LiH Scattering Paper - Referees Meeting¶
The next referees meeting on the LiH scattering paper will be on Friday 27th November at 14:00 GMT.Agenda:
- Overview of the note (John Nugent)
- Discussion (All)
[[computing-software:MICE_zoom_info|MICE Zoom Connection]] (login required)
Mariyan Bogomilov (MB)
John Nugent (JN)
Paul Soler (PS)
John Cobb (JHC)
Chris Rogers (CR)
- Overview that describes the scattering physics; description of the geometry and maths definitions; review of literature; description of the apparatus; description of the data; selection criteria; scattering distributions; comparison of data and MC; asymmetry and effect of fiducial cuts; momentum reconstruction, TOF recon vs truth; acceptance correction; convolution and deconvolution technique; comparison of data and model; systematics esp TOF systematic uncertainty; final results plots
MB: Cobb-Carlisle model has been dropped, why? JN: JHC asked for it to be dropped; JHC: CC model is identical to Moliere, so why bother?
Section 1: no comments
Section 1.1: ACTION Please clarify the wording (which shows the scattering angle). Scattering angle is the angle between the two vectors. The projected angle is the angle in the plane which contains the incident muon. Use the wording in JHC note/PDG book.
Line 75: wording "The RMS space angle is ~ sqrt(2) of RMS theta scatt". ACTION Tidy wording to indicate this is RMS.
ACTION Define r_e in the equation (believe to be classical radius of electron?)
ACTION Add a reference for eq. 12.
No mention of Moliere. ACTION Moliere should be introduced.
Fig. 2 - ACTION add station number to the diagram.
ACTION Define Tracker Reference Plane
ACTION Define hall coordinate system and position of apparatus.
Table 3: ACTION please clarify "Select muons from run at the target momentum" -> what is "target momentum?"
Line 158: why was the TOF window in MC offset compared to the data? not clear; we wrap it up in systematic uncertainty. JHC: should we shift the momentum or the TOF? JN: we shift the TOF, because it is a selection-level issue and we want to keep selection low-level. JHC expressed dissatisfaction with the procedure; if there is a problem with the momentum, it should be fixed in the momentum; CR: selection is a low-level process, we have to use low-level reconstructed variables. JHC: we should check the 240 MeV/c and 170 MeV/c as well. ACTION Please make plot 11(g) for 170 and 240 MeV/c.
Line 163: why 140 ps between LiH and data? JN: It is because of the energy loss in the LiH - we want to select momentum at the centre of LiH. JHC: should it be momentum dependent? JN: I will check. ACTION Do plot 11(g) for empty and full 170, 200, 240.
JHC: Please note the region in angular acceptance that the fiducial selection limits the acceptance. Estimate ~ 20 mrad acceptance.
Fig. 4 - MB: ACTION please correct the label on TKD - should be 0.3 m "fiducial diameter"
JHC: ACTION Clarify "detector reconstruction and beam behaviour". Should be "detector resolution". Unpack the "behaviour" to which you are referring.
JHC: ACTION Clarify "The simulation is then used for the prediction". "then" is inappropriate.
JHC: rotation discussion. Why is data and MC different in 11(d)? JN: the upstream tracks have been rotated. The plot is just included for completeness.
JHC: One line summary "the alignment has been corrected by setting the mean to 0" JN: confirm yes that's what we did
JHC: have you calculated a chi2 for 11(e) and 11(f)? JN: no, not for this; JHC: you should. This would validate the experiment. This is the main result!
JHC: you could in principle stop at 11(e) and 11(f)? JN: yes, but we would like a result "purely" for LiH by doing the deconvolution. These plots have detector effects and other materials. PS: low Z materials have less confirmation in the literature. JHC: why not do a MC vs data for empty channel, like 11(e) and 11(f) PS: well, we like to compare with Moliere as well
JN: clarified that the Bethe Bloch correction is applied going from TOF01 to absorber; but mean momentum is used for TOF12. ACTION Please clarify in the text.
JHC: where does t_e come from? JN: it is calculated by hand, there are not many electrons. You should be specific that "in the channel" means an averaged momentum between TOF1 and TOF2. ACTION Clarify wording
JN: 18(d) shows a momentum offset. JHC: why not correct the offset? JN: well I don't like to use the MC to optimise the analysis. JHC: well, e.g. you could calculate mean energy loss and fix it that way instead. JN: I have tried that in the past, referee's told me to pull it out! CR: note that the offset is present in empty and full - it is not an energy loss effect.
PS, JN: The label for Table 4 is not correct; it should be TOF01 ACTION Fix.
PS: please clarify t_e for TOF01 or TOF12; 25.40 ns is t_e for TOF01; ACTION please write down what t_e is for TOF12.
JHC: ACTION please clarify that "momentum is calculated for TOF12 in this way if there is a hit in TOF2" and "momentum is calculated in that way if there is not a hit in TOF2". Move the sentences to the beginning of the section.
JHC: what about the discrepancy in fig. 21 (TOF01 vs TOF12)? JN: 20(b) is US - DS.
JHC: can we do a data driven correction using Fig. 21? E.g. move TOF01 to be consistent with TOF12? JN, PS, CR: we can, but it is wrapped up in the systematics.
JHC: can you check the discrepancy also at 170, 240? JN: yes I can
PS: there are probably a set of fudge factors which make a consistent set of corrections between MC and data. Can we just apply them? JN: I am concerned about the proliferation of fudge factors.
ACTION Discuss and consider applying fudge factors to make means 0.
JHC: please clarify what is meant in eq. 21? JN: within angular acceptance is unnecessary. ACTION please fix.
JHC: points are asymmetric compared with the fitted polynomial. Why? PS: that is how the data is. The two curves are fitted separately for theta x and theta y. JN: the peak is not set to 1. Known MC inefficiencies are included. JHC: was there inefficiency in TKD? CR: no, that was in helicals. JN: correct using the red curve. Errors using the error bars. JHC: are you overinflating the errors? CR, JN: well, it is mainly in the tail bins. It probably isn't too bad. PS, JN: It is not dominant systematic error.
JHC: Confusing first paragraph. JN: I will clarify. JHC: mention incident angular distributions are different in x and y. Paragraph should be moved, it is not really relevant to convolution. ACTION move and clarify
Fig. 23 caption:
JHC: Moliere is a theory whereas Geant is an implementation. ACTION Please clarify.
JHC: line 280, ACTION please rework this paragraph; no where in the text does it say that you are using Geant4 and moliere.
JHC: What is happening at 50 mrad? What about 20 mrad? What is the conclusion of fig. 28? JN: the deconvolution is self-consistent in the scattering core. JHC: can we quantify that confidence? Can you look at the ratio between blue and black? CTR: one could include this in the error? PS: or just check that it is small compared to the errors, so we don't need to bother including it as an uncertainty? JHC: I would like to see the ratios. ACTION Plot the ratio fig 27 blue to black.
JHC: (skipping ahead) when you are calculating the RMS, are you including the weights? Do you take into account that there is an error in each bin? CTR: one could do a fit to a Gaussian and do a weighted fit? JHC: what are the errors included? JN: I use statistical and systematic errors. ACTION JN to check and clarify how RMS is calculated.
JHC: TOF should be an uncertainty in momentum, not on scattering angle. PS: well, bin migration is a fine way to do it as well. CTR: both ways are technically correct the analyst should decide. JHC: I think 70 ps is too big an uncertainty JN: It is the TOF resolution CTR: we see systematic effects in the TOF on a level consistent with ~50-100 ps.
Eq. (26) refer back to the first section "definition of angles"
Fig. 29 caption: should not be "theta y" ACTION fix.
PS: effect may come from beam misalignment JHC: this is residual effect from alignment of apparatus and beam
Fig. 30: for 200 and 240 data are consistent. 172 - something is going on. JN: there is no error from the deconvolution yet. JHC: there is a statistical error on the deconvolution as well. JN: yes I will add it. ACTION consider adding uncertainty from deconvolution.
JHC: why are there error bars on the black? JN: that is error on the data JHC: why are there errors on red and blue? JN: they are statistical errors CR: MC should have small errors (esp Moliere). JN: I will have a look. ACTION check error bars.
Table 11 and 12:
JHC: need to understand errors in previous plot
Can you explain 172 MeV/c p-value? JN: I found that theta_x uncertainty was unconvincing compared to theta_y. I have adjusted to use the (more conservative) theta_y uncertainty. ACTION ensure updated theta_x error is included in the note. This may fix fig. 30.
JHC: ACTION please explain that you have selected TOF bins from TOF01 and plotted the mean reconstructed momentum as per momentum reconstruction.
JHC: we are challenging the PDG value, we should be careful. ACTION we ran out of time here, need to pick this up next time and potentially be careful about wording.