Authors: J.E.Campagne,M.Maltoni,M.Mezzetto,T.Schwetz 

Title: Physics potential of the CERN-MEMPHYS neutrino oscillation project 


We thank the referee for his/her positive evaluation or our work, the
detailed reading of the manuscript, and the critical comments. Below
we answer to the criticisms, and list the corresponding changes in the
text. We hope that with these changes our work is suitable for
publication.


REFEREE: (1) One of the most surprising feature of the results
obtained in this paper is that the mass hierarchy can be resolved up
to sin22theta13 =0.03 just by combining SPL and
betabeam(Fig.16). However, the authors seem to fail to place enough
emphasis on this unexpected result. I understood that some explanation
is already offered by the authors in page 25. What are the key
experimental features which enable the resolution?

ANSWER: As we explain on p. 26 (previously p. 25), this interesting
effect appears because of a delicate interplay of the (tiny) matter
effect in all four CP and T conjugate channels. This can also be seen
from Fig. 5, where the solutions with the wrong hierarchy appear at
slightly different locations for BB and SPL. We belief that the
explanation given on p. 26 is clear enough. A more detailed
investigation of this effect goes beyond the scope of the paper.



REFEREE: (2)The authors view of the current status of the ideas for
future projects seems to be a bit outdated. Now another option of T2K
upgrade, which is called as T2KK, is available. In this option people
claimed that all the eight-fold parameter degeneracy can be lifted in
situ without relying on combination with atmospheric neutrinos,
thereby reserving the latter as a redundant cross check. Also they use
superbeam only, avoiding a costly beta beam construction. From
outsiders view it looks like the better option within the scope of
megaton class water Cherenkov than CERN-MEMPHYS. Even though the
authors do not fully treat the T2KK setting in their analysis, it
would be nice if the authors comments on comparison between these two
strategies.

ANSWER: In our work we have confined ourselves to the standard T2HK
setup, since the porpose of our work is not a T2K optimization study
investigating various configurations for that experiment. In contrast,
here T2HK mainly serves as a point of reference to which we compare
the CERN--MEMPHYS experiments. For this aim we prefer to stick to the
``minimal'' one-detector configuration at a relatively short baseline,
since two-detector setups with very long baselines clearly represent a
different class of experiments whose consideration goes beyond the
scope of the present work. We have added a paragraph to say this
explicitly on p. 4. 

Moreover, we have added at several places in the text comments on the
sensitivities of T2KK obtained in Refs. [34,35], which are the
previous Ref. [60] and the more recent T2KK publication which we have
added to the references. Specifically, we comment on T2KK on p. 22
(CPV), on p. 26 (mass hierarchy) and on p. 27 (octant degeneracy). In
addition we mention also the the wide band beam idea in relation to
the mass hierarchy deterimation, referring to the analysis [74].



REFEREE: (3)The most serious problem in appearance neutrino experiment
with water Cherenkov detector is the background issue, in particular
pi0 rejection.  While the authors refer Ref. [51] for 'tighter cut'
the reference is not informative at all. Also, one of the advantage of
the present analysis, as compared to theprevious SPL studies, is that the
authors take into account the spectrum informations. But, at such low
energies it is highly nontrivial to reconstruct neutrino energy
because of Fermi motion. The authors description of the procedure is
too brief to allow understanding such high performance of the
method. The reasonably informative description should be provided
for both aspects.

ANSWER: We quoted reference [53] (previous [51]) because it's the only
paper where the SPL analysis has been described. It had been performed
by D. Casper from the SuperKamiokande collaboration. We don't have
today all the details of this analysis, but the pi0 background can be
understood from general principles. We have added the following text
in sec. 3.3:

"Following Ref. [53], the $\pi^o$ background is reduced using a tighter
PID cut compared to the standard Super-Kamiokande analysis used in
K2K, but the cuts are looser than for T2K. Indeed, at SPL energies the
$\pi^o$ background is less severe than for T2HK. This is because the
resonant cross section is suppressed, and the produced pions have an
energy where the angle between the two gammas is very wide, leading to
a small probability that the two gamma rings overlap. This results in
a higher signal efficiency of SPL compared to T2HK (60\% against 40\%)
and a smaller rate of $\pi^o$ background."


Regarding the energy reconstruction, we think to have fully
illustrated the problem with Figure 2 and the related discussion in
sec. 3.2. We are using the same lepton momentum smearing function as
SK and we are using the most updated version of the Nuance neutrino
generator, that is worldwide considered the reference for neutrino
generators, and Fermi motion is taken into account. In ref. [34]
(previous [60]) the authors quote an energy resolution of 80 MeV for
QE events. By applying our methods we have 77 MeV resolution in their
energy range.



REFEREE: (4)The Dm2 sensitivity of T2HK as presented in Table 4
appears too good to be true. It is strongly believed that it cannot be
less than 2

ANSWER: We have added the following text to the discussion of Tab. 4:

"In the interpretation of the numbers given in Tab. 4 one should
consider that at accuracies below 1\% systematics might become
important, which are not accounted for here. We do include the most
relevant systematics (see Secs. 2 and 3), however, at that level
additional uncertainties related to, for example, the spectral shapes
of signal and/or background, or the energy calibration might
eventually limit the accuracy."



REFEREE: (5) CP sensitivity of T2HK obtained by the authors is
significantly different from that estimated by T2KK people. With the
same 5% error the former is up to sin22theta13 =2x10-3,but the latter
extends to much smaller theta13 region, as shown in one of T2KK paper,
Fig.7 in Ref.[60]. The authors of the paper include SK
experimentalists and hence their treatment of the errors might be more
appropriate than the present analysis. Therefore, at least some
comments must be made on this discrepancy.

ANSWER: In Ref. [34] (previously Ref. [60]) the systematical errors
are taken to be correlated between neutrino and antineutrino data (see
eqs. 3 and 4 of [34]), whereas in our calculation they are assumed to
be uncorrelated. We have verified that this has a notable impact on
the CPV sensitivity and can explain the different results.  We added a
corresponding remark in the text on p. 22. Note that also for the
CERN-MEMPHYS experiment we assume the errors to be uncorrelated, so we
compare all experiments on equal footing.


REFEREE: (6)It is not obvious which figure is nu only and which
combines nu and anti-nu. A comment in each caption might be helpful to
the readers.

ANSWER: We added nu and anti-nu running times in all figure captions.

REFEREE: (7)The fact that the spectral information solves the
intrinsic degeneracy is noticed for T2K I about 2 years ago in
Ref. [60].

ANSWER: We do not claim to have discovered this here for the first
time. We add references to [30,32,34] in the discussion of Fig. 6.


REFEREE: (8) The authors do not explain the reason why theta13
sensitivity has a sharp minimum at delta = pi, as seen in Fig.9; It is
unusually abrupt drop of more than an order of magnitude.

ANSWER: We add the following paragraph to the discussion of Fig. 9:

"The peak of the sensitivity curves around $\delCP \approx \pi$
appears due to the interplay of neutrino and antineutrino data. For
the Super Beams neutrino (antineutrino) data are most sensitive in the
region $\pi \lesssim \delCP \lesssim 2\pi$ ($0 \lesssim \delCP
\lesssim \pi$), and opposite for the \BB, compare also Fig. 14 in
Sec. 6.1. The particular shape of the sensitivity curves emerges from
the relative location of the corresponding curves for neutrino and
antineutrino data, which is controlled by the $L/E_\nu$ value where
the experiment is operated and the value of $|\Delta m^2_{31}|$. The
fact that the peak is most pronounced for the \BB\ follows from the
somewhat smaler $L/E_\nu$ of the \BB\ compared to the Super Beams,
whereas the shapes for SPL and T2HK are similar because of the similar
$L/E_\nu$ values.