> Dear Prof. Townsend,
> 
> We have received the referee report for your manuscript "Tick Tock: MOST 
> Times the Magnetospheric Clock of &#x03C3; Ori E'' (ApJ88991) submitted to the 
> Astrophysical Journal.  The referee recommends that the manuscript be 
> Rejected as it does not fulfill the Journal's principal criteria for publication, 
> `new and significant research'.  The results are found to mostly confirm your 
> earlier work, with no new astrophysical implications.  The Journal is receiving 
> increasing numbers of submissions, and the Editors are instructed to raise 
> the bar on marginal submissions.  
> 
> When a report challenges the foundations of a study, I do not invite the 
> submission of a revised version.  Rather I request an email from the authors.  
> If you substantially agree that the paper does not present strong results, 
> please Withdraw it from further consideration by the Journal.  If you disagree, 
> please send me a detailed response to the authors principal objections, and 
> outline how a revised version will alleviate these difficulties.  You may also 
> request that I consult a second referee.  Upon receipt of your response, I will 
> consult the referee(s) and decide whether to go forward with revisions.  I 
> hope you can respond by December 1. 
> 
> 
> Eric Feigelson, Scientific Editor
> Astrophysical Journal
> apjedf@astro.psu.edu
> 

Dear Eric --

I'm writing with a response to the referee's report on our paper "Tick
Tock: MOST Times the Magnetospheric Clock of sigma Ori E". As I have
already mentioned, I believe the report does no justice whatsoever to
the paper, and I'm writing to request that the manuscript is sent to a
different, and hopefully less-partial, reviewer.

Below, I consider each of the referee's remarks in turn, and in almost
cases rebut them as mischaracterizations or misunderstandings of the
paper and/or the literature (or, in a couple of circumstances, as
outright falsehoods). However, there is one area in which the referee
is spot on with their criticisms: our citation of the historical
literature. In attempting to report on the recent developments in the
field of massive-star magnetospheres, we gave historical studies from
the 70's and 80's short shrift. In retrospect, we should have been
much more diligent in citing these papers. I can assure you that no
malice was intended by this, and we would be delighted to amend the
manuscript to give appropriate credit where it is due.

We would likewise be happy to make some structural changes to the
manuscript, most importantly moving much of the data reduction section
to an appendix.

However, we strongly reject the *scientific* criticisms of the paper
by the referee. The paper presents novel observations of a star of
especial interest to the massive-star magnetosphere community (the
first continuous, high-precision, space-based photometry), and
likewise reaches new conclusions about the physics at work in the
star's magnetosphere; as such, we believe it reaches the high standard
necessary for acceptance by the Astrophysical Journal. We wonder
whether our misstep with the citations somehow aroused the referee's
ire, leading them to write their vexatious and slanted report. Whether
this is is indeed the case or not, as you can see from our responses
below, the majority of the referee's criticisms (quoted using ">") are
simply without merit.

> ++++++++++++++++++++
> 
> REFEREE REPORT
> 
> 
> 
> This paper presents photometric monitoring of the
> interesting and well studied Bp type star sig Ori E with the MOST
> satellite in 2007.  A large fraction of the paper is devoted to the
> detailed description of data reduction and the light-curves 
> production. 
> 
>  The analysis of the MOST data confirms  previously published results
> obtained by other instruments.  No new significant findings on stellar
> properties are obtained from the MOST monitoring campaign.
>
>  The light-curve is shown to be stable, in agreement with 
> previous observations by other instruments,
>

These statements significantly mischaracterize our work. All previous
observations have been ground-based, and -- given the 1.2d rotation
period -- have therefore been unable to monitor the star continuously
from one cycle to the next. This should be contrasted with the new
observations presented in our paper, which -- being *space-based* --
are continuous and moreover achieve a much higher photometric
precision. It is only with these new observations that we are able to
truly examine whether there are cycle-to-cycle changes that might be
produced, e.g., from centrifugal breakout episodes.

Likewise, it is only with the very high cadence of the new
observations that we are able to make an *independent* period
measurement which does not rely on observations from any other
epoch. This measurement serves as a confirmation of the ephemeris
published by Townsend et al (2010); this is a very significant result,
as the ephemeris comprises one of the first *direct* measurements of
magnetic braking in a main-sequence star.

> 
>  but implications of the light-curve stability are postponed to a
> forthcoming publication. Only a short introduction and discussion are
> presented in the manuscript. Crude estimates are made to justify the
> lack of abrupt light curve variations within the framework of the
> rigidly rotating magnetosphere (RRM)  model that  predicts an unstable
> light-curve for sigma Ori E (Townsend &amp; Owocki 2005, ud-Doula et
> al.  2006).
> 

While we agree that the Introduction needs some additional citations
to the historical literature (as we mention above, and discuss further
below), we believe that it is otherwise entirely sufficient to
introduce the paper, especially given the additional material
presented in the following Background section.

The analysis in our Discussion section is based on the best-available
models for the magnetosphere of sigma Ori E, and we are very careful
to err on the side of caution when making estimates of the
magnetospheric mass. For the referee to dismiss this analysis as
"short" and "crude" does it no justice whatsoever, and demonstrates a
worrisome lack of understanding of the models and arguments employed.

> 
>  However, the work on model refinement is postponed to a  
> forthcoming publication  (Townsend et al. in prep.).

The referee is mistaken here. The Townsend et al. (in prep) paper
mentioned in the Discussion section focuses on the the X-ray emission
from the companion of sigma Ori E, which Townsend is currently working
on; it has *nothing whatsoever* to do with refining the light-curve model.

While we do mention the possibility of future model refinement later
in the Discussion, we stress that the magnetic Doppler imaging data
necessary for this refinement are not in our possession --- they
'belong' to Mary Oksala, who is not involved in this particular
project. (It's worth also noting that our MOST observations of sigma
Ori E predate the MiMeS collaboration responsible for obtaining the
circular polarization data used by Mary).

> 
> In my view, the two main results of this paper -- the ephemeris
> presented by Townsend et al. (2010) are confirmed, and no abrupt changes
> in the light curves are observed -- do not provide sufficient scientific
> content  to justify the publication in ApJ.  The methods of the  MOST 
> data analysis and the lightcurves productions are discussed in detail in 
> the manuscript. These  may be published elsewhere, perhaps in a 
> more specialized journal that does not require significant scientific
> results.

We strongly disagree with the referee's remarks. As we discuss above,
the ephemeris published by Townsend et al. (2010) is one of the first
(and few) direct measurements of magnetic braking in a main-sequence
star. For the wide community studying the rotational evolution of
stars, it is therefore of great interest, and its confirmation in the
present paper is a very significant result.

Likewise, the absence of abrupt changes in the light curve -- together
with our careful analysis of the magnetospheric mass vs. breakout mass
-- enables the paper to overthrow the previously-favored centrifugal
breakout paradigm for magnetospheric mass loss. For the massive-star
magnetosphere community, this is likewise a very significant result.

> 
> The sig Ori E ephemeris were already published in 2010 by Townsend et
> al., who presented photometry obtained in 1977, 2004, 2006, 2008, 2009.
> The apparent motivation for the  MOST 2007 monitoring was to find 
> variability as  predicted by the RRM model (Townsend et al.).  However,
> abrupt variability was never observed previously in this well studied
> star.
> 

As we mention above, previous observations have been ground based, and
came nowhere near the continuity and precision necessary to determine
whether abrupt cycle-to-cycle changes are occurring.

>   The only indication for centrifugal break-out was an X-ray flare,
> but Sanz-Forcada etal.  (2004) have already postulated that the X-ray
> flare originated from a late-type companion, and not from the Bp star
> itself.

This is misleading --- in fact, the possibility of centrifugal
breakout is motivated by the MHD simulations by ud-Doula et
al. (2006), as we clearly state in the Background section. Moreover,
although Sanz-Forcada et al. (2004) do indeed consider the possibility
that the X-ray emission is due to flares from a companion, this is
speculation on their part; the actual detection of a companion wasn't
made until the Bouy et al (2009) observations.

>   In the  discussion, some crude estimates are made to
> demonstrate that the magnetosphere of sig Ori E is well short of the
> level required for breakout episodes to occur.   It is also remarked in
> the manuscript (sect. 4.3) that the ud-Doula's et al. and Townsend et
> al. models are lacking sufficient microphysics to predict the observed
> light-curves in different wavelength bands.

Here, the referee again mischaracterizes the paper. As we mention
above, the estimates are built on the best-available models, and
provide a very robust upper limit on the magnetospheric mass; they are
in no way crude. Furthermore, the remark about microphysics in Section
4.3 *clearly* refers to the MHD simulations by ud-Doula et al. (2006),
specifically their ability to predict the energy yield of breakout
events; it does *not* refer the light-curve synthesis described in the
Discussion section.

>   This raises the question
> what the MOST campaign was aiming to achieve?  The refinement of the
> model which takes into account the MOST results is postponed to a future
> publication.  It would seem natural to present the observed 
> light-curves together with a more in-depth model discussion  in this
> forthcoming publication. 

The aims of the MOST campaign are clearly stated in the final
paragraph of the Introduction: "The original motivation behind this
observing campaign was to better characterize the star’s light curve,
and to search for any cycle-to-cycle changes arising from the loss of
magnetospheric plasma."

> Some important previous work on sig Ori E models and observations have
> not been considered. Sigma Ori E was the subject of intensive studies
> for a couple of decades before the follow-up works of R. Townsend and
> co-workers. 
> 
>  As an example, the periodic variable features have been
> ascribed to two circumstellar plasma clouds trapped in the magnetosphere
> and co-rotating with the star by Landstreet &amp; Borra (1978). It were
> Groote &amp; Hunger (1982) who suggested an asymmetry between  the two
> clouds to explain the UV spectral features. 
> 

As we discuss above, we recognized that this was a lapse of
scholarship on our part. In attempting to reflect the recent exciting
advances in each observational discipline (photometry, spectroscopy,
polarimetry), we inadvertently omitted many important references to
pioneering historical work (some of which are mentioned here by the
referee). This problem can easily be remedied with minor revisions to
the manuscript.

> The departures from a dipole configuration of the magnetic field in
> sig Ori E are also well known, see Leto et al. (2012) and references
> therein. 

This is simply false. The evidence for departures from a dipole field
comes from the study by Oksala et al. (2012) --- published *this year*,
and so not "well known" by any stretch of the imagination! Citing Leto
et al. (2012) here would be misleading, since they do not
present any magnetic measurements of their own (in fact, they
themselves cite the Oksala study).

> The model developed by Cassinelli et al. (2002), Maheswaran
> &amp; Cassinelli (2009, and ref.  therein)  predicts a stable
> circumstellar structure in magnetic rotating massive stars, and may
> explain the steady light-curves.
> 

Again, this is untrue. The models by Cassinelli and co-workers are
aimed at reproducing the observables (steady, twin-peaked H-alpha
emission; no periodic photometric variability) of classical Be stars,
and *not* Bp stars like sigma Ori E (which show rotationally modulated
Halpha emission and periodic photometric variability). Moreover, these
models invoke magnetic fields which are two orders of magnitude weaker
(~ 100 G) than detected in Bp stars, and predict quasi-Keplerian disks
which are quite different from the rigidly co-rotating circumstellar
structures seen in Bp stars.

>    In my view, the  manuscript would
>  greatly benefit if it were more balanced and open-minded with respect
> to the work of other groups and models and included more aspects  of
> our knowledge about sigma Ori E and other similar objects.
> 

We agree that the manuscript needs to be modified to better reflect
the historical literature.

> 
> Please, find the detailed comments below. 
> 
> Abstract.  It is not clear which findings allow authors to argue that
> breakouts do not play a significant role. What exactly are these -new-
> findings from the MOST campaign?  Please explain which  findings lead to
> a reconsideration of the Townsend &amp; Owocki (2005) and ud-Doula et al.
> (2006) predictions.  

The new findings are the absence of cycle-to-cycle changes expected
from breakout; and a careful estimate of the magnetospheric mass of
sigma Ori E, which turns out to be much smaller than that required by
the TO (2005) and ud-Doula et al. (2006) models for breakout to occur.

> 1. Introduction. The previous theoretical and observational work on
> sigma Ori E is not considered in the introduction. Specifically, the
> Groote &amp; Hunger (1982) and other classical works on sigma Ori E should
> have been mentioned. 
> 

We agree that the manuscript needs to be modified to better reflect
the historical literature.

> 2. Background.  First paragraph.  The sentence on X-rays doesn't
> represent the current status in the field. E.g. Naze et al. (2010) used
> MHD modeling and demonstrated that the X-ray emission from some 
> magnetic
> O stars is not explained by the MCWS scenario. 
>  Drake et al. (1994),
> Leone (1994), more recently Oskinova et al.  (2011) considered X-ray
> emission from Bp stars and found that their X-ray properties are
> diverse.
>  The X-ray emission of Bp stars is not significantly different
> from other B stars, in apparent contrast to the MCWS scenario. 
> It would be useful to consider in this section the model by Cassinelli
> et al. (2002), as it predicts a steady  configuration without breakout
> episodes.
> 

The Naze et al. (2010) study (which focuses on O?fpe stars not Bp
stars) reports that some predictions of the MCWS paradigm (e.g.,
rotational modulation of X-rays) are confirmed by observations, but
that MHD simulations are unable to reproduce the X-ray energy
distribution. However, they nowhere dismiss the overall MCWS paradigm,
or offer an alternative narrative.

Regarding the Drake, Leone and Oskinova studies, these indeed show a
diversity of X-ray emission from Bp stars -- but again, this doesn't
challenge the MCWS paradigm, given that the emission is sensitive to
parameters such as mass-loss rate.

Finally, we note once again that the models by Cassinelli and
co-workers are aimed at classical Be stars, and do not apply to Bp
stars.

> The  last paragraph should explicitly mention that Sanz-Forcada et al.
> (2004) did not attribute the flare to the Bp star itself, but to its
> low-mass companion.  Otherwise, this sentence is misleading. 
> 

This can easily be rectified.

> 
> 3. Observations and Data Reduction.  This section is long and technical.
> For the purpose of publication in the ApJ I would suggest to move a
> large part of this section to the appendix, or omit it entirely. The
> authors give an accurate description of their data reduction and
> light-curve production. Their explanations of the observed light-curve
> variations (Figs. 3-5) are clear. But it remains unclear, how  different
> is the data analysis from the standard MOST photometry procedure. In
> case the procedure is standard (as it is remarked in the text a couple
> of times), there is no need to devote a large part of the paper to its
> description.
> 

We would be happy to move this section to an Appendix. Note that the
approach we use to reduce the MOST data is necessarily non-standard,
by dint of the fact that sigma Ori E is the first star observed by
MOST in direct-imaging mode whose light variations are comparable to
the stray-light signal. This means that a great deal of care must be
undertaken in the reduction, and the length of Section 3 reflects this.

> 
> 4. Analysis 4.1 Periodic variations.  This subsection is is very short
> and largely repeats the 'Background' section. The variations seen in
> Fig. 1 were previously reported, based on the observations with other
> instruments.
> 

The "largely repeats" assertion is untrue --- Section 4.1 discusses
features of the light curve (esp. the differences between
primary/secondary eclipses, and the brightening after the secondary
eclipse) which are not mentioned in the Background section. Certainly,
in the interests of reinforcing the narrative in the Background
section, we repeat the fact that the eclipses are caused by transiting
magnetospheric clouds -- but this repetition involves just a *single* sentence!

> 
> 4.2 Period Measurements.  This subsection presents arguments to strongly
> confirm the T10 ephemeris. While this is a clear result, it is not
> surprising. The T10 ephemeris were not questioned before.  
> 

We beg to differ. The T10 ephemeris is a recent result, and
-- given its significance -- certainly benefits from independent
confirmation.

> 
> [Scientific Editor comment:  The dashed line showing a False Alarm
> Probability in the Lomb-Scargle periodogram is unreliable and should 
> be removed.  A number of recent papers discuss the inadequacies of 
> simple analytic LSP FAPs (Koen 1990, Reegen2007, Baluev 2008, Sturrock & 
> Scargle 2009, Zechmeister & Kurster 2009, and Vio et al. 2010).]
> 

This issue is readily rectified.

> 
> 4.3 Residual Flux.  The authors demonstrate that sig Ori E exhibits no
> periodic light variation other than the rotational modulation.  The
> authors also show that there was no centrifugal breakouts during the
> observations.  However, the interpretation is  not clear.  Ud-Doula et
> al. (2006) write "Townsend &amp; Owocki (2005) ... give estimates for the
> expected characteristic breakout timescale.... For the case of sig Ori
> E, they ... anticipate a whole hierarchy of breakout events extending
> down to timescales of days." In my view, it would be interesting if the
> authors would reconsider/refine their model to explain the apparent
> disagreement between this model and the observations.  However, this is
> postponed to future work. 

As we clearly enunciate in the Discussion section, the observational
absence of any significant breakout events is consistent with the fact
that the inferred/modeled upper limit on the magnetospheric mass is
almost two orders of magnitude smaller than the mass at which
significant centrifugal breakout will occur. This is a robust result,
and will not change with further model refinement; there simply isn't
enough mass in the magnetosphere for breakout to occur at a meaningful
level.

> 4.4 The 'Magnetospheric evolution' subsection is very short, and is
> basically the caption to Fig. 5. 
> 

It's unclear what point the referee is attempting to make
here. Section 4.4 is as long as it needs to be to discuss the
calculation and interpretation of the data in Fig. 5.

> 
> 5. Discussion
> 
> A simple estimate presented in the discussion shows that breakout
> episodes should not occur in sigma Ori E, while TO2005 predict otherwise.
> It should be explained, which of the assumptions/conclusions of TO2005
> were incorrect.
> 

Again, as we clearly explain in the Discussion section, the breakouts
do not occur because there is too little mass in the magnetosphere ---
some other process must be responsible for shedding mass, and we
consider some possible candidates.

> 
> The paragraph starting 'These finding are significant because they
> challenge the prevailing narrative for mass leakage ...'.  In section
> 4.3 it was explained that ud-Doula et al. (2006) models cannot offer
> much guidance. 
> 

As we discuss above, and as we clearly state in Section 4.3 of the
manuscript, the limitations of the ud-Doula et al. (2006) models have
been their inability to predict the energy yield of centrifugal
breakout events. But this is wholly tangential to the fact that
centrifugal breakout has --- until now --- been the prevailing
narrative for magnetospheric mass loss, as evidenced by the 44
citations to date to the ud-Doula et al. (2006) paper.

> 
> Simple estimates made in the 'Discussion' show that sig
> Ori E 'magnetosphere is well short of the level for significant breakout
> episodes to occur'.  Authors should clearly explain weather some
> breakout episodes in sig Ori E are predicted in their model or not. It
> should also be noticed that 'the mass leakage' is  'the  prevailing
> narrative' only in the context of the RRM model. 
> 

As we clearly state in the Discussion, "However, there are a number of
independent arguments which favor the alternative conclusion that
centrifugal breakout simply does not occur in σ Ori E, at least at a
level where it has any impact on the magnetospheric mass budget."

The fact that somehow mass is leaking from the star's magnetosphere is
wholly independent of the validity of the RRM model. The UV
observations of the star, dating back to the pioneering Groote &
Hunger (1982) study, clearly show that the magnetosphere is being
continually fed by the star's wind. And yet, the mass held in the
magnetosphere is shown in our paper to be (relatively) small. Ergo,
there must be some leakage mechanism at work.

> 6. Appendix The reasons to re-evaluate the parameters from Groote &amp;
> Hunger (1982)   are not clearly explained. It should also be noticed
> that there is a long-standing disagreement between the spectroscopic and
> evolutionary masses in B stars.

As we clearly state in the Appendix: "These findings [the Groote &
Hunger parameters], however, stand contrary to a number of
observational results indicating that σ Ori E is a bona fide member of
the cluster rather than a background star.". We then proceed to
describe these observational results.

Put another way, our reason to re-evaluate the parameters is simply
because the Groote & Hunger parameters place sigma Ori E at a distance
which is inconsistent with its membership of the sigma Ori cluster (as
revealed in multiple diagnostics), *and* also inconsistent with its
young age (as inferred from the spindown rate measured by Townsend et
al. 2010).

