----------------------------------------------------------------------
Report of Referee A -- LU10303/Achterberg
----------------------------------------------------------------------

This is a topical result and of interest to a wide readership so I
think it deserves to be published in PRL. However there are a few
changes that I think would make it a better paper and they are listed
in the following paragraphs.

The introduction needs to be more compelling in order to convince the
people that are not normally following the subject that they should
take a look. The first sentence could state (as can be found in the
abstract for reference 5) that the flare under discussion was "much
more luminous than any previous transient event observed in our
Galaxy". That alone should make the case that looking for neutrino
events from this source is not a priori hopeless.

In the third paragraph of the introduction there is reference to
three different sets of measurements and a line of argument that
states that the source is non-thermal (ie has a high energy tail). I
found this to be confusing and vague (eg "similar considerations
apply ...").  The part about hardening of the power law below 10 keV
before the flare seems out of place since it is so far away from the
TeV scale we are concerned about.

What is clear from reading the references is that four instruments
(RHESSI, Swift/BAT, RXTE and Chandra) have been used, by three groups
and they all suggest that there is a power-law component to the
source spectrum. Why not just state it like that and argue that this
means an important high energy component is suggested?

The first complete paragraph on page 11 needs to be rewritten in a
more linear way. It is really convoluted and confusing as it stands.

If I understand the argument, one assumes a power law flux of photons
hitting the earth and calculates with CORSIKA etc the number of muons
that should be seen by AMANDA for a given power (gamma) and
normalization (A_90). The fact that no muons are detected is then
used to set a limit on A_90 depending on the value of gamma assumed.
Meanwhile, in reference 10, the number of muons is calculated
analytically and a power law gamma-ray spectrum with hard cut-off (at
200 or 500 TeV) is assumed. The non-detection of muons can rule out
some of the candidate spectra in this paper. The first paragraph on
page 11 talks about "the same energy limits as in ref [10]". Does
this just mean that the power law spectra were cut off at 200 or 500
TeV and this changes the corresponding A_90? What about the effect of
gamma-ray absorption on the interstellar medium? You say it is taken
into account - here or in ref 10 or both? What does it do to the
power law? There's a lot going on here for a single short paragraph.

Perhaps a figure with integral flux vs energy would be helpful. At
low energy, one could have the satellite measurements as points and
at higher energies some AMANDA upper limits for different power law
assumptions. Theoretical predictions (eg with different powers,
normalizations and cutoff energies) can then be drawn as smooth
curves and one can see clearly what is ruled out and what survives.
(If you need the space you could omit figure 1 which is pretty bland
and can be given as a formula.)

The conclusions are a bit weak for something that is supposed to be
so important as to merit a PRL. There is a claim that "underground
(sic) neutrino arrays such as AMANDA and IceCube (not defined in this
paper) can be used as TeV gamma detectors for transient events."
Strictly speaking that has not been demonstrated. What has been
demonstrated is that for a short time window and an angular cut
consistent with the detector's angular resolution there were no
background counts and this allows an upper limit on flux to be
computed. You can say that about a lot of detectors. It is worth
emphasizing that AMANDA has a large duty factor and a wide field of
view so that it is sensitive to a large fraction of possible
transients.

Is the upper limit interesting? It is admitted that the neutrino
limit is not very competitive but what about the gamma limit? Does it
rule out fluxes expected from reasonable extrapolations of lower
energy data or expected from reasonable models of what might be
happening at the source? I suspect that has to do with the paragraph
at the top of page 11. If that paragraph is spelled out more
explicitly it will be easier to follow it up in the conclusions.