and the data fit errors (dA1 and dA2) for each w bin.
   • Recalculated the deuteron DF systematics due to nucleon model. (the results posted on oct 30, 2007 were wrong because an incorrect version of MC was used.) see Shige's webpage dated dec18, 2007
3. General:
   
   • Next meeting will be on Jan 29.

11/29/2007

1. Shige
   
   • Final systematic : 6% on Deuteron DF.
2. Karl
   
   • Running the simula model to compare with data.
   • Will work on paper over holiday and after PAC.
   • Mark points out that there is a way to submit to prl from the arXive which preserves the citations.
   • Goal is to have someting on the ArXive by April.
   • Mark favors 1 PRL, not two papers.
3. Oscar
   
   • Points out some ambiguities in the references. For Bjorken sum rule gv/gA evolved to Q2=5, must use the PDB value divided by 6: Gamma_1^N-Gamma_1^P = gA/6
   • Oscar has a link to website for getting alpha_s at any scale. From alpha_S, can derive Lambda_QCD.
   • We should be assuming two flavors, since charm threshold is near Q2=1.3. Oscar points out that NF,Lambda,Beta are all scale dependent, but there is very small change withing error.
   • Oscar has 3 parameter fit to SLAC deuterium with good chi2 for use in Bjorken.
   • Not certain whether to use LO or NLO alpha_S.
   • Believes we should use Nachtmann moments instead of C.N. moments.
4. General :
   
   • Next meeting December 18

10/30/2007

1. Oscar :
   
   • Updated RSS configuration
   • Changed VNC configuration to use 256 instead of 16 colors. We should now be able to use acroread properly.
   • Going over long archive paper.
   • Trying to incorporate his work into systematic section.
   • Suggests some modifications to Shige's outline. Mark modifies the posted outline with new suggestions and to reflect the responsible person for each section. Also restructures the systematic section.
   • Oscar will add some background definitions to the asymmetries section to make it more appropriate for a general target.
2. Shige :
   
   • Working on the deuteron DF systematic study.
   • Evaluated the uncertainty due to nucleon model, but would like to recheck.
   • Discussion of the "B-factor" in propogation of deuteron DF systematic. Oscar believes Shige can probably use the proton "B-factor", since it will be an overestimate. From Feb. 09, 2006 webpage B=1.
3. General :
   
   • Shige will be out of country. We will meet next on Nov. 27th.

10/17/2007

1. Oscar
   
   • What to do with Wilson Coefficients? Nothing, unless we want to compare measurements at different kinematics. For example, RSS vs SLAC, or if we want to correct our data.
   • d2 can use only 2 flavors : u and d. Oscar prefers we use two flavors for RSS, 3 for SLAC.
   • No corrections to d2 at leading order. Our prl values is fine.
   • Oscar discusses the various contributions to the nachtmann moments from different kinematic regions.
   • Oscar recommends a book from T. Mutan (Reference 8) to understand QCD.
   • Notes a large discrepency with SLAC deuteron d2 Nachtmann moment. Indicates something wrong with SLAC or us, but Cornwell Norton moment agrees???
   • Oscar asks Karl to verify his moment calculations.
   • For d2, we can use g2ww to estimate uncertainty, but can not use it to estimate unmeasured part directly (d2 has no twist-2).
   • Final word on RSS d2 : "3sigma indication of higher twist", "1st measurement ever of definite higher twist".
   • Regarding the long paper: Oscar incorporated Shige's new systematic errors, and is working on the target section. Mark suggests reorganizing the systematic discussion. Karl will send Oscar discussion on the R.C. systematics.
2. General :
   
   • Next meeting in two weeks : 10/30/2007

09/25/2007

1. Shige
   
   • Has completed the Deuteron dilution factor systematic study.
   • Provided result to Oscar and Frank.
2. Oscar
   
   • Discusses topics for the moments prl.
   • Oscar points out that d2 is pure twist 3, so it is incorrect to estimate the missing low x contribution using g2ww(twist-2) as Karl has been doing.
   • Oscar notes that at low x our g2 data is consistent with g2ww, so that we can assume that the twist-3 part is zero (ie there is no DIS contribution to d2). Then we can use the result of the g2ww integration as an estimate of the systematic uncertainty.
   • Oscar added a new section on this topic to his note, along with a section on the flavor structure of the 3rd (and higher) moments.
   • He proposes that we evolve our d2 up to the asymptotic value instead of vice versa as we have been doing in the plots. Similarly for Bjorken: we may find that we have a very competitive error bar.
   • Regarding the moments paper, Oscar proposes the following structure:
     1. Karl's intro
     2. 1st moments Gamma_1, Gamma_2 and Bjorken. Both Cornwell-Norton and Nachtmann.
     3. Higher twist
     4. If space permits, a further section on Efraemov-Teryaev sum rule.
   • Oscar will repost an updated version shortly.
3. General
   
   • Next meeting: Oct 19th.

09/04/2007

1. Oscar
   
   • Produced first draft of moments technote focusing on the QCD radiative corrections, Nachtmann moments and NLO corrections (section in preparation).
   • Basic conclusion : Oscar believes that our Nachtmann results don't need any QCD corrections: they are fine at LO. The Wilson coefficients are used to relate measurements at different Q^2, so we can use them to evolve from SLAC down to RSS or vice versa (assuming pQCD works), but our results are just fine as they are in LO.
   • Ji et al. have also worked out the NLO corrections, which would help us in getting a better value for the asymptotic d2 from our measured one. Oscar is working out the proton and neutron results from the singlet and non-singlet ones that are provided in Ji's papers.
   • Oscar requests comments on the draft. He plans to circulate among a few theorists also.
   • We will incorporate a summary of this work into the letter.
   • Oscar points out that there is several different conventions relating G1,G2 and g1,g2 in the literature. This can lead to significant error if not treated properly, of the same order of magnitude as the "standard" corrections : TMC, low-x resummation....
2. Shige
   
   • Working on draft of long paper.
   • Using this opportunity to organize his study of Packing Fraction and other systematics.
3. Karl
   
   • Preparing talk for crimea2007.
   • Several suggestions/comments for Karl's presentation are discussed.
   • Oscar has some comments for the moments letter: would like us to include Bjorken and ELT sum rules. For this we may need a fit to the neutron data.
4. Frank
   
   • Will produce systematic error bars for deuteron from Oscar's tables. Needs final Deuteron D.F.
5. General
   
   • We will meet again on Sept. 25.

08/14/2007

1. Karl
   
   • Updated proton RC section for long paper.
   • Circulated first draft of moments prl.
2. Shige
   
   • Updated neutron results with Mark's new Deuteron fits. Did not change much. The changes in the fit arose from the TE/calibration constant. There was a small variation in the error for g1n.
   • Shige finds g1,g2~0, except for one point at x=0.65 which is about 3 sigma from zero. He has posted a table of his results.
   • Expects to have packing fraction yields by next week.
3. Oscar
   
   • Working on g1d duality. Finds good agreement for global. Also split the region in two for a limited local test : 1.09Oscar discusses several theory papers relevent to RSS. He is working on a technote to summarize what he has found.
4. General
   
   • We now move to bi-weekly meetings.

07/31/07

1. Karl
   
   • Updated moments technote to reflect recent comments from Oscar.
   • Posted first draft of proton RC section for long paper.
2. Shige
   
   • Working on updating his neutron results.
   • Progress interrupted by JLab network failure.
3. Oscar
   
   • Wrote target section for long paper. Mentions that best source he has found after reviewing old papers is from Mark's elastic paper.
   • Suggests that Shige should not worry too much about the exact formatting requirements of PRC for this initial draft.
   • Oscar has evaluated duality integrals for the deuteron using the GRSV and AAC tables from Peter at Q2=1.2 and Q2=1.4 GeV2. He interpolates these results to RSS kinematics.
   • Global duality looks good (although partially this is due to larger deuteron errors).
   • He will look at some limited local duality integrals over a restricted W range.
   • Oscar has also evaluated d2D and d2n. d2n is now consistent with world data, although with quite large uncertainty.
   • Also evaluated the systematic errors for Gamma_1^D and d2D. They are slightly larger than the corresponding proton errors.
   • Oscar is considering what we should publish of the higher moments beside d2.
   • He is also working on a technote to discuss the moment's logarithmic QCD corrections. So far, he has concluded that we do indeed need to evaluate the Wilson Coefficients as described in S. Matsuda and T. Uematsu. He believes the discussion in the Ji paper refers only to the BC sum rule.
   • Oscar may not be available to meet next week.

07/17/07

1. Oscar
   
   • Applied fits to Deuteron to get sum rules, d2 and duality integrals. Not quite finished.
   • Oscar is finding d2n approximately zero.
   • Oscar recommends that we set the DIS part of the A2D fit to constant if we use a constant value for A2D.
   • Finds d2D=0.0048, with very little dependence on Q2 evolution. Compare to our proton d2p=0.0057. When deuteron is corrected by D-state correction we get d2D approximately equal to d2p. So d2n is about zero, but slightly negative.
   • This is very different from what Karl found for d2n (see 01/16/2007) with Shige's early neutron results. The old results where consistent with the hall A results, and Slac results within error.
   • Oscar: It's not clear we are doing anything wrong, but we have to understand this.
   • Shige notes that he is now using Peter's F1, which is per nucleus.
   • Oscar has begun to evaluate the deuteron duality integrals. He has GRSV and AAC from Peter (tables at Q2=1.2 and 1.4 interpolated to RSS), but he does not have Soffer's statistical model for deuteron. Oscar doesn't believe it is necessary to have too many PDFs for Deuteron, since we found there is relatively small differences among versions for proton.
   • Oscar is now using the fit instead of data values to evaluate the systematics that are proportional to central value. But, this still hasn't smoothed out the sytematic fluctuations, in fact it made it worse.
   • Oscar points out that we need to evaluate the ELT sum rule.
2. Karl
   • Posted "final" table of proton moments, and moments technote.
   • Found 15% relative error on Gamma_2 DIS contribution. This is larger than what Oscar had recommended, but only slightly larger than the Gamma_1 DIS error.
   • Added theory discussion and introduction which can serve as starting point for PRL.
3. Shige
   • Working on long paper.
   • During general discussion sections of the paper are assigned to group members. It is agreed that we should all have a draft of our relevent contributions ready in two weeks. The assignments are:
     • Experimental description: Mark.
     • Polarized Target : Oscar.
     • P.F. : Shige.
     • Engine : Frank.
     • Proton RC : Karl.
     • Nitrogen Correction: Mark/Oscar.
     • Fit to Data : Oscar/Frank.
     • Deuteron analysis and results: Shige.
     • Deuteron RC : Mark.
     • Smearing and Neutron Extraction: Shige.
     • Systematics : Oscar.
   • We will meet again in 2 weeks: July 31.

07/03/07

1. Oscar
   
   • Working on the deuteron systematic. Noticed that the systematic is very "jumpy". This is because some of the errors are directly proportional to the asymmetry which has large statistical fluctuations. Oscar notes that we will publish results in 45 MeV bins.
   • Karl suggests to use the fit instead of the data to avoid statistical fluctuations in the systematic error.
   • Oscar mentions that the new UVA postdoc is Vladis T. He will start in September.
   • Reading the paper of Song et al. He is confused by the discussion just after eq. 28 "... the combination of gT and g1 relevant to d2 receives no radiative correction." Does this mean we don't need to evaluate the Wilson coefficients? Oscar requests that we read the paper so that we can discuss.
2. Shige
   
   • Working on Deuteron dilution factor systematics. He contacted Peter, who estimates the deuteron model uncertainty is 5% (3-4%) in the quasielastic (inelastic) region.
   • Shige is following the same procedure as for the proton: He increases the cross section by the uncertainty and then observes the change in the packing fraction (about 1%). He then reevaluates the dilution factor with the new values. Dilution factor changes by 2.6% (2.5%) for the parallel (perp) setting.
   • Also looking at the small uncertainty due to run selection.
   • Shige is very busy with new job, but still working on RSS in his spare time.
3. Mark
   
   • Is up to date on all his RSS tasks.
4. Karl
   • Updating his moments note and data table.
   • Planning to have draft publication in July.
   • Oscar points out we should evaluate the ELT sum, and compare to hall A.
   • Oscar requests that Karl relay to his hall A collaborators the utility of producing a fit to the hall A data which can be evaluated at any Q2.

06/26/07

06/19/07

06/12/07

1. Shige
   
   • Following Peter's suggestions to evaluate deuteron systematics. Now running jobs on the farm. Peter claims 3-4% uncertainty on the inelastic deuteron model, and 5% in the quasielastic. These are the values Shige will use in his DF calculation.
   • Shige needs about 4GB of disk space for his work. Mark will move some of the Bigcal files.
   • Shige is following the procedure outlined in his Feb. 09, 2006 webpage entry.
   • Oscar expects 4.5-5% uncertainity.
2. Mark
   
   • Since there are new Deuteron polarization values, he redid the A1, A2 fits. Notes that Shige should use new fit parameters that Mark posted on his webpage. Shige will update.
   • Mark sent an email regarding the fits to the RSS group.
3. Oscar
   
   • Shows graphical errors on deuteron g1,g2.
   • Compares the deuteron systematic to previous proton result. The deuteron systematic is slightly worse than the proton.
   • Oscar expects to have the uncertainty on Gamma_1, Gamma_2 next week, then we should have all the ingredients for the papers. The goal is to have drafts of both papers in July.
   • Oscar has been looking through a paper of Ji et al, for the needed Wilson coefficients.
   • Points out an error in the paper of S. Matsuda and T. Uematsu. They were looking only at 3 quark operators, and neglected gluon operators. Ji et al. reanalyzed and have values for the singlet and non-singlet case. From these we can get the neutron and proton coefficients.
   • The other option is to just quote the ratios and not explicitly evaluate the coefficeients (example M23).
   • Papers: X.D.Ji, W.Lu, J.Osborne and X.T.Song, ``One-loop factorization of the nucleon g2 structure function in the non-singlet case,'' Phys. Rev. D62, 094016 (2000) and A.V.Belitsky, X.D.Ji, W.Lu and J.Osborne , ``The singlet g2 structure function in the next-to-leading order,'' Phys. Rev. D 63, 094012 (2001).
   • No meeting for two weeks, until after SANE readiness review.

06/05/07

1. Mark
   
   • Working to incorporate the change in deuteron polarization due to new T.E. values into the RSS analysis.
   • Reran Frank's script to produce updated asymmetries.
   • Redid the deuteron radiative corrections.
   • The polarization is 3% lower than what we thought previously.
2. Oscar
   
   • Performing additional systematic studies.
   • Oscar points out that the Nitrogen asymmetry correction is larger in ND3 than in NH3.
   • Oscar has requested that Shige evaluate the deuteron dilution factor systematic.
   • Still need to evaluate the contribution from F1, R. Oscar expects he will need a few weeks to finish this.
   • Oscar mentions that he is reading a very instructive early paper by Nachtmann "Is There Evidence for Large Anomolous Dimensions?" Nucl.Phys.B78:455,1974.

05/29/07

1. Oscar
   
   • Reviewing the Deuteron T.E.'s to determine systematic. Oscar found that the standard deviation of the results was about 10%. Assuming we know GEP to a few %, the we would get smaller systematic from Q.E. asymmetry. But this would require detailed study of the Deuteron dilution factors in the QE region.
   • Oscar's results: Bottom Para gives 10% syst from standard deviation of T.E.s, slightly smaller value from the absolute deviation. Bottom Perp : 3.6% variance. Top para: 4%.
   • In regard to Frank's g1d experiment, Oscar mentions that the GEN experiment achieved 2.5% uncertainty on polarimetry, but it required detailed study. Also, the deuteron target can take 150 nA easily. The RSS target average was 120 nA.
   • Oscar has also been doing some further research into target mass corrections.

05/22/07

1. Oscar
   
   • Discusses his spreadsheet used to evaluate deuteron systematic errors. The spreadsheet combines the radiative correction systematic with all remaining errors. Work in progress.
   • Oscar is following the same procedure as with the proton. However, Pb and Pt will be separate contributions.
   • He is using the 12 configurations of RC results that Mark provided to evaluate RC syst.
   • In general, he finds something like a few percent RC error for inelastic, and a slightly larger error at low W.
   • For total systematic, he finds a similar trend: several percent uncertainty everywhere, but rising to 10-12% at low W.
   • The statistical errors are generally much larger than the systematic.
   • Oscar needs reasonable numbers for the polarimetry uncertainties in order to complete his work. He reviews the deuteron T.E. data which suggests 6-8% systematic for the target polarimetry.
   • With regard to the moments paper, it is not clear how we can evaluate the Wilson coefficients that are discussed in the S. Matsuda and T. Uematsu paper. We should publish M23 (=Wilson coefficient * d2).
   • Sum rules paper should include Bjorken non-singlet, ELT sum rule.
2. Shige
   
   • Rechecking his codes.
   • Lots of training to go through for his new job.

05/15/07

1. Karl
   
   • Evaluated M23, and M11 integrals to compare to Oscar's earlier result.
   • Seems to be factor of 2 problem with the error when compared to Oscar. Both will recheck their results.
2. Oscar
   
   • Oscar's conclusions: Plenty of real twist-3 in our data. The TMCs contribute less than 1/3 of the Cornwall-Norton result.
   • Oscar summarizes Dong's new paper: "To properly treat target mass effects, use nachtmann variable as discussed in the S. Matsuda and T. Uematsu paper." Integrals such as Int g1 dxi (where xi is the nachtmann var) are meaningless.
   • Need to evaluate the Wilson coefficients E1N,E2N.
   • We may need to mix the neutron/proton results to get singlet/non-singlet pieces.
   • Dong/Matsuda and Uematsu are the first believable treatment of Higher Twist in spin structure functinos.
   • Frank: How difficult will it be to persuade others that this is the proper treatment? Oscar: Matsuda and Uematsu paper is a classic, nobody questions the approach. This is a good approach for any experiment which measures both g1,and g2. The only reason it was not applied before is because of lack of g2. The Ridolphi and Piccuni paper, along with Matsuda and Uematsu clarifies all issues.
   • Oscar is working on the Deuteron systematics. It is straightforward but tedious.
   • We should include M23, M11 in our sum rules paper, in addition to standard moments.
3. Mark
   
   • Found that varying A2 model in the Radiative Corrections has very little effect, only varying A1 matters.
4. Shige
   
   • Produced an outline of the deuteron/neutron long paper.
   • This will also be the long paper of the experiment. It should contain the final proton results, but no need for further discussion.
   • The detailed theory discussion should focus on neutron extraction.
   • Probably will only be able to evaluate global duality integrals, not local.
   • Sum rules paper should have more theory.

05/08/07

1. Mark
   
   • Testing the deuteron RC systematic by varying the A1,A2 fits by their errors during the RC procedure. Mark evaluated 12 different combinations that arise from the follow possibilities: A1, A1+dA1, A1=dA1. A2(constant), A2(constant)+dA2,A2(constant)-dA2. A2(fit), A2(fit)+dA2,A2(fit)-dA2.
   • Will provide these results to Oscar.
2. Oscar
   
   • Working on d2 and target mass corrections. Reading the new Stamenov and Siderov preprint.
   • Oscar mentions that Dong's latest paper advocates using the TMC method proposed in 1980 by S. Matsuda and T. Uematsu. Nucl. Phys. B168, 181 (1980).
   • Comparing various versions of R to what Peter's code uses for Deuteron. Oscar sees a very large difference among R's at low W. This will increase the systematic.
   • Points out that the relevent Wilson coefficient must be included in order to compare data to Dong's prediction of d2.
   • Karl asks if Oscar's Gamma_1p systematic study can be applied to Gamma_2p. Oscar will look into it.
   • Oscar shows a study he performed evaluation d2 with the Nachtmann variable, with zero and nonzero mass. He concludes that the target mass corrections are small. In particular, Karl can use Scenario I for his Gamma_2 DIS estimate. The g2WW formed in this scenario only has a 10% contribution from target mass corrections. The rest is twist-2.
   • Oscar now looking into the issue of "higher spins".
3. Karl
   
   • Working on including LSS, BB pdf's in his code for evalution of Gamma2_DIS estimate.

04/30/07

1. Mark
   
   • Worked on Deuteron RC systematics. First he tried to vary the DIS fit form as discussed last meeting. But he found that there was really no significant variation in the total fit anywhere but at threshold. So now he is going to vary the A1 and A2 fits by their error bands. Mark will look at the following possibilities: Hold A2 fixed at a constant value and vary A1, Hold A2 fixed with a functional form and vary A1, Hold A1 fixed and vary A2 by its error band.
2. Karl
   
   • Discussion on error from g2 integral Dis contribution. Conclusion: Karl should use "Scenario I" to form g2 DIS integral, as was done in E94010 and E155x. Then estimate the error strictly from variation of PDFs, no Regge fits. Oscar suggests to normalize the PDFs in the resonance region to the ratio of PDF to Regge fit in DIS. This is due to the fact that the TMC PDF's do not match the data even at low x. Also, Karl should include more PDFs, LSS in particular, but maybe also BSB.
   • Karl : if we normalize the PDF's to match world data, then shouldn't we do that also for our duality integral?
3. Oscar
   • Looking at Dong's newest paper with Mark. They would like to discuss several issues with Wally. In particular the relationship between the a2(3?) matrix element, and the explicit calculation of TMCs.
   • Oscar has found that in the rush to publish PRL, we underestimated the contribution to the sytematic from the elastic tail. It is a pretty small effect and will only effect the PRL plot near threshold. No effect on any published numbers.

04/23/07

1. Shige
   
   • Posted datafiles for g1,g2 proton(smeared), deuteron, and neutron.
   • Using F1,F2 from Peter's code which is for free F1,F2. Shige needs smeared F1,F2.
   • This week Shige produced A1n, A2n from g1n, g2n.
   • Talked to Peter regarding the smearing of F1,F2. He modified Peter's code F1F2INEL.f to give smeared quantities.
   • Shows comparison of his latest A1n, A2n to what he found last year. At that time he did smearing on DeltaSigmas. Now he forms A1,A2 directly from g1,g2.
   • Shige is assuming an arbitrary 10% uncertainty on F1n (smeared). Will spend some time trying to quantify this number.
   • Finds that the D-state correction is signifcant at large W. The x-dependent correction differs from the (traditional) constant correction only for 1st W point.
   • Shows a plot of g1n,g2n with new binning, compared to world data.
   • Writing a report on neutron extraction.
   • Frank: "what remains to complete the neutron analysis?". Shige wants to recheck all codes, and quantify the F1,F2 error estimate.
   • Oscar will check the sytematics as he did for proton.
   • Mark has done some checks of the deuteron radiative correction model dependence. He'll summarize.
2. Mark
   
   • Our g1d is consistent with the EG1b data. Mark notes that A1d is essentially independent of Q2 (within the error bands of the Eg1b data) for the Q2 range relevent to the RSS radiative corrections.
   • Points out that for the deuteron, all available models are quite poor. It's not reasonable to use them to evaluate the RC systematic. Oscar suggests that Mark instead quantify the effect on the RCs of using different functional forms for the deuteron fit. The variation of the additive term Arc is what is relevent.
   • Peter's code is posted on his webpage.
3. Oscar
   
   • Requests Frank update his deuteron SUMMARY files.
   • Oscar recently reread Wally's PRC and Zein-Eddine's higher twist PRL, along with Dong's new paper. a2 matrix element (3rd moment) should represent TMC term, but this is not what he finds when he tests with the PDFs. He is in discussion with Wally about this
4. Karl
   
   • Performed study of proton g1 higher moments.
   • Ratio of PDF/RSS becomes larger for each successive higher moment.

04/16/07

1. Shige
   
   • Applied the Melnitchouk et al x-dependent D-state correction. Found very little effect except for highest x point which changed by about 10%.
   • Shows plots of g1n,g2n. They are consistent with world data. Also consistent with zero over most of x range.
   • Shige will test some other bin sizes and post tables of his results on the web.
   • Oscar requests a short writeup describing the fits that Shige used so we can proceed with the integrals: table of smeared proton results and final neutron, along with code to get A1n, A2n.
   • Shige is using F1n,F2n from Peter's code F1F2IN06.F. It's not clear whether they are free or smeared. Shige will discuss with Peter.
   • Oscar suggests Shige should go directly from g1n,g2n to A1n,A2n.
2. All
   
   • Mark discusses new target mass corrections paper from Dong.
   • Oscar reread the earlier Dong TMC prl regarding the application of TMCs to the twist-2 part of g2. He concludes that since we get g2ww from our data, we don't need to apply TMCs to our g2ww, they are already in by definition.
   • Mark raises the issue of using nachtman variable for d2 as recommended by Dong. Not clear whether we should do this.
   • Oscar points out that we need sum rules for the Deuteron/Neutron, then we are done RSS.
   • To estimate the DIS contribution to Gamma_2, Karl uses g1(resonance) to form g2ww. Oscar requests Karl to look at 4th and 6th moments of ratio of PDF/Data to estimate the contribution of higher twist in our resonance g1. Significant high twist contributions would show as increasing relative differences between our fit and the pdf's with increasing moment. Oscar points out that this is discussed in A. Deur's Bjorken sum rule paper.
   • Oscar mentions that we might want to extract f2, and polarizabilities from RSS data.
   • Oscar will be at JLab next Thursday.

04/10/07

1. Karl
   
   • Updated moments note to discuss Gamma_2, and the uncertainty of DIS contribution.
   • Oscar suggests that it may not be valid to extend the Regge fits into the resonance region for "Scenario III". Also, believes that NLO-2 versions of the pdfs can be assumed to be independent of NLO-1 versions.
   • Oscar suggests several possibilities to evaluate the uncertainty of the DIS contribution.
   • Karl shows comparison of data to chiral soliton models.
   • Group reviews and critiques Karl's talk for DIS07.
2. Shige
   
   • Sees interesting structure in neutron g1,g2. Mark remarks that it probably arises from the F2D that Shige is using to form the structure functions.
   • Mark suggests that Shige should not mix fits and data when extracting the neutron.
   • Shige raises the issue of how to treat the A2 fit error propagation. Presently, Mark fits the entire resonance region A2 to a constant with an error given by the sigma/sqrt(N). Does this mean that we must treat the entire resonace region as a single bin?
   • After lengthy discussion, conclusion appears to be that if we use the fit we are indeed left with a single bin for the resonance region. Oscar suggests that Shige should perform the deuteron-proton subtraction using data, not fits. On the other hand, the fit should be used for integrations.
   • Shige shows size of the D-state correction in the structure functions: Not very significant for x>0.7.
   • Mark used Peter's R for radiative corrections, not R1998. He requests that Shige should doublecheck which F2,R he is using to create his g1d fit.
3. Oscar
   
   • Discusses the D-state correction. He remarks that the correction, in fact, has an x-dependence. It is not constant. This topic is treated in a paper by W. Melnitchouk et al: PLB 346, 165 (1995).
   • Oscar digitized the correction using WINDIG (windows digitizer) and created a text file which Shige can use to compare to the 'standard' prescription. Mark notes that the correction is apparently independent of Q2.

04/02/07

1. Shige
   
   • Improved his calculation of g1,g2 statistical error. Previously he evaluated the error by simultaneously varying A1 and A2 by their respective errors dA1 and dA2.
   • Now he adds the errors in quadrature: ( d(g1)/dA1 )^2 dA1^2 + ( d(g1)/dA2 )^2 dA2^2.
   • The error is reduced by a factor of approximately 0.7. Oscar notes that this latest error is similar in magnitude to Shige original attempt which came from propagating the relative error of rebinned data through the smearing.
   • Shige is using version 6c of the fit for his work.
   • Oscar suggests that we discuss the sytematic effect of using other fits in a long paper.
2. Mark
   
   • For the deuteron radiative corrections, he suppresses the R2 (~1350 MeV) resonance. The three Delta resonance parameters are allowed to vary. For R3, and R4, he allows the amplitude to vary, but fixes the centroid and width to the proton fit values.
   • For the DIS part he is using the V6c fit : DIS ~ (b_0 + b_1x + b_2x^2 + b_3x^3)x^alpha, allowing only the alpha parameter to vary. The remaining parameters are fixed to the proton fit values.
   • Mark also tested the V622 DIS fit : DIS ~ b_0x^alpha. Both look reasonable, but the V6c seems to reproduce the data better. Mark will post the chi2/N.
   • Mark is using 15 MeV binned data for the fitting.
   • Karl notes that the fit produces a tiny error near threshold, which is not representative of the data.
   • Oscar believes this is due to having a single (hence well defined) free parameter in the DIS.
   • Mark notes that the DIS fit is dominated by the more precise data at larger W.
   • Discussion follows. Perhaps we need to include a systematic error in our fit uncertainty due to the effect of various functional forms. Oscar suggest to take the variance of several different fit forms as the systematic error.
3. Karl
   
   • Evaluated two new DIS fit forms DIS ~ b_0x^alpha, and DIS ~ (b_0+b_1x)x^alpha.
   • Both give excellent reduced chi-squared.
   • The parameters of the first form have smaller errors, which is consistent with Karl's earlier study that found that only one or two parameters are needed to satisfactorially reproduce the DIS background.
   • Oscar notes some oddities in Karl's spreadsheet and asks for clarification. Karl promises to look into it quickly.
   • Also working on DIS07 talk. He will circulate a version for next monday's meeting.

03/26/07

1. General discussion
   
   • Shige notes that the RSS work disk is almost full. Mark will try to find some space for the SANE/GEP files that are currently being stored there.
2. Shige
   
   • Still having some difficulty with the fitting routine. He tightened the convergence criteria and found that the error matrix elements go to zero.
   • Shows g1p smeared with new errors which come from the A1,A2 fit.
   • Shige believes the resulting error may be too big.
   • Mark suggest that Shige use ( d(g1p smeared)/dA1 )^2 dA1^2 + ( d(g1p smeared)/dA2 )^2 dA2^2. The necessary derivatives can be calculated numerically. This method assumes no correlation between A1 and A2.
   • Shige discusses the effect of rebinning on the errors. Mark points out that we now have a fit with continous error band so rebinning should have no effect.
   • Mark points out that we should use a fit to deuteron when we subtract out the proton fit. Presently, Shige subtracts the proton fit from the deuteron data to get neutron.
   • Oscar notes that the fit integrals should conserve the data relative statistical error, as we do with d2.
   • Shige is working on systematic error study for deuteron.
   • Shige is leaving at the end of April, but will continue working on RSS.
3. Mark
   
   • Found that the new nitrogen asymmetry correction has negligible effect on the deuteron radiative corrections.
4. Karl
   
   • Studying the DIS fit to A1,A2, by evaluating several different functional forms to see which parameters are significant.
   • Oscar suggests Karl add two further forms to his study : DIS ~ b_1 x^alpha and DIS ~ (b_0 + b_1 x) x^alpha.
   • Karl will send these results to Oscar.
   • Karl has talk at DIS07 around April 18th. Mark suggests to try to finish this week.

03/19/07

1. Mark
   
   • Redid the Deuteron RCs with new Nitrogen asymmetry correction.
   • Made some adjustments to T_in, T_out thicknesses. Did not see any significant change.
   • Mark is trying to decide how to implement the new changes in the Quasielastic. He notes that PWIA does not reproduce the Deuteron Quasielastic very well.
   • Peter has been working on this issue for CLAS.
   • Mark must leave early since he is Run Coordinator this week.
2. Shige
   
   • Catching up on the RSS emails sent while he was away.
   • Having some difficulty with the fitting routine : Matrix elements don't converge.
3. Oscar
   
   • Asks Frank what causes the asymmetry to be set to zero below the pion threshold in the RSS_SUMMARY files.
   • Frank will check it out.
   • Mark will post latest Nitrogen asymmetry correction files with values below the pion threshold.
   • Discusses evolution of data to constant Q2.
   • Comes to the conclusion that the systematic error is conservative.
   • It appears that we are basically assuming A1 is independent of Q2. The Q2-dependence comes mainly from the F1 fit.
4. Karl
   
   • Oscar noticed that errors on the DIS fit parameters were huge in pass VIb.
   • Karl: seems to be an issue with the RC iterative procedure. Ran pass VIc radiative corrections, this time truncating the iteration at 5, instead of 11.
   • Very little effect on the RCs, but the fit parameters have more reasonable errors.
   • Karl suspected that this is due to using too many fit parameters (5) to describe the DIS background.
   • Evaluated 5 different DIS forms for goodness of fit
     1. no background
     2. constant
     3. linear
     4. quadratic
     5. cubic
     6. cubic*x^alpha
   • All forms gave reasonable chi2/N. Best Chi2/N comes from assuming the DIS is a constant. From this Karl concludes that the current DIS parameters don't hold much significance.
   • Oscar suggests Karl test the forms :A1~x^a, and A1~b x^a, A1~ (b0 + b1x)x^a, which are more justified.
   • Oscar also notes that a fit to sig_TT, sig_LT could use the RSS dis Regge fit.
5. Frank
   
   • Regarding the compression of x-range caused by evolving to constant Q2: Asks Oscar whether crossover of the original/final curve is significant.
   • Oscar states that it is purely kinematical.

03/12/07

1. Karl
   
   • Performed new pass of radiative corrections (VIb) to include the new Nitrogen Asymmetry correction advocated by Oscar. (New Nitrogen asymmetry correction uses radiated, acceptance averaged cross sections instead of born xs.)
   • Only significant difference (to the data) from pass VIa is just at threshold.
   • Oscar points out that the DIS portion of the A1,A2 fit has changed. Asks Karl to investigate.
2. Oscar
   
   • Reran the duality integrals.
   • Found small differences with previous version.
   • Oscar asks Karl to review the RC systematic method we used last year. He requests Karl to send updated model file to evaluate the systematic.
3. All
   
   • It is agreed we will attempt to update the prl plots and tables before publication.

03/05/07

1. Shige:
   
   • Working to include DIS and elastic contributions into his smearing technique.
   • He uses Oscar's Regge fit for DIS, and the standard formulation for elastic in terms of form factors.
   • For g2, he assumes g2WW, with g1 in the resonance region given by the Regge fit. Karl notes that there is significant difference in g2ww depending on which g1 is used in the resonance region.
   • Shige finds that the contribution from elastic and DIS are very small: 0.001 absolute maximum contribution to g1 and g2. He plans to recheck his analysis.
   • Regarding the error matrix, Shige still finds some differences with Karl's tables. They will meet to discuss.
   • Shige will be away next week.
2. Karl:
• Updating his moments technote to include Gamma_2.
• Discusses g2WW estimate of DIS contribution to Gamma_2. g2WW depends on which g1 model is chosen. He uses the RSS fit in the resonance region, and has checked the effect of using various DIS g1 models in the unmeasured region. He finds a maximum variation of 10% in g2WW depending on which DIS g1 he chooses.
• Oscar requests that Karl also quantify the effect of using other models (instead of the RSS fit) in the resonance region.
• Karl discusses the elastic contribution to Gamma_2. He is using the Mergell et al Form Factors (from 1995), and assuming 5% error on this contribution to the BC sum rule.
• Mark points out that 5% is probably an overestimate, and that 3% may be more reasonable.
• Mark will also forward the most recent FF parameterization, that he used for the RSS elastic analysis.
3. Frank:
• Mark inquires about PRL status.
• Frank: the PRL has received "editorial acceptance", but there may still be some space issues.
• Frank will wait till this is clarified before informing the collaboration.
4. Oscar:
• Has an improved Nitrogen correction for protons, which has the advantage that it does not diverge at the pion threshold.
• Oscar points out that the Nitrogen correction should utilize radiated, acceptance-averaged cross sections, not Born cross sections.
• The main difference from the previous results is in the first few W bins (plot).
• Since we discard the first bin and quote a very large error on the remaining points around threshold, there should be little impact on the results. However, we will need to verify.
• Oscar emphasizes that any formulation of g2WW should not include an elastic contribution. As pointed out by Wally Melnitchouk, g2WW is twist-2 and therefore can not include elastic.

02/12/07

1. Shige:
   
   • Working on propogation of A1,A2 fit parameter errors. He is having problems reproducing the official error matrix.
   • Will continue his debugging.
2. Oscar:
• Working on DIS fit for g1p,g1n.
• Neutron fit uses E143 (neutron) and E155 (D-P).
• Oscar notes the big correlation between the exponent parameter and the higher twist (Q2) coeficient.
• Suggests that we use his fit down to x=0.1 and then use E154 value down to zero.
• Will try to include Hermes and E154 data in his fit.
3. Karl:
• Working on moments.
• Has set up a new summary page.
• Compares Bianchi and Thomas DIS fits (p and n) to Oscar's fits.
• Notes that Bianchi and Thomas DIS fit does not include E155. We should contact the authors to see if they have an updated fit.
• Oscar suggests we look at Frank's E155 g1/F1 fit.
• Oscar requests that Karl send him Bianchi and Thomas DIS neutron table. He will make global comparison slide.
• Oscar points out that Karl's neutron Gamma_1 slide is using data, not a fit. Also, it is the integral of smeared neutron, not unsmeared.
• Shige suggests we compare integral of smeared/unsmeared proton fit to see the effect on the integral.
4. Mark:
• Notes that there will be a plenary Spin talk in Munich. We need to consider which slide(s) to provide the speaker for RSS and SANE.

02/19/07

02/12/07

1. Frank:
   
• Sent prl back to editors on Friday before he could implement Mark's comments.
• Frank will not go to April APS.  Missed the deadline.
• Oscar mentions that James Maxwell can present some RSS slides at APS during his SANE talk.
  
2. Mark:
• Completed Deuteron RC with Arenhovel's new calculations.
• Mark shows the Q.E. asymmetry model curves which are large at large W.  Points out that these large values are suppressed by f_rc.
  
• Mark shows f_rc and A_rc, which are very flat beyond the Delta (==> so will not change data shape very much).
• Discusses the structure of his data file.  Karl requests unimplemented entries like sigma_LT be set to some obvious value (like zero or -1E6)
• Mark is using Shige's latest Dilution factors. He found a non-negligible difference of about 10% near threshold and elsewhere from the old DFs.
• Shows fit he is using for A1 in R.C.  For A2 he uses a constant: the weighted average.
• Most A1 fit parameters seem significant.
• Only one iteration needed in radiative corrections.
  
• Karl can use these fits for the Q2 evolution of Deuteron.
3. Karl:
   
• Working to organize the moments work.  Asks what topics will be in paper.
• Mark : Proton, Deuteron and Neutron moments should be in prl.  Neutron extraction and full neutron results should go in Shige's archival paper.
• Oscar : include plots only when we can compare to existing data : Hall A for neutron, Hall B for proton, etc.
4. Oscar:
• Discusses neutron smearing.
• Oscar made a fit to E143 and E155 g1p data (1<Q2<1.5) for low x (plot 1 plot 2).
• Suggests that we compare Bianchi and Thomas prediction to this 3 parameter fit.
• Shige can use the SLAC W>2 GeV data to constrain his error propagation through the smearing.
• Oscar points out that g2WW at low x agrees well with our g2 data as shown in Frank's prl.  So as long as we use a reasonable g1 fit, g2ww should be good at low x.
• Oscar and Mark point out that the range of EG1 Gamma_1 integral has to be checked.  First they don't have same resonance region range.  Second, their "total" integral actually includes data that they measured above W=2GeV.  Karl promises to look into this.
• Regarding ELT sum rule: Oscar states that we can not get ELT directly from the proton and deuteron integrals.  The x-dependence is not the same for proton and deuteron.  We really need g1n(x) and g2n(x).  Might be best to work from fits to the data.
• Oscar points out that ELT is an exact sum rule, and that the sea contribution cancels.  Left only with valence.
5. Shige:
• Working on the propagation of statistical error through the smearing.
• Shige made a fit to the stat error of A1 and A2.  From this he can predict error at any x.
• Using this fit he propagated A1+-dA1 and A2+-dA2 through the smearing.  He finds huge error at low x which he doesn't think is reasonable.
• Extensive discussion follows.
• Mark points out that the A1,A2 fit has errors on the parameters which reflect the statistical uncertainty of the data.  In light of this, we probably should not propagate the error point by point. He suggests that Shige could randomly vary each fit parameter simultaneously over a Gaussian distribution with a sigma the size of each error bar.  Then evaluate the resulting Gaussian sigma of the fit at each W-point. 
  

02/05/07

1. Oscar
   
• Oscar discusses the differing formulations of d2. ie, how to get from the (g2-g2ww) form to the expression using g1 and g2.
2. Frank
   
   • Going to Florida in April. Will not be able to go to DIS07(April 16) in Munich.
   • Discussion of who can go instead.
   • Oscar: Need answer for Rolf this afternoon.
   • Frank expects to send latest version of the PRL this week. Not totally certain how to address all comments yet.
   • Regarding the referee's request to include some standard equations (A1, A2, etc) it is decided to reference the E143 paper instead.
3. Mark
   
   • Included the new Arenhovel QE results in his radcor code.
   • The new grid covers 3755 < E0 < 5755 and the full range of W.
   • He started running today and expects results tomorrow.
   • Mark will also include Shige's latest Dilution factors.
   • Next step is to evaluate the Deuteron systematic.
   • P_b : uncertainty will probably be same as in the proton paper.
   • P_t : we have more TE's but the error is slightly larger.
   • Mark mentions that the structure of his deuteron radcor files is slightly different from the proton files.
   • Mark points out that none of the available deuteron models do a very good job reproducing the data. It may be best to vary the RSS model over some reasonable range and use this to evaluate the RC model dependence.
4. Shige
   
   • Regarding the deuteron dilution factor systematic, Shige will discuss with Peter.
   • Neutron systematics will be even harder to evaluate.
   • Shige is also trying to properly propogate the statistical error through the smearing. It's not clear how to address the very large statistical uncertainties at low and high W.
   • Frank points out that the neutron results will have systematic contribution from proton, neutron and also the extraction itself.
5. Karl
   
   • Evaluated the Q2 evolution of deuteron and neutron g1 structure function and moments by assuming g1/F1 is independent of Q2.
   • Oscar prefers that we use functional form of A1D and A2D as we did for proton.
   • Mark thinks he turned off DIS Q2 contribution in his fit, will double check.

01/29/07

01/16/07

1. All
   
• Review of last few meetings. Shige summarizes the details.
2. Karl:
   
   • Regarding quark polarizations: Following Oscar's suggestion to form delta u/u and delta d/d for W>1.77 GeV Q2>1 RSS data.
   • Shows latest result. They don't seem correct. Mark suggests that Karl should use "per-nucleon" values for the deuteron structure functions.
   • Karl has formed neutron Gamma_1, Gamma_2, and d2 from Shiges tables of smeared neutron RSS structure functions. Shows comparison to Hall A E94010 neutron results at slightly lower Q2.
   • Karl has not yet evolved the RSS data to constant Q2, and the W coverage is not the same for the two experiments. Nevertheless, the data seem to match up reasonably. He will make the needed corrections.
   • Oscar requests that DIS,elastic, etc. contributions be included so we can evaluate the full sum rules. For DIS, fit the lowest x data and then extrapolate to x=0 (Regge form is probably best).
   • Regarding Q2 dependence: Mark states that his fit of deuteron A1 contains no Q2 dependence. He removed it because it seemed to give unreasonable results. Peter suggests we assume g1/F1 is constant and get Q2 dependence solely from F1. He points out that there is a plot in his latest proposal showing g1/F1 for all world data that demonstrates the relative Q2-independence for W>1.8, Q2>1.
   • Karl also tried to extract the neutron moments from g1p, g1d using only D-state correction. Did not get reasonable results, which was surprising, given that nucl-th/9709015 claims this to be good to 3%. Will repeat more carefully.
   • Points out nice webpage of deuteron and proton data. Peter warns that some of this data may be incorrect. The page was set up by a summer student and may not be up to date.
3. Shige:
   
• Shows plot of RSS g1n compared to hall A data.
• All : discussion of quark polarizations and "per-nucleon" issue.
• Peter points out that his code F1F2IN06 provides per-nucleon results.
• Oscar notes that the RSS data at large W used for the quark polarizations satisfies Q2>1.15, so we have similar constraints to Hall B results.
• Shige is working on his Hall C talk. Will send out something to the group to review shortly.
4. Mark:
   
   • Notes that deadline is approaching fast for several conferences: INPP 2007 in Tokyo, Japan, Pan Pacific High Energy conference in Vancouver (August), and Baryons 07.
   • Peter will probably go to Pan Pacific. Karl is considering Baryons 07.
   • Mark is writing new code to interpolate the latest Arenhovel tables. He will then run them through the R.C. code again and post the new results. Does not expect large change.
   • Notes that Frank has not received any response from the referees yet.
   • We will resume regular meetings next monday.

01/08/07

1. Shige:
   
• Checking to see if his smearing method agrees with Peter's code.
• Peter first looks at the momentum distribution functions. He splits the range 0 to 1 into 20 evenly spaced bins, each with equal probability.
• Shige is modifying Peter's code so that it will also provide cross sections. Then he can perform a direct comparison.
• Regarding the Hall C talk on Jan. 25th, Shige has about 15 slides already prepared.
• Mark and Frank suggest that Shige probably does not need to show the asymmetries.
• Karl will send the sum rule/moment xmgrace files to Shige.
• Peter mentions it would be nice to compare our d2n to hall A's result.
• Karl will form d2n from our data if Shige wants to show it in his talk.
• Peter can add RSS to the hall B duality plots. He will make some plots for Shige.
• Mark will send the relevent tables to Peter (g1p & g1d vs x at constant Q2).
• Peter suggests that we assume g1/F1 is constant and then assign all the Q2 dependence to F1.
• Peter is using code F1F2IN06 for smearing.
• Peter points out that John Arrington found some problems with Eric's proton fit. The problem arises when W2>4GeV2. They hope to include some of Nadia Fomin's data to improve the fit, or maybe just generalize to make a smooth transition to NMC or other DIS fit.
2. Karl:
• Karl discusses his results for Delta u/u and Delta d/d.
• Needs to double check Delta d/d result.
• Discussion of sum rules paper: Karl has limited time due to PAC, but promises to try to finalize.
• Hall B is also working on a sum rules paper for EG1b, so time is tight.
• Peter points out that since we published g1p & g2p plots, it is fair game for anyone else to form the moments (and publish without us).
• Last major topic to finalize is how to properly include the DIS contribution.
• (Peter points out that he has 'machinery' for fitting world F1d data. May be useful to form world g1d fit.)
• Peter suggests that Karl use LSS06 for DIS contribution of moments. This fit is very new (Dec. 2006). Other possibilities would be AAC or Bianchi and Thomas (II).
• Peter suggest that Karl contact Stamenov for LSS function that directly returns g1.
• We should also form Bjorken sum rule. Peter points out that Bjorken is especially sensitive to DIS since the Delta contribution is minimized.
3. Mark:
• Received Arenhovel's calculation for quasi-free deuteron over wider range than previous table. This will change the radiative corrections somewhat.
• Mark will post new version on his website when completed.
• Short discussion of whether the calculation is reasonable at high x.
• Peter points out that it's possible to gain information on the deuteron wavefunction from high-x deuteron data (especially EG4 data). The counts are very limited in this region so fits/calculations should not really be trusted for now.
• Mark mentions that the deadline is approaching for Baryons 07 in Seoul Korea, June 11-15. Shige does not think he wants to go. Karl may submit an abstract. Deadline is in February? Peter mentions small workshop in Trento (15 people) at the same time as Baryon07. Mark mentions that the deadline for April APS in Jacksonville is quickly approaching.
• No meeting on Monday due to holiday.

12/11/06

1. Mark Oscar and Shige present
   
   • Karl not present. (Thanks to Shige for recap of the meeting.)
   • Oscar briefly talked about a recent seminar given at UVa by Igor Strakovsky who discussed the resonance at ~1355MeV (see Oscar's email on Dec 11).
   • Shige: talked about the new deuteron DFs which were posted on his webpage.
   • Mark discussed the RC. He will get new deuteron asym results using the new DFs.

11/27/06

1. Shige
   
   • Comparing RSS deuteron model to Peter's model.
   • Redid PF study: Data/model comparison repeated. There is not much change.
   • May ask Peter to repeat his fit using version of R that is consistent with the rest of our analysis. But, this will probably have very little effect.
   • Oscar: We should just use Peter's current model (without changing R). The model looked excellent compared to other data and looks reasonable with our data. The only major difference is R subroutine and this has really small effect.
   • Shige will proceed with DF calculations using Peter's model.
   • All : disussion of anomolous point in Deuteron A_perp at around W=1300 MeV. It looks quite strange in A2 (sign flip). Mark will make some plots from Frank's tables to check systematics. Mark will also look at raw asymmetries to see if the anomoly is present in the raw data, or is a result of one of the corrections.
2. Mark
   
   • Compared our deuteron data to EG1B deuteron and to the hall B models.
   • A2 models do much better than the A1 models.
   • Karl offers to send Mark the Hall B RC table.
   • Mark will begin systematic study of the model dependence.
   • The A1 model dependence is atleast the size of the stat error, or even worse, but the models are quite bad, so this may not be the best way to evaluate.
   • Can we compare Hall B proton + Hall A neutron with our deuteron? No, Hall A neutron is not yet available at our Q2.
   • Arenhovel sent Mark a new calculation of deuteron QE with larger grid in E,E', and W.
3. Frank
   
   • All : Discussion of the paper.

11/13/06

1. Frank:
   
   • Next week he will probably be ready to send new draft to the collaboration. Not ready yet. Maybe the monday after Thanksgiving.
2. Shige:
   
   • Tried Peter's model with RSS ressf.f
   • He is checking different versions of deuteron xs fits to see if it's worth to repeat the dilution factor/packing fraction calculation.
   • The biggest difference seems to be a low W (around the Delta), not much difference at large W.
   • Peter is using an old version (March) of Eric Christy's code.
   • Conclusion: If we use Peter's code (which agrees well with data) then we should use it with our version of Christy fit.
   • Shige also varied pion value in the Christy code: found that it had not much effect.
   • Shige believes that Peter's model is probably better, but we need to ask Peter to repeat with July 2005 Christy routine, which would be consistent with our analysis.
   • Whatever we do, we should use same model consistently throughout analysis.
3. Mark:
   
   • Working on Deuteron R.C.
   • Oscar requests the error matrix.
   • The current file has some incorrect leftovers (?).
   • Using 45 MeV bins.
   • Mark is treating A2 as a constant.
   • He sees an anomaly at x=0.65.
   • Starting to look at the systematics, he will create a short write-up when complete.
   • Will produce results in 15 MeV bins so others can rebin to whatever they like.
   • Mark has requested a larger spread (in energy) in the Q.E. calculation from Arenhovel.
   • Karl expresses some concern regarding the (small) size of the deuteron radiative corrections relative to what he found with the proton. Mark points out that the deuteron spectra has much less structure which would lead to smaller corrections. Karl notes that we could compare the new code to Eg1 deuteron radiative correction tables.

11/08/06

1. No meeting
   

11/01/06

1. All
   
   • Discussion of latest PRL draft.
   • Frank recieved comments from Oscar and Shige (local duality=Bloom-Gilman).
2. Shige
   
   • Regarding dilution factor: using the current model of deuteron in the M.C. instead of Peter's model results in the Packing Fraction increasing by 2-3%.
   • The comparison of data to MC is slightly better at lower W for Peter's model.
   • However, "our" model is better for perpendicular comparison.
   • Oscar compared the models to data in Donal's database from NE11.
   • His conclusion is that there is just no good model for Deuteron Q.E. All available codes have some problem. He has not yet tried Peter's code, but will try shortly.
   • Shige is also calculating g1d, g2d. Not sure which R to use. Proton R from Eric's code?
3. Mark
   
   • Working on Deuteron R.C.
   • New webpage where he will post all results.
   • Arenhovel's cross section in the Q.E. region is much smaller than either Hall C code or y-scaling result.
   • The comparison with asymmetry is reasonable though.
   • Mark shows A1, A2 with Q.E. tail subtracted. He finds a weird point in A2 around the Delta region.
   • Mark has started to fit A1 using the Oscar/Stein Breit-Wigner form.
   • We will meet two weeks from today. No meeting next week.

10/??/06

1. Oscar:
   
• New Dong paper not yet on Spires
• Simula fits and Bianchi and Thomas fits did not incorporate the E155 (and newer) data.
• They use only E143 and before, such as Q2=0 Bonn/Mainz sigma_LT sigma_TT data.
• This is why they are featureless in the resonance region. For this reason, Oscar is reluctant to use them.
• Oscar requests a comparison of E143 and E155 data to the fits.
• Karl should send his Bianchi Thomas fit to Oscar.
• The only existing g2 fit seems to be to Hermes data.
2. Shige:
   
• Working on Peter's routine for smearing: creating a table of Fermi momentums.
• Regarding comparison of data to the M.C. : Old comparison looked better (both for proton and Deuteron).
• Using same proton (Eric) model. What else changed? C, He and Nitrogen were adjusted in QFS in order to agree with hall B.
• Previously: we made modifications to QFS scaling factor. Now, Shige uses scale=1.0.
3. Mark:
   
• Working on Deuteron R.C.
• Slowly modifying the code.
• Will discuss with Arenhovel to get quasi-free deuteron calculation.
• Oscar want's to know if Mark will treat Q.E. tail similarly to how Karl treated proton elastic.
• Mark: can do both methods, i.e. as tail subtraction with associated f_rc +A_rc, and also simply as another inelastic channel.
• Oscar would like to have f_rc in order to properly handle statistical error.
4. Frank:
   
• Frank is working on PRL comments.

10/09/06

1. Frank:
   
• Received new comments from Shige on the PRL.
2. Oscar:
• Converted the Hermes A1 data to g1: Over a huge range in Q2 they still have really large error, and agree with RSS.
• Also found that Hermes is not covering the entire range of W.  The resonance coverage is 1.3<W<1.8 GeV.
• Each Q2 point only has local W coverage.  Need to average over all the Q2 points in order to cover the full resonance region.
• For these reasons Oscar is reluctant to devote much space in our prl discussing the Hermes data, as requested by the referee.
• Oscar then discussed the recent paper of Y.B. Dong(Nucl.Phys.A744:293-306,2004).  Apparently the paper compares g1&g2 parameterizations with pdfs, so it's not clear if it is so meaningful.  But the paper also claims that the Wandzura-Wilczek sum run does not need to be corrected for TMCs.  This directly impacts how we handle the DIS contribution to the Gamma_2 moment.
  
• Oscar then points out that Dong uses the nachtmann variable for his integrals, which had been strongly discouraged previously by the Hall C group.  It appears no one is really certain whether the nachtmann variable is appropriate or not.
  
• The paper claims onset of duality should be at about 1.7-2.0 GeV2.
• Oscar also briefly discusses a new paper by Fernando Stephens and Wally Melnitchouk.
3. General Discussion of PRL comments
• Oscar notes that it is not rigorous to use g1,g2 and x_bjorken at our kinematics.  We should really use G1,G2 and W.  May need to add a brief sentence explaining that we are examining transition from DIS to resonance.
• Frank is still not sure how to reduce the jargon in the first paragraph. 
  
• Frank raises the issue of whether the OPE is applicable at low Q2 if you just include more and more terms.  Not clear, we should talk to Wally.
  

09/30/06

1.   Mark:
   
• Does not think new changes to the PRL will be enough for the referees.
2. Shige:
• Working on smearing effect.
3. Peter:
   
• Has some useful codes for that.  Can provide a grid (200 bins) of probability of finding momentum pz, p^2 .
• Then for any given (nu,Q2) can calculate the distribution using minuit for a free proton and free neutron structure function.
  
• Comparison to the existing deuteron data provides neutron parameters since the proton parameters are fixed using Eric's code . 
  
• Peter breaks the entire region into 20 bins, and assumes each bin has 5% probability.
  
• Peter also modified Eric's RESMOD routine by hard-wiring all the variable parameters. This provides a factor of 10 improvement in speed.
• Notes that there is no way to properly smear proton and neutron near threhold which properly takes breakup into account.
• Peter is using Laget's code for breakup, but needs to somehow account for 2-body currents etc.
• Oscar: comments that Peter is building a "real" model of the deuteron.  We may only need a phenomenological fit.
4. Oscar:
• Using our free proton fit and "something" for the neutron, can iterate till we converge to the deuteron data. Some of the resulting neutron parameters may be insignificant, but this will be revealed by the fit errors.
• Peter believes it is incorrect to use arbitrary mass and widths for the resonances.  We could fix them using SAID, for example.  Notes that what we really are trying to determine is the Q2-dependence of the form factors.  The helicity amplitudes should not change with Q2. Inclusive scattering unfortunately does not give much information except to constrain the overall form.
• Oscar states that our main goal should be to perform a reasonable subtraction of the proton to get the neutron x-dependence.  This is important for d2n and moments.
  
• Peter notes that in EG1b they found that duality seems to work better for deuteron than proton.  Also, neutron does not appear to work as well as proton.
• Oscar et. al : Discussion of problem of extracting neutron in resonance region for Deuteron as well as He3.  Karl notes that Hall A never extracted resonance region neutron spin structure functions, only moments.
• Peter notes that the P+N final state in EG4 can really get neutron information.
• Oscar reiterates that he prefers to integrate a fit to the data, (instead of the data itself), for the moments.
• Peter can produce sigma_3/2 and sigma_1/2 in Laget's code which Mark could use for the radiative corrections.  The results will in effect cover the region 0<Q2<10 GeV2 and 0<W<3.2 GeV.   Peter also has table covering Q2<0.4 GeV2 from Arenhovel.  Mark has similar table at Q2=1.3 GeV2.
• Oscar notes that we should compare the old (hallC) deuteron model to Peter's new results.  This would tell us if its worthwhile to redo the packing fraction and dilution factor. Peter believes there will be a significant effect only below the delta resonance.  Mayber 10% elsewhere.
• Oscar would like to see how well the models reproduce our data (See June 14th yield plots on Shige's webpage).  Also, we need to update a few portions of our fit.
  
• Peter's model uses Donal's y-scaling in quasi-elastic.   Using a smeared R from Eric's fit, and assuming smeared neutron R = smeared proton R.
• Peter will send the code to Mark for the M.C.
  
• Oscar may be here for next weeks meeting.
5. Mark:
• Recieved some tables from Arenhovel.  Peter comments that Laget believes Arenhovel's code is more reliable near threshold, but this is probably not significant at RSS kinematics.
• Mark shows a comparison of Arenhovel's Apar and Aper to our data : There is 'rough' agreement which seems to indicate sizable contamination from the Delta on large W side of the Quasi-elastic and dip region.
• Oscar points out that Stein directly subtracted a Q.E. tail from the data for radiative corrections.  Karl mentions that Hall A treated Q.E. as just another inelastic resonance.
• Peter : our subtraction will be model-dependent, but the models will improve over the next few years.
• Mark will ask Arenhovel for wider W-coverage.
• Arenhovel is using the Walcher form factor which is better than Galster at our Q2.
• Peter uses a different form factor fit for the polarized and unpolarized calculations.  It's not obvious how to handle this properly.

09/18/06

1.   Mark:
   
• Talked with Peter RE new model he is working on for Eg4 to  fit all Deuteron data.
• Iona's model/data has problem at threshold (for W<1.3).
• We may be able to use this new model for P.F. etc.
• Oscar: Shige found quite high P.F. for deuteron compared to NH3.  Perhaps this will help there.
2. Karl:
• Shows moments plots.
• Oscar points out that not all the SLAC data is shown. They did not concentrate much on the 16 GeV and 9GeV data, mostly focused on the 27 GeV data instead.
  
• But the SLAC dilution factor may be bad.  They used F2 from NMC and R1990 which is not great at our W.
• Also, there were no good models of g2 at the time.
• We need to find out how well Eric's fit joins to DIS.
• Oscar will look at what a pure data DIS contribution to Gamma_1 would look like.
• Karl : shows Gamma_2. For DIS contribution, he tested what the Stratmann Bag-model would give, compared to g2WW, and Weigel's chiral soliton.
• Oscar points out that we probably need to apply TMCs to the models (Stratmann's g2ww atleast).
3. Shige:
• Now using R1998 (from Mark)
• Shows plot comparing R1990 and R1998.  There is about 15% difference, which looks like a "resonance".
  
• In fact, this has <1% effect on the smeared A1 and A2..
• Shige is not sure which kinematic variables to use for the neutron.  The Deuteron and Proton values differ by about 5% below W=1.25.
• Above W=1.25 they agree within 1%.
• Frank suggests he take the statistics weighted average.
4. Oscar:
• Working on DIS contribution to moments for next week.
  

09/11/06

1.   Shige:
   
• Shows plot of neutron smeared cross section, along with Apar & Aperp, and A1 & A2. 
  
• Using 30 MeV bins, does not see much structure.
• With 60 MeV bins has nice plot.  Both A1 and A2 are negative on average. 
  
• Probably won't have to unsmear since he is using 60 MeV bin.
2. Karl:
• Shows plots of the moments
• Mark points out Karl is using old Soffer-Teryaev curve.
• Oscar : it would be interesting to quantify the comparison of PDF's to the E143 DIS data.  Could use this
  to determine how much of DIS "leaks" into Resonance and vice versa.
3. Mark:
• Working on Deuteron RC.
4. Frank:
• Regarding the PRL referees comments : it will be difficult to satisfactorily cover both duality and g2 in the paper.
• Believes it is best to remove duality table from paper.  Oscar agrees.  Can use extra space to discuss that 50% of our d2 comes from twist-2.
• Oscar: we should really use G1,G2 instead of g1,g2.  We need to add discussion about Hermes A1 duality result.  A1 duality may be accidental cancelation in g1 and F1.  Perhaps can use this in the motivation for RSS section.
• After Frank rewrites, he will recirculate to collaboration.
• Oscar here on Weds-Thursday.
  

08/28/06

1.   Shige:
   
• Working on smeared neutron cross section = Deuteron XS - smeared proton.
• Using 5752.515 MeV for E0, and kinematics from the ntuple vars.
• Peter: Asks if spectator nucleon is on-shell? 
  
• Oscar: There is no off-shell correction applied.  They are using a convolution of W1 and W2 with momentum distribution.  Will also try some other prescription.  Peter can provide his code for a cross-check.
• Shige shows smeared neutron Apar and Aperp.  He uses DeltaSigma/(smeared neutron xs) and R1990. 
  
• Peter questions why we don't use HallC R?  Could smear F1,F2 and get R from that.
  
• There should not be much difference between R1990 and R1998 at our Q2, W.  Difference should be at low x.
• Peter: his code 'externals_all.f' has improved QE for all targets but deuteron.
• Mark notes that if our W-bins are big enough, we don't need to (should not)  unsmear the neuteron results.  Average smearing is about 100 MeV.
2. Karl:
   
• Shows plot of Gamma_1. 
  
• Now evaluating from the model at Q2=1.279
  
• Peter recommends to get revised version of Eg1b from Yelena. Also that we should look at replacing contribution at threshold with model, as they did in Eg1b.  This can have big effect on d2.
  
• Oscar points out that E143 did high x extrapolation which will be useful for us.
• Karl should look at gamma_0=intx^2 A1, using fit, not data.
• Propagate relative errors to fit result.
3. Mark:
• Arenhovel will respon in 2 weeks.  In meantime he will continue using simple model for Deuteron in RC.
• Peter has arenhovel's code, so does Vipuli.  there are two versions.
• Peter also has Laget's code.  He is allowed to run it, but not share it.
• Peter will run the codes for Mark.
4. Peter:
• For NE11 data, Laget looked better than Arenhovel, at Q2=1.7 unpolarized.
• Steve Rock created a parameterization of Arenhovel's tables.  It is in Peter's externals code: Option 3.
5. Oscar:
   
• The proofs for the elastic paper are back.  They look fine.
• We will not meet monday.
  

08/17/06

1.   Shige:
   
• Last week he found negative cross sections.  Now he is using new deuteron parameterization.
• Needs kinematic variable phi, but which phi to use for A_para , A_perp?
• Oscar : Phi-dependence only important for A_perp.
• Assuming R_n=R_p.  Getting R from Frank's table.
• Oscar: probably better to use R=0.2 or R1990 instead, because of all the wiggles in the Hall C R.
• Karl : will send Shige email regarding which ntuple variables to use.
• Eric: Don't trust Deuteron cross section below W=1.15.
• Oscar : Steve Rock studied many different models for Q.E. There was no report, but he produced a series of plots.

08/10/06

1.   Shige:
   
• Working on developing code and debugging.
• Procedure: Smears W1^p and W2^p and then creates smeared cross section.
• Shige needs unpolarized structure functions for the deuteron.
  
• Frank : warns that the structure function quantities in his deuteron files are actually proton quantities.
• Oscar : Shige can use the Niculescu/Stewart/Keppel parameterization.  The only assumption would be that R is independent of A.
2. Mark:
• Working on the RC for the deuteron
• Plot
• Needs radiated version of QFS.  Karl will send his version.
• Will get the Q.E. parameterization of asymmetry from Arenhovel.
3. Oscar:
• Arenovel is retired :(
• The tables Arenhovel produced for GEN only go to Q2=1.2 GeV^2 and E=3 GeV.
4. Frank:
• Submitted prl to archives : nucl/ex and hep/ph
• Oscar points out a 'new' treatment of Target Mass Corrections by Piccioni and Ritali (1998).

07/20/06

1.   Shige:
   
• Shows g1 and g2 D using 180 MeV bins.
• Needs dilution factor in Q.E. region.
• What model to use for the deuteron?
• Oscar has tables from Arenhovel at our kinematics.
• Shige finds that g2n is about 0.
2. Frank:
• Received comments from Ingo.
• Discussion of Jlab acknowledgement requirements.
• Will circulate new acknowledgement.
• Will leave off c^2 in all momentums.
3. Karl:
• Start work on sum rules paper: d2, GDH, BC, ELT, elastic to all
  
• Include the deuteron.
• E155x did ELT.
4. No meeting next week.
   

07/13/06

1.   Oscar:
   
• Should g2ww include x=1? He will follow up on this.
2. Frank
• Talked with EG1b people.  They request we average their Q2=1.2 and 1.4 GeV data.  They claim there is also more data coming.
• EG1a did not use a radiative dilution factor.  This is why the errors from Eg1b are worse.
• Discussion of PRL comments.
• Will drop data tables from appendix to prevent someone else publishing before us.

07/06/06

1.   Oscar:
   
• The g1D and g2D listed in official tables were incorrect.
• We were not using the correct unpolarized structure functions (either resonance or DIS).
• ~rondon/rss/analysis/deuteron
• Noticed this because raw Apar proton and Apar deuteron are about equal, but dilution factor introduces a factor of two.
• Concludes we should use R1990 (which agrees with Eric within error anyway).
• Overall about 1 sigma smaller g1.  g2 doesn't change much.
• Deuteron Asymmetry seems to be dominated by the proton contribution.
• Discussion folllowed regarding how to correctly apply DF.
• Elastic contribution to g1,g2: Oscar's result agrees with what Karl got.  Both disagree with Wally.
• Karl should get full DMT model.
• May have to double iterate on RC and smearing inversion.
• Spreadsheet:  updated with new flags to evaluate at data kinematics or at fixed Q2.
2. Shige:
• Shows his latest g1d,g2d.  Now better agreement with Hall A.
• using 30 MeV bins.
• g2n constent with zero and Hall A.
3. Frank:
• Re prl : d2 range inconsistent with plot data.  Not much impact. Will investigate further and fix.
• Eg1a/Eg1b data: Eg1b has worse statistics at our Q2 than Eg1a???
• Discussion of prl follows.
• Oscar discusses propagation of systematic errors for integrals.
  

06/28/06

1.   Karl:
   
• Shows 'final' pass VI results.
• Regarding A2 fit: Stops the fit iteration when the delta resonance drops to ~130 MeV.
• For A1 : Using Oscar fit, Hall B fit, Fit to world.
  
• For A2: Using All fits except Maid, and the Soffer Limits.
• Oscar requests tables have internal/external stages added.
• Oscar requests standard deviation of d2 integrand from all models used.
  
2. Oscar:
• Sees difference in elastic contribution to moments/g1,g2 from what Wally has in his paper.
• Requests Karl send the values he calculated at RSS Q2.
3. Shige:
• Slac g1d is divided by two (per nucleon value)
• Still not clear how to get A1n.
• Oscar: there is a paper by Tony Thomas and Wally M.  He will send it to Shige.
• There is a problem with using constant w_d when correcting for the D-state. But this led to only a small error in E155.
• Submitted his proceeding today.
4. Frank:
• Try for next week to get PRL ready, but probably not ready till week after.
5. Mark:
• Already received comments from PRC on elastic paper.
• Mostly minor issues.  Looks good.

06/22/06

1.   Karl:
   
• Discusses Pass VI RCs
• Suggests to try 16 parameter fit instead of 17.  Also, could stop fit iteration when the Delta width goes below some limit.
• Needs to talk to Sebastian or Keith regarding some of the Hall B A2 models (DMT and Soffer/W).
• General consensus : use OR, Hall B, and world fit for A1.  Use all A2 models except A2 = 0 and Soffer limits.
2. Shige:
• Asks for comments on his proceedings
• Sees big disagreement with his g1n and Hall A (X.Zheng data)
• Checked his code with 3 different fermi momentum distributions.  Finds negligible effect.
• Trying to figure out how to get A1n from A1p and A2d.
  

06/08/06

1.   Shige:
   
• Working on smearing the proton.
• Keeping relative errors constant through the process.
• Using 30 MeV bins now, may need to go wider.
• Shows g2n.
• What is binsize of Hall A duality?
• Oscar suggests looking at A1 A2 of deuteron and A1N, A2N.
• Eric: Peter found that Deuteron =(smeared proton + 0.6 smeared proton) is good prescription at low Q2. Describes the Jan 05 data well.
• Oscar : constant factor over entire x-range may not be suitable.
• ~tajima/rss/ssf/note.ps (Oscar requests comments)
  
2. Eric:
• R is still independent of nucleon # A for DIS.  Suggests we just use proton R.
• Systematic error on theory: suggests run several different potentials. Find out if it a big effect.  Oscar says this is addressed in his technote.
  
3. Mark:
• Has Shige been able to go back to unsmeared from smeared result?
4. Karl:
• Shows new RC results
• Needs to produce the error matrix for the fit.
• Needs to fit sigma_TT, sigma_LT, but for short paper we will use the A1,A2 fit, fixing the width of A2.
• Needs to talk to Vipuli to get latest data.
• Peter: There is low Q2 Aperp data from Bates and Nikhef.

06/01/06

1.   Shige:
   
• proton smearing: optimize run parameters getting V3 parameters from Frank.
• Gets similar result for 15 and 30 MeV Bins.
• Not sure how to propogate errors through the smearing?
• For now, keeping relative errors constant.
• Will create web page summarizing his work.
• Oscar: Smearing (nucleon fermi motion) is the perfect duality.
• Shows plot of smeared g1n, may need to go back to 30 MeV binning or larger.
• Karl and Oscar will send g1,g2 world data to Shige.
• Shows smeared g2
• May need to use a variable binsize.
2. Oscar:
• Using elastic contribution to improve duality.  Discussing with Thia and Rolf.
• TMC make pdf value for g1 at 0.8<x<1 non-zero.
  
• Different approaches to solve this problem can have huge impact on F1.  It's just not clear how to add elastic to the PDF.
• Discussion with Wally regarding elastic.

05/24/06

1.   Shige:
   
• Working on smearing proton.
• Trying to unsmear smeared proton.
• Got fit program, needs Phi from Frank (kine weighted vars are in "kineperbin" files).
• Oscar: suggest to try with finer binning.
2. Oscar:
• Discussion with Cole Smith : there are 19 3/4 resonances. adjusts multipoles wrt/ Liang's code.
  
• 1511 MeV : Eta production threshold.
• Wally, Thia and Rolf paper.
• MEK_DUAL.EPS in Oscar's directory: rondon/rss/analysis/systematics/

05/19/06

1.   Shige:
   
• Working on DeltaSigma smearing
• D-wave function from Oscar, who wrote a technote.
• Following technote, debugging code.
• Shows plot of smeared proton deltaSigma using Paris potential.  Will run with other two potentials.
• Using fit function of Oscar for A1 A2 to get g1,g2.  With Eric's fit can get Delta sigma.
• Results look reasonable.
• Using constant Q2, wants to use actual Q2.
• Next subtract from Deuteron data.
• Mark: suggests to try to reproduce smeared quantity w/ fit.  Try to go from smeared to unsmeared. Asks Karl to give code to Shige.
• Oscar: can try using either 3 or 4 peaks.
• Shige is also working on his DIS proceedings, due by end of June.
  
2. Mark:
• Put off the elastic paper for now.
• Only got comments from Ingo.
• J.P. Chen suggested adding stat. and systematic  in quadrature.
• Generally, we prefer keeping them separate, but Mark has mixed some of them in the plot for the other experiments.
3. Oscar:
• 4 new students on NMR.
• AO can give asymmetries >1
• Cole Smith did not like us fitting A1, A2 instead of cross sections.
• Oscar pointed out the E143 paper which shows A1~sigma_TT' , A2~sigma_LT. So they are directly related to the multipoles.
• Because of this, his objections to our fit have been withdrawn.
• ~rondon/rss/analysis/rss-eta-a1.eps, rss-eta-a2.eps
• Cole Smith ran Maid at our kinematics, got same results as Karl.
• We could do a multipole decomposition. 
  

05/11/06

1. Oscar :
   
• Wants to finish by June deadline.  Discussed with Donal and Don, July probably ok, and will free Shige and Karl to work on other projects (Frost)
• Discussion of Duality.
• We will invert our duality integrals to match the world convention.
• Created a Makefile for Radiative corrections.  Can be used for debugging: /group/rondon/rc/subs/src
2. Karl
• EG4 almost over.  Working majority of time on RSS again.
• Updated minutes (finally)
• RC started again.
3. Shige:
• Discusses OVC can for SANE runs.
• Had discussion with Simonetta in Japan: large x resummation should be included in duality integrals.
  

03/23/06

1. Karl
• Working on RC pass 6.
• Oscar suggests using the envelope of statistical errors to see systematic effect.
• could use minuit.
• Needs to recheck his Radiation lengths using the final Packing fractions.
2. Frank:
• stat weighted average is much different than before, now gets much better agreement with Mark.  Frank weighs Q2,W with the stat average of asymmetry.
• Karl will do RC at central kinematics, then Frank can redo the stat weighting.
3. Oscar:
• g1 has not changed significantly since first try.
• Now Eric has new model, with big difference at Delta.
• P.F. is very insensitive to the model, but DF will vary.
• Frank will get commitment from Eric.
• Oscar: single pi production near threshold might be useful for threshold.
4. Shige:
• Method of combining df for high/low weighted average is not equal to combining all into one bin and then calculating error.
• Solution: combine high and low ntuples (weighted by charge for each case) then dont need to combine.
• Proton smearing program : comparing several non smeared xs should have something next week.

03/16/06

1. Shige:
• 20 min talk at Japan, Vipuli also.
• Proton smearing calculation: needs some kinematic variables.
  
2. Mark:
• Questions the statistical propogation of errors.
3. Oscar:
• two new notes on Radiative corrections.  The second is an example of how not to do it.  Can't explain why new f_rc <> (old frc)^2 but otherwise very clear.
• Mark: model systematics comes into DeltaRC only, otherwise in ratio.
• Ionization is already taken care of in the engine.
• Straggling is not: Most urge to ignore straggling since it is a fraction of 2 MeV.
4. Eric :
• has newest version of his fit. Changes:
• Q2 dep of FF
• no extrapolation to real photon point
• Best BR for doing relativistic B-W and threshold effects.
• Should be significantly better at threshold
• pi, 2pi, eta
• We'll compare to previous version to see if we need to implement it.
• renewed request for error on Eric's fit.  It's possible, but he needs time.
  
5. Peter:
• W<1180 MeV we have no measurements, error is so big. Similar to RSS results.
• look at polarazibility gamma_0
• Vipuli working on this problem for hall C, can discuss with her, she has 12 models of g1.
• Peter has smearing  routine for radiative corrections. works pretty good.  Eric has code from Wally which is very sophisticated.
• Peter will provide his code to Shige, but it may also be useful for RC.
• Looking at N15 elastic, Karl will send some info he gathered for PAC.
  
6. Marc:
• Has new version of elastic PRC
• got comments from Peter
• will add more comments on 2gamma effect.
• will add formulae for cross section.

03/09/06

1. Oscar:
• Updated systematic errors
2. Karl
• Looking for error in RC version 5
3. Oscar
• f_rc : many different definitions. E155 covariance of fit from models which is very different from our deficintion.  Should only be used in statistical error propagation, and should reflect the net change in counts in a given bin from radiative corrections.
• f_rc ~ sigma_born/sigma_T
4. Frank
• will send out draft of PRL with appendix.
  
• Karl should send the fit's diagonal matrix to Frank.
• Frank doesn't want to circulate yet.
  

03/02/06

1. Frank:
• Discussion regarding PRL.  Final comments before sending to collab.
2. Karl:
• Discussion regarding RC systematics.  Will send table to Oscar. Also needs to evaluate the statistics weighted average for PRL table.
  
• Looking at lowest W-bin data.  Trying to salvage so we don't have to discard this large error point.
3. Shige:
• Not present (working in Hall A), but posted new results for Deuteron DF on his webpage.
• Oscar points out that this DF is consistent with maximum expected (6/21 for N15D3).
• Frank will generate new deuteron asymmetries.
• Perp-Top is still TBA.  Mark suggests it is not high priority since recovering this data will only decrease error bars by 8%.  Would prefer that Shige work on the neutron extraction.
  
4. Mark:
• Would like to bundle up all useful background info (tables, fit parameters, etc.) and put in an appendix for Los Alamos archive version of PRL.
5. Oscar:
• Discusses d2 value. Looks like d2 doesnt depend much on Q2.  Still many sigma from zero.
  
• Mark brings up "factor of two" issue.  He wrote a note about it previously, requests that we review to ensure general agreement.
• Discusses correlations of systematic errors.  Refers to "excellent treatment"  of said issue in T.J. Liu's thesis (E143).  Bottomline, no big change in his estimate of sytematic error.
• Mentions that Eric has requested we reference the Hall C elastic Rosenbluth paper (in our elastic paper?).

02/23/06

1. Karl:
• Discusses pass V RC results.
• Not clear how to establish systematic uncertainty.  Right now looking at variation due to choice of model input (Oscar's model, Maid, and Hall B).
  
• Oscar suggests to discard Maid, atleast above W=1400 MeV where we know it is unreasonable.  For W>1400 MeV, use the difference between Oscar model and Hall B. Below 1400 MeV can use the variance of the three.  For overall number divide by the additive RC correction and then average.  Believes the elastic uncertainty is small (~1%).
• Notes that g1,g2 are significantly smaller than what he found in Pass IV.  Oscar points out that this is expected since the DFs are now bigger than previous pass.
  
2. Oscar:
• Working on systematic evaluation for PRL.
• Finds a large systematic for A1, A2 (10% and 30%).  The kinematic factors seem to be unfavorable, especially d'.
• Regarding the large x resummation: Simonetta provided estimate for the global integral ratio.  The ratio should be increased by factor 1.214.
• We will mention in text, but leave duality table unchanged.
• Compared Karl's pass V tables with his own and Frank's.  Found agreement to about 0.2% for almost all bins.  The exception is in bins where the Q2 varies dramatically due to the momentum overlap.
• Frank : points out that the peak positions in the fit graph do not match the values the fit code produces.  There is a small shift.  Oscar has noticed this before but will reexamine.
• Oscar did not take correlations of uncertainties into account.  May reduce systematic significantly.
• Points out that fit is basically zero for first three W points in d2, while the uncertainty is quite large.  We may be taking an unneccessary hit in overall systematic.
• Karl concerned that fit does not follow the points very well in this region, will look into it.
  
3. Frank:
• Discusses PRL
• Working on space constraints.
  

02/16/06

1. Frank:
• Regarding rules of affiliation, PRL wants only who paid you to do work listed.
• Will update his script to use latest Nitrogen asymmetry files.
• Working on RC paragraph, other comments for paper.
2. Mark:
• Regarding nitrogen assymetry correction: between version 5 and 6 the proton xs might have changed.
• Believes first point of asymmetries should be discarded.
3. Karl:
• Released preliminary results for RC pass 5.  Hopes to finish within week.
•  Oscar and Frank will proceed with these preliminary results, update if there is a big change.
4. Oscar:
• Had discussion with Simonetta Liuti on usage of Nachtman variable. She is a leading investigator of duality and it's still not resolved if Nachtman is appropriate.  Her opinion is that in the integral it is only a dummy variable. Believes it is proper for us to state that we find better agreement when using Nachtman in the paper.  Oscar's integral results do not indicate the Nachtman behaves simply as a dummy variable.
  
• Large x resummation: She would need 2 weeks to recode her procedure to directly correct g1.  Instead she will give Oscar a parameterization in the next few days.
• We'll reference S. Liuti, Bianchi et al in the paper and thank her for discussions.
• In the future, it will be crucial to reduce the PF uncertainty (0.37*1.77*5% in his note).  Could achieve this with 6mm disk targets instead of pellets.
• Wants to move on Mark's PRL quickly.
• Will be here next week for GEN.
  
5. Shige:
• DF systematics: 4.8-4.9%. 
  

02/9/06

1. Shige:
• Regarding the systematics due to the nuclear model: Finds 2.4% for high P, 3.3% for low P.
• Now gets a total systematic for D.F. of 5.3%., detailed on his webpage.
• He will write a formal document to summarize.
• There seems to be general agreement that 5.3% is consevative.
• Oscar: mentions that Shige will need to redo the Deuterium D.F, since QFS has changed.
2. Frank:
• Working on applying the latest D.F. to the raw asymmetries. 
  
• ifarm machines very slow.
• Mostly complete, should be done in next day or so.
3. Oscar:
• Discusses paper
• Open question: Implementation of the large x resummation (correction to the TMC).
• It becomes most important at large x.
• Necessary for unpolarized structure functions.  Without it, global duality doesn't work.
• Details in paper by S. Liuti (UVA) et al.
• The issue is that the correction should be applied directly to the PDFs.  We can't do that since the code Oscar uses outputs g1 directly.
• Oscar will ask Simonetta to provide a parameterization that can be applied directly to g1.
• This should be the last open issue regarding our duality conclusions.
• Regarding TMC: Using Peter's code which agrees with published paper results.  It's not clear what effect the TMC have on our twist-3/d2 conclusions, since some of g2bar is from TMC.  Would be a problem if TMC differed for g1 and g2.  Refers to X. Sung paper :"Modified g2ww formulae with TMC"
• Still not sure if we should use/mention the nachtman variable.  Both the Bloom-Gilman and DGLAP papers mention that they find improved duality when an "improved" variable like nachtman is utilized.  We many want to mention this in the paper, since we also find improvement when using xi, but don't change the results otherwise.
• Mark: Prefers we leave it out, atleast till we speak with Wally.
4. Frank:
• Discusses several questions regarding the PRL.
• Points out that there are two points with very similar x in our plots, due to the overlap of the high and low momentum setting.  Would like to average them.
• Oscar mentions that the PDF errors are not resolved yet.

02/2/06

1. Oscar:
• Finds good agreement with preliminary hall C and our own carbon data (field off) by tweaking DIS part of QFS.
• Scale factor of 0.92 applied to DIS part of QFS.
• Continuosly adjust QFS, no discontinuity now between low and high momentum.
• ~/rss/analysis/qfs_dis/c_no_b_mc_data_opt_dis.eps
• We (i.e. Shige) will need to run MC one more time w/standard P.F.
• Should be last modification to proton Apar Aperp.
2. Shige
• Has already recalculated the P.F. There was no big change.
• Also did DFs : ~1% change  at low W, 3% change at high W.
• will need to double check everything, but looks final.
• Will finalize systematics for df for next week.
3. Karl
• Now must do next pass of RC
• Frank will run his code to apply DF to raw asymms
• Karl will send latest fit parameters to group when he has them.
• Needs to evaluate RC systematics on this pass.
4. Frank
• PRL : Latest version only differs by space cutting.
• Mark : Change Pb to 2.7% (relative) 1.8% (absolute).  This increase from the previous values arises from removing some bad Moller runs from consideration.
• Oscar: May need to add TMC explanation.
5. Oscar
• RE moments paper : has Prof. Sungs TMC, they agree exactly with Linda Stuarts code.
• He will be onsite next week.
6. Mark
• Received comments on elastic paper from Shige and Oscar.  Will incorporate those and send out a new version.
  

01/25/06

1. Oscar:
• Has new spradsheet with duality integrals (see latest email).
  
• Also now has Peter's TMC as a subroutine.
• Mark: Stat PDF disagrees w/ others at low x, He would expect better agreement there.
• Has new DF note
• Question: Use scaled(to our C data) QFS or match to preliminary Hall C dat for d.f.?
• Difference is part of our systematic, but we need to choose on of them for central value.
• Mark: would like that we use a continuous function of W, instead of two discrete scaling factors for high and low momentum.
  
2. Shige:
• Systematic due to scale factor.  Using 0.96 for low P and 1.04 for high P.
• When he looks at top para and top perp, he finds a +-3% change, which leads to a 1% systematic.
• Mark: suggest take error on xs and propogate to d.f.  Can be done by hand (no code needed)
• Oscar: refers to page 7 of his TN.  Estimates 2-3% from this method , 4% max
• Shige will write up a one page summary of the procedure he uses.

01/19/06

1. Oscar:
• We're in good shape for the packing fraction.
• Discusses latest version of PDFs from AAC.  All curves have target mass corrections (TMC). 
  
• Mark noticed that in fact the TMC are quite big.
• Xiaodong Jiang pointed out that there are errors generated for Delta u, Delta d.
  
• However, Oscar says they are very difficult to propogate.  Naive formula for g1 doesn't work.
• Mark : suggests we ask authors to generate g1 at our Q2 to avoid mistakes.
• Some of Oscar's results at ~rondon/rss/v3/fit/dual
• We will need to make clear which version of TMC we are using in the publication.
• Only stat model from Soffer doesn't have TMC.  Oscar will try to include them.
• All our previous conclusions on duality still hold.
• RE Thia's objection to our integrating over single resonance.  This was apparently a misunderstanding.  Oscar is following the standard procedure from the earlier hall C paper and Bloom Gilman paper to choose our range of integration.  After speaking with Eric, Oscar says there is no issue.  The confusion arose because Thia/Eric average over Q2 before performing the integrals.  We simply can not do that with our data.
• Also, increasing range of integrals beyond nominal width of each resonance does not improve the comparison with the PDFs but makes it worse.
• Oscar gets similar results if he uses the measured A1, A2 or A1, A2 evolved to constant Q2 using his fit. 
  
• Karl points out that latest fit has almost no Q2 variation.
• Oscar: possible to use g1 or |g1| in the delta region, although it seems unnatural to use |g1|, and breaks global duality as well.
• We will need to perform final pass. But modifying g1 by 5% has only 5% effect on the duality integrals.
• New Technote on DF/PF: looked at individual components of QFS compared to the Hall A N14 data.
• Concludes 2-3% uncertainty on model at low W, larger at higher W.
• Next step: Shige will run the MC with SCALEIN <> 1.0 for systematic study.
  
• Frank : discussion of removal of PF paragraphy from PRL.
• Oscar : d2 = 0.106 from 3g2bar. 0.074 from 2x(g1+3g2). 0.065 at constant Q2.  Will finalize after DF are completed.
• RE factor of two in lattice version of d2.  We use E155 definition, lattice should be divided by two.
  
• Effect of Q2 dependence is small.
2. Shige:
• Working on proton smearing.  Should he use central angles theta,phi or modify somehow?
• Mark has some kumacs that might provide the (statistically weighted?) correct variables.
• Frank offers to run through the ntuples to get the proper phi for Shige.
• Also working hard on GEN in hall A.
• May need to help hall B gdh experiment. 
  
• Oscar: cryo group and Italians should have sufficient expertiese and manpower to handle it.
  

12/15/05

1. Shige:
• The Packing Fraction varies very little when Peter's model is substituted for QFS.
• QFS agrees much better w/ our Carbon data, the new model.
• PB model not so good in QE and below.
• Oscar : PF seems very insensitive to whatever choice we make.
  
2. Oscar:
   
• PB model not a good match below W=1.4 GEV, but QFS has a llittle bit toowmuch dip in region just above Delta.
• Did not look at individual channels of QFS yet.
• Fairly convinced now that the RSS packing fraction/dilution factor method is correct.
• Yield is dominated by Nitrogin. 
  
• To get the PF we need to use yields, not cross sections.
• Doesn't like using two differenct PF for high and low momentum, but it might be consequence of changing magnetic field.
• Mark points out that it would be bigger effect for perp, not para, which is opposite what we find.
• Wrote new technote on packing fraction and dilution factor.
• In Eq. 5 we were previoulsy scaling the N15 xs to match our carbon data, but not scaling hydrogen xs. Now we will scale neither. The dilution is very sensitive to thes ratios, but the PF is not.
  
• Peter model undershoots Hall C data below W=1.4.
• Shows plot of Asymmetries, only corrected for PbPt.  Already features visible at 1350 MeV.
  
3. Mark:
• To get MC elastic to come out right you'd need to replace 20% of NH3 with Nitrogen.  This whouldn't have a big effect anywhere but the elastic..
• Will double check the RC in MC , compare directly to SimC.
• Oscar: The packing fraction and DF should ultimately only depend on the model chosen.
4. Karl:
• Pass IV RC complete.  These results include correction for tilt in scattering plane and use statistic weighted kinematics.
• Latest fit to A1, A2 and stt, slt.  Notes that slt does not seem to go through zero at delta as he expected since M1 multipole should dominate.  Mark points out that as Q2 increases, it is expected that the other multipoles should be more important than M1.  Karl shows He3 slt from hall A,which is dominated by QE contribution. Will provide zoomed in plot of delta resonance.
  
• Submitted proceedings for Spin Praha 2005.
• Shows comparison of Peter's model and Hall A Nitrogen, He3 data.
• Oscar: notes that for Q2~0.5, the RSS version of QFS should do very good job.  Chris Harris worked hard to get it to agree well with data.
5. All:
• Oscar suggests adjusting our MC to latest Hall C data.  Mark points out it will be small (4%?) adjustment, and the data is far from final.  Might be better to just adjust our systematic error.  Conclusion: Models won't be tweaked anymore.
• Mark suggests using averge of low/high PF, and the spread as our systematic.  Difficult to explain why there would be two different PFs otherwise.
• Oscar: Shige will need to run MC with average PF, and the two extremes, then get DF with average and +_ the spread.  Would also be valuable to run Peter's Model at the same time for one additional systematic estimate. 
  
• Oscar: Frank will then need to regenerate the asymmetries, and Karl the RCs.
• Summary by Mark:
1.  Take the average of the packing fractions for the "HI" and "LO" settings using the unscaled RSS QFS model.
2.  Using the average packing fraction calculate the dilution factor using the MC which use the unscaled RSS QFS model. We use
    the MC to getting the radiated cross sections to use in the dilution factor.
3.  Use the new dilution factor to update the Aperp and Apara using Frank's scripts.
4.  Update the radiative corrections.
   
• No meeting next thursday.
  

12/08/05

1. Mark:
• Note from Dave Mack on BCM calibration. Systematic is about 5%.
• Born XS from RSS model is about 20% larger than Peter's new model. (RSS model includes scale factor of 1.10 to match our carbon data.)
• RCs: Ran the MC with and without Radiative Corrections to determine size, then compared with Peter.  They agree well except in the Delta, where they disagree by about 15% for hydrogen.
2. Shige:
• Uses two different PFs for high and low momentum settings, since one PF does not describe data over whole range. Oscar notes that Hongu also observed that a long time ago.
3. Oscar:
• Peter's and our model differ by 20% (except for hydrogen. We both use Eric Christy's model for hydrogen, although Peter is using an older version.)
• Neither agree with the Hall C data, which is in between.
• How can we count more than we expect? Two unknowns:
1. Packing Fraction: If we knew it exactly, we could write the xs as a ratio of nuclear to proton cross sections.
2. Solid angle. (The low current is well known).
• Points out that a normalization due to acceptance (1.10) is not equal to the error on the acceptance, which should be much smaller.  The normalization has error which is not equal to 100%.
• Alternative: Look at difference of models spread around data.  Make normalization, and take spread as the error.
• Another alternative: Go directly to XS measurement.  Get DeltaOmega from other hall C data.
• Calculate PF with well understood XS.  From MC/data we would know how to adjust.
• Frank : maybe effect is rate dependent.
• Oscar is concerned that we are so sensitive to xs-type quantities.  Never was an issue for the SLAC asymmetry measurements.
• 10% systematic on PF ==> 10% total systematic.  It will dominate everything else.
4. Mark:
• XS models compared to 12C give correction factor (CF).  CF * NH3 ---> yield.  Yield gives PF.  So the PF comes strictly from radiated models.  Use Peter's code.
  
• Systematic uncertainty?  3% on H model. 5-10% on A>3 model.
• Conclusion: use known xs to normalize our yields.
• Already put Peter's model in the MC.
• Has Peter's RC code, will use to radiate.
• Will update the CVS.
  
5. Oscar:
• The yield is a linear function of Packing fraction, which is why it is fine (actually best) to use linear interpolation to obtain the value between 50 and 60%.
• Suggests we try to turn off field for parallel setting in MC and engine.  Take a look as shifts in data and M.C. Data will reveal size of real shift.
• Frank is skeptical, thinks best way would have been to take data with no field.
• Peter's new model will also affect the EG1 data.  Their g1 will come up, ours will come down.
• Possible solution in future: Solid NH3, ND3 disks sliced from frozen cylinder.   LiD even more machinable.  Only problem is placement of ND3 coils.
6. Frank:
• Requests Karl send curve for A1 and parameters from latest fit to our data.
• Oscar will send g2WW to Frank.
• Questions "wiggly" PDF (Soffer's?)  Oscar: wiggles come from Q2 evolution to CLAS Q2 of 1.2 and 1.4 GeV2.  Not present in original.
• Requests Karl send SLAC data.

12/01/05

1.   Peter:
• DF from EG1b and RSS differ y 10%. (in the right direction)
• difference seems to be from radiated xs.
• Other possibilities : different xs fit (born) or SIMC bug for RC.
• Mark will compare unradiated  xs for all targets.
• Peter will update his proton fit to Eric's latest.
• His new model agrees well with most of the nuclear target data from last fall in Hall C.
• D model smeared with some fermi and EMC corrections of about 10%.
• EG1 evalutated DF in two ways: 1) models 2) directly from the counts from Carbon, corrected for presence of He (Ylena).  Peter did both and they agreed well, except near resonances.
• RCs depend mainly on W and a little on Q2, sees same size RC for 4 GeV & 5.7 GeV data at same Q2.
  
2. Mark:
• top Apar, comparing DF's.
• His and Shige's DFs agree pretty well.
• We never got 5% xs, but instead absorbed the accceptance effects, etc into the model.
• Peter points out that this could be dangersous if you only apply that to Carbon model, but not to Hydrogen model.  Also, there is newer/better data we can compare to.
3. Shige:
• Changing the xs by 10% changes the DF by 6.3%, bothh para and perp.
  
• looked at DF systematics: About 1%.
• Oscar: suggests running the MC with and without the RC, make a table and send to Peter.
• Liang vs 7/22: changes DF by 0.9% para, 0.8% perp.Peter points out that the two models are completely correclated.  Suggest Using Linda stuarts H2 model or Atwood/Bodek fit instead for systematic check.
• Oscar: Overal 2-3% uncerainty in proton data, bigger for N. Liang Thesis syst~2-3%.  Would need 20% error on xs to get 10% effect on DF.
• Mark: need to look at N.  Shige will modify scale factor in N.
• Peter: 52% PF seems quite low. Oscar rrecalls scrounging for Ammonia during the ru just to fill the cell, so  its not so surprising.  Target material came from 4-5 different sources(Dustin Mcnulty).  Typical in the pas was 55-58% for this target.
• Peter: Eg1 ~70% but it's actually PF*length.
• Shige: summarizes PF systematics.  Total ~1%.
4. Oscar:
• Last year he compared our N xs to Ylena's parameterization of N.  Looked very good to W~1.6 GeV.  But, was never able to find out what parameterization of D was used by Ylena.  Will forward those plots to the group.
• We had a pure Carbon target, only aluminum would come from tailpiece.
• The density was  based on micrometer measurments.
• Mark points out that it would have been useful to take data with field on in DIS.
5. Karl:
• Shows some RC stuff.
6. Oscar
• g1 plot compared to E143, good match of our fit to DIS fit.
• d2 = 2g1+3g2 shows a big difference from Q2 to const Q2.  Oscar evaluates d2(Q2=1.3) from fit, the error is the relative error at the actual measured Q2.
• choice of DIS extrapolation not so important for d2.
• In paper, integrate up to pion threshold.
• For Karl: Use 16 GeV 17degree data for dis data (E143).
• Sane meeting on thursday.
  

11/17/05

1. Shige :
• Working on proton smearing code.
• Running Oscar's code, and working on smearing technique. 
  
• Needs  (W,Q2) value.  Should he use average Q2 or measured?
• Oscar: use measured stat-weighted Q2 from Frank's files.
• Proceed with uncorrected (RC) deuteron asymmetries to get smeared A1N, A2N, then unsmear.
• Paris/Bonn potentials should be accurate enough at our momentum
• Oscar: may need 45-60 MeV bins.
2. Karl:
• Sees problem with NC at low W.  After discussion with group, it is clear that it must be combination of elastic tail subtraction + NC.
• Oscar: suggests set A1 to zero at threshold: whatever is not delta is non-resonant BG.  (The cross section in mostly non-resonant).
• Mark : Maybe can massage the NC at threshold to get value you want.  Error is so big, have some flexibility to adjust central value.
  
• Will have self-consistent RC over the weekend.
3. Oscar:
• Global duality must hold if wish to do extrapolation (to high x), but it is not interesting or unexpected.
• Using fit at constant Q2=1.3, compared to Peter's pdfs.
• Target Mass corrections: at large Q2, Mp ~0.  Callan-Gross relation holds.  At lower Q2, corrections needed to account for mass of proton.
• Correction is applied to g1p.  Peter is using Blumlein approach.
• Oscar uses the  resonance widths (from his fit) for the local duality integral limits.
• Sees glogal duality to about ~5%.  Local to about 30%. Ires/Idis ~0.7
• The table in PRL should be per resonance, not per pdf, using some average of the pdfs perhaps.
• Have to document model in prl, because duality study must be done at const Q2.
• Frank: wants to detail fit in this prl, not the next.
• Oscar: Fit uses Q2 dependence of Stein et al., so its already well documented.
• Oscar: can add discussion of 2nd derivatives resonance test in prl.
• Oscar: A1, A2 peaks don't have to have same centroids (but they do!)
• Systematic study of R-dependence: Looked at Liang, 07/22, 06/14 and DIS fit.  The effect on 1350 peak is 3%. Two of tthese fits are clearly not applicable, yet the spread is still only 3%
4. Frank:
1. Oscar: PRL in good shape now.  Just some tweaking, don't reference anything not published yet.
2. Hall B has paper on global g1 analysis, and the model: Osipenko PRD 71, 054007 (2005).
3. Lengthy discussion of paper. 
   
4. Oscar has brilliant idea: Frank should calculate the systematics in his tcl script.
5. Systematic error on PF/DF: Shige can modify xs by +-10%.
6. Requests Karl send reference for RCs.
7. Requests Karl send latest d2.
8. Cant use simple g2WW formulae, general expression.
9. Requests Karl send appropriate E94010 reference.

11/10/05

1. Peter:
• Discusses comparison of EG1 Apar/D vs. RSS.
• Apar/D = A1+eta A2
• He finds difference of 0.05% compared to Oscar's file.
• Requests Karl put Apar/D or A1+etaA2, D in next table.
• Peter has a new improved XS model (quasiy8), and will use it to evaluate his own radiative dilution factor.
• Has an analytic model for rates in 3He, D, carbon and aluminum in hall C.  Works well compared to data.
• Still needs to get Eric's latests fit.
• EG1 measured C+MT targ bin by bin and directly subtracted it from data.
• Oscar: Shige performed similar study showing Helium from target and Helium.
• Peter: EG1 does not get zero in super elastic region after subtracting dilution factor.  Shige/Mark do.
• Peter: Averaging 3 EG1 Q2 bins.
• Oscar: Karl and Oscar results agree to 0.1 %.
• Peter: Comparing A1+etaA2 from two experiments, at threshold eta-> 1, A2 dominates.  The two experiments have almost identical eta.
• Requests that Oscar adds error bars.
• Eg1a did NOT use an frc.  frc=1 so they have quite small error bars at threshold.
2. Karl:
• Oscar has made fitting code public so Karl can now use it for iteration.
• Looking at RC systematic.  Initial estimate is difference between radiated model and data.  General agreement that this is too crude.  EG1(K. Griffeon) looked at 14 or so models and used the envelop as the systematic error.
• Frank: suggests running RC on DIS model and comparing to our data, (avoids features).
• Peter: suggests we vary the input model.
• Karl: requests Oscar send TN from Linda Stuart regarding target mass corrections.  Oscar will send to group, but emphasizes that Todd Averetts thesis should be more detailed and reliable.
• Oscar : switching to nactman variable NOT EQUIVALENT to applying target mass corrections.
3. Shige:
   
• Working on subroutines for smearing proton.
• Recieved necessary files from Oscar, needs to still get Oscar's deuteron model.
• Oscar: problem is impossible to solve analytically, will need to perform numerical integration, suggests Runge-Cutta or similar.
  
• Mark: Many routines available from Numerical Recipes.
• Oscar: was able to recover his old VAX files at UVA with calculate Paris, Bonn, etc. momentum distributions.
• Asks if there will be target cooldown soon.  Oscar: probably not till Jan.
• Oscar: would like Shige to come to UVA to learn how to analyze target data.
  
4. Mark:
• Working on Elastic paper for next week.
• Not clear what to use for Pt polarization systematic.
• Standard is 2.5% quoted in NIM paper,but what should we use?
• Oscar: the dispersion of good T.E.'s is about 2.4%, so the number used in E143  is not too far off.
• Oscar: bottom target: only two points so its hard to estimate, also there were problems with those T.E.s
• Oscar: However there were lots of deuteron T.E.s.  Suggests that we stick to 2.5% or 3% to be superconservative.
• Mark will work on paragraph to explain all this.
5. Oscar:
• Regarding the 1350 resonance, there was a suggestion from John Ralston to look at the 2nd derivative in order to locate resonances.  He sees strong indication from this technique that there is something at 1350.
• Fitting code is based on Bevington.  The resulting variance for all parameters is very small compared to central values, so the fit is *very* good.
• Conclusion: We definitely have 4 peaks at well defined positions.
• To Do: looking at A1,A2 w/ Whitlow fit for R to see how big our R dependence is.
• To Do: Duality integrals with Marks pdfs.
• Suggests that we have separate paper for g2/twist-3 stuff. Mark can just lay the groundwork for that in this prl.
• Oscar may be here next week for meeting.
  

11/03/05

1. Oscar:
• Discusses d2.  Understands why the different formulations of d2 integrand differ.  Most important is the total integral  which must give same result regardless of formulation.
• Also, it seems that we must apply target mass corrections as detailed in Linda Stuart 's technote (E143?) and in Todd Averett's thesis.  These are important because the gamma^2 (=nu^2/Q2) factor is not very small in resonance region and can not be neglected.
  
• Mark and he are discussing this topic with a theorist (Christian) to further understand d2 in the resonance region.  Initial thoughts: d2 in resonance region probes nucleon interactions instead quark-gluon interactions.
• Regarding "bump" at 1350 in A1: Oscar varied the corresponding points in Apar and Aperp by 3.5 sigma and observed that the peak in A1 only decreases to 80%.
2. Mark:
• Peter's comparison of (A1 + eta A2) between EG1b and RSS implies that our A1 should be about 20% less, since our A2 is 20% bigger than the hall b model. This follows from eta being about one at our kinematics.
3. General
• Discussion of systematics: PbPt comes entirely from elastic.  RC: correction is small, about 3-4% purely additive.
• Frank: inquires about Oscar's fit.  Oscar says it should only be run with the 1,4,4 or 2,4,4 options for resonances.
• Mark: Regarding Eric's fit, it is very difficult to get the error matrix since Eric fits FL, F1 separately.
• Discussion: about how great it would be to have the error matrix, and how difficult it would be to get.
4. Frank
• Has new asymmetry files with statistically weighted W.  Recommends Karl use that for RC.
• Working on PRL.  Maybe something to circulate by next week.
5.  Mark:
• We are all using newest E. Christy fit (7/22/05) as copied from Mark's newest MC update.
• There is a small region below our lowest W point but above threshold where the new fit gives zero.  Mark can "fix" this by adjusting the pion mass, but this affects the overall behaviour of the fit.  In the end, he chose to leave it alone since it will not affect even our lowest point.
• May request TN from Eric regarding fit.
• Newest Eric fit includes some new data that was previously not included.
• For neutron extraction, will need initial neutron model.  Maybe possible to use Hall A He3 data (E94010) as starting point. Get A1,A2 using any R.  Karl will take a look.
• Need someone to present RSS at Jan 5 Hall C meeting.  Oscar suggests Karl who will be onsite at that time.
6. Oscar :
   
• discusses his newest (draft) TN on smearing.
• Before next meeting he will try to get duality integrals using his fit compared to pdfs.
  
7.  Shige:
• Working on Neutron extraction.
• Discussion of smearing functions.
• Oscar: may need to go to 45-50 MeV bins for deuteron to help statistics.
8. Karl:
• Latest batch of RCs sent to Oscar for iteration.
• Needs to produce plots of before/after again.
• Shows Gamma_1, Gamma_2, GDH for proton.
• Mark: There is an newer Soffer curve.  He will forward article to Karl.
• Someone (Oscar, Mark?) suggests using PDFs for DIS contribution to Gamma_1.
  
• Oscar: can use E143 g1/F1 fit (along with any F1; NMC for example) for DIS contribution.  The kinematics of RSS were deliberately chosen to match one of the E143 settings. This was mentioned in RSS proposal specifically for the integrals.
• Mark: Fatemi nucl-ex/0306019 (2003) for EG1a Gamma_1.
• Might be nice to see Gamma_T plot.
• Frank : urges Karl to finish RC error estimate before starting work on other stuff.

10/27/05

1. Oscar at PANIC.  No weekly meeting.
   

10/20/05

1. Oscar:
• DF's need to be redone using Eric's 7/22/05 fit.
• But, concerned with R from 7/22/05 fit.
• Fit: A1 DIS goes like x^alpha + cubic polynomial. A2 DIS goes like x^alpha + cubic polynomial + sqrt(Q^2).
• Discussion of fit characteristics.
• Discussion of Nitrogen asymmetry correction for Deuteron.
• Will be at PANIC all next week (Tue-Sat). Requests updated plots from Karl.
• Regarding d2: used to be 0.101, then 0.073, now 0.098.
• Regarding publication: goal to have all plots final by mid November, before Thanksgiving break.
  
2. Mark:
• There are four new tables of nitrogen asymmetry, with correct bin centers.
  
• Affects the fine binned data only.
• Will ask Eric about R.
  
3. Karl
• Shows latest iteration of Radiative Corrections.
• Suggests that we all use Mark's MC version of 7/22/05 fit, to avoid any small discrepencies in analysis ( i.e. pion mass definition, wdif cut). All agree.
4. Frank:
• Has added statistical weighted W to his newest output file.

10/13/05

1. Oscar:
• Can "see" roper in the raw RSS asymmetries, before application of dilution factors.
• Believes we need to use Eric's 7/22/05 fit.
• We should iterate on the data.  Asks Karl to include nitrogen correction in RC analysis.
2. Eric
• 7/22/05 fit: Was performed because of the anomoly in previous fits in region just above the delta.  That problem arose from including Roper strength in photoproduction.
• The E94110 data on his website is not quite final.  It need to be corrected for few small effects:
• Fan speed was too slow on target, so overall density was wrong.  Data should be increased by about 1/2%.
• 0.5-2.0% effect due to optics of dipole/quad  mismatch.
• The table on web does NOT include these corrections.
• 3rd resonance: fit includes input from much lower energy data.
• Newest R should be consistent with Liang but smoother
• Oscar: asks Karl to produce plot of newest R compared to previous.
• Eric: suggests to compare Liang to 7/22/05 fit to estimate error.  Both are fit to the same input data.
• Plateu in asymmetry at ~1400 MeV just about right for Roper, since uncertainty of Roper in W is a few hundred MeV.
• Oscar: this plateu is also just above the 2 pion threshold.
• Eric : 7/28/05 fit should NOT be used.  It is "overfit" in some way.
  

10/06/05

1. Oscar:
   
• His fit form is from Stein paper, eq. 24.  Asymmetries are Breit-Wigner form. 4 resonances + cubic polynomial background.
  
• The fit now has Q2 dependence which seems reasonable.  A1 almost goes to -1 in delta region at Q2=0, which is to be expected for pure 3/2 transistion.
• Should be usable for RC.
• Finds "interesting" resonance at around W=1.35 GeV.
• The Delta moves to lower W.
• Mark at nuclear conference.
• Requests Karl check what resonances MAID puts at 1350.

09/29/05

1. Oscar: Regarding Karl's pass II radiative corrections.
   
• V1 vs VII  errors similar but bigger.
• Stat. error on A born.
• frc should come only from elastic.
• A_rc = A_rc^inel + A_rc^el
• A_rc has structure, but asymmetry itself does not. Why?
  
• rss_proton_as_rad_cor.sdc
• Measured asymmetry much different than born.
• Concern w/size of inelastic R.C.

09/22/05

1. Oscar:
• Inserted new pass II r.c. into his spreadsheet.  Makes comparison with Pass I.
• Sees same problem with errors that Karl had.  Explains why.
• Asks what Karl's "Quick Fix" was with first Pass RCs.  Karl doesn't recall, will check.
• Karl: offers to send code to Oscar to show how he is computing errors.
• Oscar: Probably fine, as long as Karl is adjusting statistics to account for loss of events due to tail subtraction.
• Oscar needs reasonable errors for his fit to work, so requests Karl put higher emphasis on fixing his error estimate.
2. Karl
• Working on input model for RC.
• Would like to use Oscar's fit, but it has no Q2 dependence (for now)
• Mark: suggests to use MAID to determine scale factor between energies, then apply this to our data to obtain model at lower energies.
• Oscar: or could just scale by Mott.
• Oscar/Mark : CLAS has data around 4.6 GeV.  Can see if scaling model makes sense.

1. Oscar
• ~rss/v3/rc/
• Shige: People at DNP seemed interested in our Elastic data, would be good to have paper on this topic.
• Mark will start writing a rapid communication.
• Re PRL: maybe don't need so many references to duality theory, since it is fairly well known now.
• Mark: would like more discussion on higher twist.
  

09/16/05

1. Karl:
• Pass II RC complete.  Shows plots of various physics quantities with the new correctioins.
• Oscar:  Why such big errors on A1 near threshold?  sigtt,siglt or g1,g2 don't have big error for same kinematics.  Looks like there is an inconsistency with the way Karl is propagating errors.
• There is a very large change in the radiative corrections around W=1400 MeV, visible especially in A1p.  Some concern expressed by all present that this looks a bit strange. 
  
• It is agreed that because of these issues, Shige will show Pass I radiative correction results at the DNP.
• Two issues Karl must resolve : Why such a large correction around W=1400 MeV, and why are the statistical errors inconsistent?
• Mark: suggests Karl try to turn off some resonances in MAID to improve d2 integrand curve.
• Karl: MAID d2(Q2) for proton is pretty bad, but MAID d2(Q2) for neutron reproduces HAll A data quite well.  Weird.
• Mark:  stt, slt errors are correlated.  Can't just propagate Apar, Aper errors blindly.
• Oscar : Still would like F1,R errors from Eric.
• Oscar: Yelana's GDH2004 talk mentioned issue with low W, low Q2 radiative corrections. 
  
• Oscar: need to compare Pass II radiative corrections to Keith Griffeon's results.
• Karl: will try to run polrad with only model.  no data input.  Will check d2(Q2) integral.
  
2. Oscar
• Fit: 4 Breit-Weigners along with constraint that A2<1.  Concludes that less "wiggles" is better.
• Will not be able to provide better g2ww plot for DNP. 
  
• Shige should use original plot for DNP talk.
3. Shige
• Shows DNP talks.
• Requests plot of full radiative corrections, not just inelastic, for Pass I.
• Requests plot of A2p vs. W.
• Requests plot of GDH(Q2) from Karl.
• Oscar will provide new Deuteron plots.
  

09/09/05

1. Shige:
• Working on talk for DNP (Sept 21). Recieves comments from all. Leaves on Saturday 17th. Talk is Weds. 21.
  
• ~tajima/rss/haw05
  
• Mark and Karl will provide new plots.
  
• Peter: Apar radiative corrections are suprisingly big.
  
• Peter: MAID not expected to work above W=1.4 GeV, since it is not multipion version.
• Mark recieved no response from Tiator re multipion MAID.  Karl previously recieved no response from his request to Drechsel.  Karl will try to contact Drechsel again, this time with a plot attached.
• Conclusion: A1 for single pion agrees with result for  all channels.  Interesting in its own right.
• A0 model is published. Perhaps we should show that instead.
• Other possibility: MAID with solid curve until W-1.4 GeV. Then dashed curve.
• Our data will be used to improve MAID.
• Peter: For Radiative corrections, better to use A1 as input.  From A1 A2 (+ F1,R christy) we can get everything else.
• Oscar: his A1 fit has no Q2 dependence.  Peter: Q2 dependence should be small.
• Peter : suggests showing MAID only up to Delta region.  Also requests bigger lables on figures.
• Oscar: g2p plot should not have MAID.
• Peter: Add neutron data to d2 plot? Also would be nice to compare to a model to see q2 dependence.
• Include a GDH plot?
• Deuteron Talk: Using Deuteron fits to data.
  
2. Peter
• Will send EG1A data to Mark and Karl to ensure we are all using same tables.
• Comparison of EG1a and EG1b: plots
• Findsa reasonable agreement between EG1a and EG1b for the partial integrals *Q2 (duality-type integrals)
3. Oscar
• ~rondon/rss/v3
  
• Shows effect of using fit in g2WW versus straight integration of g1 data: Not very big.
• Fit reproduces g1 very well.
4. Karl
• Discussion of proton form factor fit.  Peter: Jim Kelly fit is the best.
• A1 from Apar, Aper doesnt agree with A1 from g1,g2.  Not clear why.
  
• Peter: has technote with all relevent formula from E143, will provide to Karl.
• Will post tables of all quantities from Pass III.
• Oscar will translate his fit into Fortran, for possible use in RC.

08/26/05

1. Oscar: 
• Discussion of target and beam polarization, along with related systematics.
• A1,A2 fit:
  
• A W^2 Gamma_1^2    1/(W1^2-W^2)^2 +(W1-Gamma_1)^2
• Discussion of smearing technique for extracting neutron from deuteron.  Technique takes into account offshell kinematics  but not dynamics.

1. Karl
   
• Dipole form factor subroutine used for proton elastic radiative tail had incorrect value of Mup (too big by 30%)
• Will update with the subroutine used in monte carlo. 
  
• Mark:  montecarlo  Gm > dipole  by only few %.
• Mark: Some plots http://www.jlab.org/~jones/rss/mc_notes/notes_on_elastic_xn.html
• Conclusion: dipole probably good at RSS kinematics to within 5%..
  

08/26/05

08/16/05

07/25/05

07/15/05

07/08/05

06/17/05

06/13/05

05/27/05

05/19/05

05/13/05

05/05/05