RR Lyrae Variables as Tracers of the Galactic Bulge Kinematic Structure

Round 1

Reviewer 1 Report

I have read the manuscript `RR Lyrae variables as tracers of the Galactic bulge kinematic structure' 
by A. Kunder.
The paper presents a review of the main results obtained in the last years
on the Galactic bulge formation history by using RR Lyrae stars, 
along with a comparison with those obtained with other diagnostics and 
and stellar tracers.

The paper reads well and in my opinion it deserves publication.
I don't see major flaws, but I have only few minor points that I would like 
the author to take into account.

•   Section 2, lines 107-109
    I think this statement is confusing: RR Lyrae are necessarily old and 
    preferentially metal-poor stars

 

•  Section 2.2
   Does the author think that the lack of calibrations accounting 
   for variable [\alpha/Fe] abundances can be a major issue for RR Lyrae     studies? I would expect them to be most likely α-enhanced.

 

• Section 2.3, lines 177 - 179
  I suggest the author to spend a few lines to explain better how RR    Lyrae are used as age indicators. I believe that a simple reference to a specific plot might not be sufficient in the context of a review paper.

 

• Section 2.3, lines 190 - 191
  It is hard to catch the meaning of this sentence. First the author refer 
  to an age gradient and then to a (chemical?) enrichment. Please clarify.

 

• Section 3.1 - I find quite counter-intuitive that here the description is mostly focussed on the RR Lyrae luminosity-metallicity relation, while the dominant one is the relation between luminosity and period (with eventually some dependence on metallicity, as reported in Eq. (1)). I think this approach is also not particularly useful for the following sections whose results deal with RR Lyrae intrinsic luminosity mainly coming from period analysis.

 

• Section 3.2
  Results from Kunder et al. (2020) appear to go in the opposite direction respect to other analysis. Do the author think it might be only related to the different sample sizes of the different surveys? I ask the author to discuss this point in some more detail.

 

• Section 3.2
  What's the impact of the reddening estimates and uncertanties on the derived RR Lyrae distances and eventually on the different results found by different authors?

  

• Section 4.2, lines 368 - 373
  How can the discovery of some of the past Milky Way accretion events, such as Kraken (Kruijssen et al. 2019, MNRAS 486, L20) or Gaia/Enceladus - Sausage (Helmi et al. 2018, Nature 563, 85; Myeong et al. 2018, ApJ 863, L28), fit within such a picture?

 

• Section 4.4 
  Apparently results from Kunder et al. (2016) and Du et al. (2020) are quite in disagreement. I kindly ask the author to comment on that and eventually to detail more on the impact of the contamination of the 
  Kunder's sample.

 

• Section 4.4
  I don't think it is necessary to introduce this section by going back to concepts that have been already discussed earlier in the paper. 

Author Response

> The paper reads well and in my opinion it deserves publication.
> I don't see major flaws, but I have only few minor points that I would like 
> the author to take into account.
>
>

I would like to thank the referee for a critical reading of this paper, and for
taking time to make astute comments that strengthen the presentation of
the paper.

>       Section 2, lines 107-109
>       I think this statement is confusing: RR Lyrae are necessarily old and 
>       preferentially metal-poor stars

RRL within the MW can have an age range from ~9.5 Gyr - ~12.5 Gyr.
RRL in globular clusters or in the solar neighborhood will have
different ages than those in the bulge, although yes, they are all
necessarily old (between ~9.5-~12.5 Gyr).  The idea here is that bulge stars
will have a larger range of ages than the RRL, although most are predominately
old.  Because the [Fe/H] distribution of the RRL is so peaked at FeH ~ -1, 
targetting bulge RRLs means we are sampling an older/more metal-poor
population than the average bulge stars.  

This has now been clarified in the text (see bolded text).

>      Section 2.2
>      Does the author think that the lack of calibrations accounting 
>      for variable [\alpha/Fe] abundances can be a major issue for RR Lyrae
>      studies? I would expect them to be most likely #-enhanced.

Even within the solar neighborhood, the RRLs vary in their
alpha-abundances.  Prudil+19 and Zinn+20, for example, show that a number
of local RRL originated from the Gaia-Enceladus system and these have alpha/Fe 
values that are ~0.20 dex lower than the RRL in the halo.

Deriving metallicities of pulsating stars like RRLs are inheritantly
challenging -- even with high-resoultion spectroscopy, it is important
to observe the RRL when it is in a quiescent state for best results.
Within ~0.4 dex uncertainties, which is not untypical for FeH metallicities
of RRLs, not accounting for alpha/Fe is most likely
not an issue.  I do think pushing the photometric metallicities to be
trustworthy at the 0.1-0.2 dex level requires correcting for the
alpha/Fe of a star.  This is something I am working on with a few
other RRL enthusiastists by observing local RRL with a range of
different alpha-abundances.

This has now been included in the text (see bolded text).

>     Section 2.3, lines 177 - 179
>     I suggest the author to spend a few lines to explain better how RR
>     Lyrae are used as age indicators. I believe that a simple reference to
>     a specific plot might not be sufficient in the context of a review paper.

This has been added. 

>     Section 2.3, lines 190 - 191
>     It is hard to catch the meaning of this sentence. First the author refer 
>     to an age gradient and then to a (chemical?) enrichment. Please clarify.

This has been clarified.

>     Section 3.1 - I find quite counter-intuitive that here the description
>     is mostly focussed on the RR Lyrae luminosity-metallicity relation,
>     while the dominant one is the relation between luminosity and period
>     (with eventually some dependence on metallicity, as reported in Eq. (1)).
>     I think this approach is also not particularly useful for the following
>     sections whose results deal with RR Lyrae intrinsic luminosity mainly
>     coming from period analysis.

The RR Lyrae luminosity-metallicity relation in the V-band
is the most widely discussed relation in the literature.  Because this
is a review paper, it is therefore not unreasonable to begin a section
on absolute magnitudes of RR Lyrae stars by reviewing this relation.
We prefer to keep the paragraph discussing the luminosity-metallity
relation in the V-band in the manuscript,
due to the importance of the luminosity-metallicity relation within the
RR Lyrae community.

The referee is correct that in recent years, there has been a shift to also
other luminosity relations, particularly those in the the longer wavelengths.
In the longer wavelengths, the period of pulsation is more important
than the metallicity of the RRL.  We feel Section 3.1 is more
comprehensive with a discussion of this progression of RRL luminosity
relations.

It is also worth noting, as we did, the limitations of this relation,
in particular that theoretical
models do not support a luminosity-metallcity relation in the V-band.  So 
although the use of a relation between luminosity and period is limited in
the sense that it has not been studied in as much detail,
one advantage compared to the more common luminosity-metallicty relation
is that it is theoretically motivated.

I have added a few sentences to motivate discussion also of the
RR Lyrae luminosity-metallicity relation in the V-band, and included
a few sentences discussing the RRL luminosity relations typically
used for RRLs in the Galactic center.

(Reviewer 2 requested including more discussion on the point that
theoretical models do not predict a simple linear realtion between
Mv(RR) and FeH, so this is why there are additional sentences
on this point.)

>     Section 3.2
>     Results from Kunder et al. (2020) appear to go in the opposite
>     direction respect to other analysis. Do the author think it might be
>     only related to the different sample sizes of the different surveys?
>     I ask the author to discuss this point in some more detail.

A paragraph has been added to discuss this point (see bolded text).

>     Section 3.2
>     What's the impact of the reddening estimates and uncertanties on the
>     derived RR Lyrae distances and eventually on the different results found
>     by different authors?

A paragraph has been added to discuss reddening estimates and uncertainties
on the derived RRL distances.

>     Section 4.2, lines 368 - 373
>     How can the discovery of some of the past Milky Way accretion events,
>     such as Kraken (Kruijssen et al. 2019, MNRAS 486, L20) or
>     Gaia/Enceladus - Sausage (Helmi et al. 2018, Nature 563, 85;
>     Myeong et al. 2018, ApJ 863, L28), fit within such a picture?

These major accretion events (and others) undoubtedly extend to the inner
Galaxy, but they have been shown to belong to the halo and not the bulge.
They shape the Milky Way halo -- including the extension of the halo
in the inner Galaxy -- but these have not been linked to the bulge.
This has been discussed now at the end of S4.2 (see bolded text).

>     Section 4.4 
>     Apparently results from Kunder et al. (2016) and Du et al. (2020) are
>     quite in disagreement. I kindly ask the author to comment on that and
>     eventually to detail more on the impact of the contamination of the
>     Kunder's sample.

The Du+20 and Kunder+16 results are actually in agreement
with each other.  Both survey's find slower rotation within the inner Galaxy
RRLs.  Neither Du+20 nor Kunder+16 had 3D velocities to clean for halo
interlopers.  This has been clarified in the text.

>     Section 4.4
>     I don't think it is necessary to introduce this section by going back
>     to concepts that have been already discussed earlier in the paper. 

We have removed the first paragraph introducing this section.

Author Response File: Author Response.docx

Reviewer 2 Report

see attached file

Comments for author File: Comments.pdf

Author Response

We thank the referee for bringing up important points that both clarify
the manuscript and increase the scientific significance presented.

> Line 56: It is the oldest stellar populations in the bulge can reveal...
> it should be The oldest stellar population in the bulge can ...

Thank you, yes, this has been fixed.

> Line 59: It would be helpful to have the presentation of the sections
> at the end of the introduction, as it is common practice

This has now been included

> Line 100 ..bulge red clump stars from the Blanco DECam Bulge Survey (BDBS)...
>
> A few details can be added here. Since the bulge has a complex structure,
> as discussed in the coming sections and there is a metallicity gradient,
> if would be interesting to know if the two surveys are sampling the same
> region of the bulge. If not, then a caveat can be included. The main
> result will not change, but the text will be more precise
>

Thank you, this is a good point.  We have now included a few sentences
regarding this.

> Line149:[Fe/H] -1.4 dex should be [Fe/H]= -1.4 dex

done

> Line 186: The ages of the bulge RRLs are compared in Figure 2 to bulge
> field stars determined Hasselquist et al.(2020)[47].
> What are the uncertainties on these age determinations?

sent Nataf an email

> Line 206: Theoretical models do not predict that a simple linear
> relationship between MV(RR) and [Fe/H] exists.
> Indeed theoretical models (e.g., Caputo et al. 2020) suggest that the
> MV - FeH relation is not linear but that the slope is going from
> 0.17 pm 0.069 mag/dex at FeH < -1.6 dex to 0.359 pm 0.027 mag/dex
> at FeH > -1.6 dex.  The authors can discuss this briefly.

We thank the referee for this reference.  We have discussed this in the
manuscript.

> Figure 3. A conceptual picture..
> This figure is not useful. I suggest dropping it.

OK, we have removed this figure.

> Line 228: the slope of the MV(RR) - [Fe/H] decreases to a = 0.25 pm 0.05
> mag/dex, more in-line with previous estimates.
> It should be noted that this includes also the correction for the
> zero point parallax offset of Gaia DR2 data, what is a source of uncertainty

Noted

> Line 233: Not only are longer wavelengths of light less sensitive to
> reddening, but passbands such as the infrared K-band are shown to have
> less dependence on metallicity.
> Also the dependence on evolutionary effects is less relevant at longer
> wavelengths.

We thank the referee for this comment, and have now noted this.

> Line 238: Therefore, currently RRL PL relations are well characterized from
> optical (R and I) to mid-IR wavelengths.
> Can the author provide more detail on the slopes here?

More detail has been added here.

> Line 264: and another, those that are most centrally concentrated.
> It should be  ...and another that is more...

done

> Fig.4. A comparison of the spatial properties of the RRLs in the inner Galaxy
> (red) compared to the giants in the inner galaxy (blue)
> The top panel is not very significant, showing only the selection effects
> due to observations and extinction. I suggest to remove it

We woud prefer to leave in this panel.  RRL in the inner Galaxy from
OGLE do not fall within b+-2 degrees of the plane, due to heavy reddening
in the opitcal passbands.  Investigations of RRL in the inner Galaxy
generally focus on the OGLE RRL, and so therefore, generally focus on the
RRL |b| > 2 deg.  The VVV RRLs from Dekany & Grebel+20 are different
in this regard.  They are the only statistical sample of RRLs that cover
the plane.  Highlighting explicitly that we are comparing RRL
directly in the plane with giants directly in the plane (from APOGEE) is
important when discussing the different spatial distribution of the RRL
as compared to the giants.  The Dekany & Grebel +20 paper has only 5
citations (it is a recent result).  We worry the reader will not
appreciate the VVV RRL sample probe an area of the bulge overlying the
plane, and therefore, where the barred signature should be strongest.

Two sententes have been added to illustrate the significance of the
spatial distribution of the VVV RRLs.

> Line 315: that the BRAVA kinematics (Figure 2) it should be
> ...(their Figure 2)

done

> Line 365 : are a kinematically hotter than the stellar should be are
> kinematically hotter

done

> Line 371; and even the metallicity gradient may be reproduced in these
> models unclear here if the discussion is about radial or vertical
> metallicity gradient. For clarity it should be explained that the
> metallicity gradient was for a long time considered a signature of a
> classical bulge

done

> Line 442: metallicities from came from Layden should be metallicity
> came from Layden

This paragraph has been removed due to a comment from a reviewer.

> Line 446: This has been confirmed by future studies with Gaia proper
> motions should be
> This has been confirmed by studies with Gaia proper motions

This paragraph has been removed due to a comment from a reviewer.

> Line 455: A population of RRL disk stars would point at a younger age of
> these RRLs since the thin disk is very cold and can not be more than
> 5~Gyr old
> please re-phrase it, the thin disk is older than 5 Gyr

This paragraph has been removed due to a comment from a reviewer.

> Line 462: and (given the high metallicity of the surrounding star cluster)
> the central supermassive black hole.
> The relation with the central black hole is not discussed in Kunder(2016).
> The argument presented here is unclear. I would simply drop this part
> of the sentence, or provide a reference

done

> Line 487-488: this analysis can be repeated using not only more precise
> proper motions from eDR3, but also there is - RR star and a Gaia proper
> motion. This sentence should be rephrased.

we have broken this sentence up into two sentences

> Line 491: Orbits are calculated employing a non-axisymmetric model for
> the Galactic gravitational potential
> It is unclear to which sample this applies BRAVA+Gaia EDR3 data?

Yes, it is BRAVA-RR+Gaia EDR3 data.  This has been clarified in the text.

> Figure 9. Caption Left: The mean velocity for RR Lyrae stars in the
> BRAVA-RR survey as a function of galactic latitude
> ... it should be longitude

done

> Figure 10 Caption Right: The velocity dispersion profile for the same
> stars. The lines represent the model by Shen et al. (2010) with a pure
> bar and no classical component...
> This is not what is plotted in the figure

Thank you!  This has now been fixed.

> Line 532: kinematic fraction ... in kinematic fractionation

done

> Line 572: These stars lie in the bulge, but do not sharing its abundance
patterns or kinematics ...
> in ...
> but do not share its..

done

> Line 583: today there only a few low metallicity stars
> ... in today there are only ...

done

> Line 597: although their is a population of RRLs
> ...in... there is a population

this phrase has been rephrased

> Line 606-611:  Comparing these observations in ... RRL observations.
> This sentence is too long and difficult to read

This sentence has been split up into three sentences now.

Author Response File: Author Response.docx

Reviewer 3 Report

Report on the review paper by Andrea Kunder “RR Lyrae variables as tracers of the Galactic  bulge kinematic structure”

 

The author presents a comprehensive review of usage of the bulge RR Lyrae stars to study the inner Galaxy structure and dynamics. I have only two comments:

 

Discussing the Subsection 3.2 the spatial distribution of RR Lyrae stars in the inner Galaxy, in particular whether the population of RRL stars trace the Milky Way bar, the author does not touch the important question of the completeness of RRL stars. The completeness of the tracer stellar population should be discussed in the paper.

 

In the description of kinematics of RR Lyrae stars in context of the Milky Way formation, i.e., why the barred population of stars co-exists with the unbarred stellar population, the author invokes the buckling instability, assuming that while the Galaxy evolves, the disk stars will trace the bar in different ways:  the hotter and older initial population would weakly trace the bar, whereas the cooler and younger population would follow a barred spatial distribution.

 

The origin of different stellar populations in the inner galaxy has, probably, a different explanation. The MW bar is formed by gravitational instability of a stellar-gaseous disk (see, e.g., Deg et al. MN, 2019, 486, 5391;  Khrapov et al. Galaxies, 2021, 9, 29). The bar instability, developing in the stellar-gaseous disk, involves disk stars in the bar but not hot halo stars with higher velocity dispersion. After its formation, the bar undergoes a second process, the buckling instability, which forms the observed ‘X-shaped’ Milky Way bar/bulge (see, e.g., Martinez-Valpuesta & Gerhard APJL, 2011, 734, L20). The author should discuss this scenario in the paper.

Author Response

>  The author presents a comprehensive review of usage of the bulge
>  RR Lyrae stars to study the inner Galaxy structure and dynamics. I have
>  only two comments:

We thank the reviewer for these pertinent comments and taking a critical look
at our paper.
 

> Discussing the Subsection 3.2 the spatial distribution of RR Lyrae stars in
> the inner Galaxy, in particular whether the population of RRL stars trace
> the Milky Way bar, the author does not touch the important question of the
> completeness of RRL stars. The completeness of the tracer stellar population
> should be discussed in the paper.

A discussion on completeness has now been included in Subsection 3.2.

> In the description of kinematics of RR Lyrae stars in context of the
> Milky Way formation, i.e., why the barred population of stars co-exists
> with the unbarred stellar population, the author invokes the buckling
> instability, assuming that while the Galaxy evolves, the disk stars will
> trace the bar in different ways:  the hotter and older initial population
> would weakly trace the bar, whereas the cooler and younger population would
> follow a barred spatial distribution.
> 
> 
> 
> The origin of different stellar populations in the inner galaxy has,
> probably, a different explanation. The MW bar is formed by gravitational
> instability of a stellar-gaseous disk (see, e.g., Deg et al. MN, 2019,
>  486, 5391;  Khrapov et al. Galaxies, 2021, 9, 29). The bar instability,
> developing in the stellar-gaseous disk, involves disk stars in the bar
> but not hot halo stars with higher velocity dispersion. After its formation,
> the bar undergoes a second process, the buckling instability, which forms
> the observed `X-shaped' Milky Way bar/bulge (see, e.g., Martinez-Valpuesta
> & Gerhard APJL, 2011, 734, L20). The author should discuss this scenario in
> the paper.

We have now started this section by discussing the
Martinez-Valpuesta & Gerhard+11 paper to
illustrate how different populations can come about due to evolution
of the disk.  We agree with the reviewer that there are a number of different
explanations that have been put forth regarding the formation of the
inner Galaxy.  We have focussed on the scenarios that naturally explain 
two or more different stellar populations that co-exist in the
inner Galaxy, where the inner Galaxy is within a Galactocentric radius of
3.5 kpc.
Our understanding is that the Martinez-Valpuesta & Gerhard+11 paper
focusses on the barred/boxy bulge and how it coincides with the long bar.
The long (in-plane) bar has not been discussed in this review, as this
is not where the RRLs discussed in this review reside.  This long bar
extends out from 3.5 kpc to ~5 kpc.
We prefer not to elaborate on the specifics regarding the
formation of the long bar, but to use this paper to illustrate the
complexity of stellar populations that can arise from secular evolution
of the disk.

 

Author Response File: Author Response.docx