(Last Updated: Feb xx, 2005)

Comments and Answer (Internal review round 1)

[I.Kato] [Y.Hayato] [T.Ishii] [M.Sakuda]

I.Kato

(Paragraph 3)

Q. Here you describe the SciBar detector. Why don't you cite SciBar NIM paper?

A. In fact, it is refered at the last sentence of Paragraph 2. But I modified some sentences around
there and that position is moved to Paragraph 3.
(Paragraph 5, Line 12: ... verified by a timing distribution.)
Q. I think it cannot be understood by people outside the K2K.
A. Yes, I agree with you and I received same comment from Ishi-san and Nakaya-san in fact.
I'll remove this sentence along your suggestion.
(Paragraph 6, Line 4: ..., where the proton is below threshold or overlaping with other particle tracks.)

Q. What is the fraction of these events?
A. Roughly 2:1 (threshold : overlaping)
(Paragraph 6, Line 12: The events with the energy deposit less than.....)

Q. I think there is no description on "no additional hits around the vertex strip" in your analysis note,
while there is on vertex activity. What kind of criterion do you apply about the "no additional hits"?
A. In my Note, that term correponds to "2D track cut", where '2D track' means hit cluster
(consists of more than or equal to 2 hits including vertex strip) around the vertex strip which is not
associated with visible(=3D) track.
(Caption of figure 1: Energy deposit distribution in the vertex cell)
Q. You should use the word "strip" instead of "cell", to make a consistency through the paper.
A. Yes, I agree. I'll modify it.
(Paragraph 6, 4th line from bottom: ... the nuclear potential set at 27 MeV.)
Q. Could you provide a reference for this number?
A. Reference is Nuclear Physics. A292, 445, 1997. But I don't think it's necessary to include.
(Paragraph 7, Line 2: In order to reduce the uncertainty due to ..)
Q. What you did is to constrain the uncertainties based on our data. Therefore, I think the wording
"In order to reduce the ..." seems to be somehow misleading.

A. Yes, I'll modify it.
(Paragraph 10, Line 7: The uncertainties in QE and CC1pi interactions are estimated by changing the axial vector mass
by +/- 0.10 GeV/c2.
Q. What was the reason for this error? Do we need to have a reference for this number? Furthermore,
I thought that we previously quoted +/- 0.2 GeV/c2 errors on axial-vector mass of CC1pi. Why is it decreased?

A. At the oscillation analysis '02, we conservatively quoted 20% error on MA(CC1pi) to explain the lowQ2 deficit
by its uncertainty. But as you know, this is the maximal possible error and ~10% is preferable from the viewpoint
of experimental data. Because we already know the main cause of this deficit now, I think 10% is reasonable error.
(For your information, If we employ the 20% error, the final result is not changed at all.)

(Paragraph 10, latter half part (from Line 11): description on the error on the CC1pi suppression)
Q1. At first, I think CC1pi suppression is not clear for people outside the collaboration.
I think the method to suppress CC1pi sould be mentioned, or at least sould be cited from
oscillation paper.

Q2. What you did here is corresponding to "a kind of" simultaneous fitting for MC tuning using
entire q2 regions of 1-track/QE/non-QE-proton samples and q2>0.1 region of non-QE-pion to
constrain an ambiguity of CC1pi suppression within our observation.
In this sense, why didn't you employ really simultaneous fitting? Is there any reason
why you did not use low q2 region of 1-track/QE/non-QE-proton sample for MC tuning?

Q3.Further, there is some contribution from CC-coherent in the low q2 region of 1-track and
non-QE-proton samples. How did you deal with it in estimating the amount of CC1pi suppression?


A1. Yes, I'll do so.

A2. As you know, some contribution from CC-coherent pi can be seen in lowq2 region of 1track/nonQE-proton.
Because BG fraction should be evaluated with samples which don't include coherent-pi componente, I didn't
use these events.

A3. I used the center value of the final result obtained from non-QE-pion sample for CC-coherent pi. After
applying this lowQ2 correction on 1tack/QE/nQE-proton (q2<0.1) sample, I checked the possible difference
between Data and MC and treated it as a systematic error on the background subtraction.
(Paragraph 11, Line 10: There are other models predicting lower cross sections, but ...)
Q. I think we need to have references to these models.

A. Yes, I agree with you. I add there references.

Y.Hayato

(4th paragraph)
Comment1 : As for the Bodek&Yang correction, I think the word 'cross-section' is little bit ambiguous.

Comment2 : Last two sentenses (nuclear effects) are not correct. As far as I understand, their model
does not rely on the experimental data which you mentioned. Those data has been used to see
the agreements with other experimental data.

A1. Yes, I agree. I remove the word of 'cross section'.
A2. Yes, you're right. I'll correct it as follows. 'which reasonably agrees with the experimental data.
(7th paragraph)
Comment : sentense 2 is too long. Could you split? Also, could you explain how bad the initial parameters
were? Are those fitted parameters reasonable within errors?

A. Yes, I agree and will modify it. For the goodness of initial parameters, could you check P15 of analysis note?
I showed q2_rec distributions before fitting there. For the fitted result, I can say they are reasonable because those
values are within expected uncercenties.
(8th paragraph)
Comment : How about adding 'As a result,' or something like that at the beginning of the last sentense?
A. Yes, I'll add it.
(9th paragraph)
Comment :`10049 events are found in the MRD events' sounds strange. identified as MRD matching events
might be better.

A. Yes, I'll modify it along your suggestion.
(10th paragraph)
Comment : Could you explain how did you change the Bodek and Yang correction by 30%? (change the correction
factor of q^2?)
A. Now, we apply the weight of q^2/(q^2+0.188) on DIS. So correction factor is (1 - weight). I changed this correction factor by +-30% (=apply the weight of 1 - (1+-0.3)*(1-weight)) and check the variation of the final result from its central value.

Comment : Normalizations of some of the figures are not clear. please clarify.
A. In the paper, all figures are normalized by the number of MRD events with q^2(rec)
more than 0.10. I'll add the explanation in the caption of Fig2.
(Figure 1)
Comment : There are two figures in Figs.1 I think it is better to add figure ID, for example (a) and (b).
A. Yes, I'll do so.

Comment : If you just say that we can use reconstructed P_nu, obtained by assuming quasi-elastic,
can be used to calculate reconstructed q^2 for coherent pion production, people may be
confused. You just say the resolution and shift is small but you didn't mention about the
difference of the reconstructed and the true energy, which is one of the essential numbers.

A. I agree with you that the value is one of the essential numbers. But, I think Enu(rec) is less important
in this analysis because it is not barely used and (as I described below) that bias(~250MeV) doesn't affect on
the q2 reconstruction at all. Considering the length of the manuscript is strictly limited, I think I'll not add
the explanation in this paper.

Comment : I think the reason why the reconstructed q^2 is small is that outgoing lepton generated
by coherent pion production tends to go forward because of the characteristic of this
interaction. Therefore, there is no problem to use reconstructed q^2 for the event selection
but I think it is better to have little bit more explanations for non experts.

A. Yes, as you say, that small q^2 mainly comes from small angle of outgoing lepton.
(If muon mass is neglected, q^2 is written as 4*Enu*Pmu*sin^2(theta_mu/2) and 0.05 (GeV/c)^2
is expected for k2k case, where (Enu)~1.6GeV, (Pmu)~1.3GeV/c, (theta_mu)~9deg for coherent pion.).
I'll add the explanation in the part for coherent pion signature in SciBar detector.

Comment : Also, you cut at 0.1GeV/c^2 for q^2 but there is no explanation why you choose this value.
Of course, it can be easily understood by taking look at the figures but it is better to
add one sentense.

A. Yes, I agree. I add the explanation together with above description.

Q. One additional thing. Which flux did you use for this analysis?
A. I re-mesured(fitted) the flux with only 1Kt and Scifi (w/o SciBar) and used the result in my analysis.

T.Ishii

(P4 L11)
Comment : collide with an aluminum target -> hit an aluminum target
A. Yes, I'll modify it along your suggestion.
(P5 L25)
Comment : non-neutrino induced events ... 0.01% level : Is this true? non-beam-induced events ?
A. This plots is a event timing for all SciBar MRD events used in this analysis(Figure).
From this timing distribution, BG fraction including sky-shine is less than 10events at most. So order of magnitude is 0.01% level.
But, as I mentioned in Kato-san's session, I'll remove this sentence.
(P8 TABLE 1)
Comment : Track direction : Is this cut explained anywhere?
A. In fact, the explanation is found after the vertex activity part in the line14-15 of paragraph 6,
"events are required to have two forward going tracks...". I'll correct it in order.

(P9 FIG.4)
Comment : In the figure caption, you should say this is MC.
A. Yes, I agree.

M.Sakuda

(In page 4, right; a 20% suppression is allowed for CC1pi)
Comment1 :I guess that your estimation of the suppression depends on the amount of
coherent coherent that you assume. The suppression should be small when
you assume a small coherent pion cross section, and that can be large when
you assume a large cross section.

Comment2 : The size of total systematic error is 10 events below q2<0.1 (GeV/c)^2,
in Fig.3, the same as the size of the statistical error 10. This means that +-0.14
should correspond to 5 events in the plots. This looks to be too small if you plot
it in Fig.3. I may say 20% suppression may mean 20 events, which lead to +-0.60.

A. Sorry. My description was not good enough. Here, 20% means 20% of CC1pi events with q^2(true)
less than 0.1(GeV/2)^2 and doesn't mean 20% suppression in the lowq^2(rec) events. Including this part,
I'll modify this part more cleary.



A.