The XSTAR Atomic Database

Round 1

Reviewer 1 Report

  See attached file.

Comments for author File: Comments.pdf

Author Response

We thank Reviewer 1 for the suggestions, which have led to improvements of the paper. The actions taken are indicated after each point and in bold face in the text of the manuscript.

 This paper reports updates to atomic data in the used x-ray modeling code XSTAR useful to astronomers and spectroscopists, and should be accepted with revisions according to following comments.

Major comments:

1. Whereas based on extensive prior work by the authors the paper focuses on important upgrades for K lines for x-ray diagnostics, the database also includes other data whose sources and accuracy need to be better delineated and described. For example, there are a number of references to Distorted Wave calculations for collisional excitation by Bhatia et al, Sampson et al., etc. Are those still the most uptodate, and not superseded by R-matrix calculations from the Iron Project and elsewhere?

ACTION: The XSTAR modeling code is mainly used in X-ray astrophysics, so most of the atomic data upgrades have focused inner-shell transitions particularly the K shell. Most of the atomic data for valence-shell transitions are from the original version of Bautista & Kallman (2001) as indicated in the last paragraph of Section 2. We have nevertheless included a new section (Section 6.6) to discuss the issue of atomic data accuracy.

2. The most significant upgrade is the inclusion of high-density effects on autoionizing resonances in DR resulting in continuum lowering, and resulting changes in model predictions. However, that leaves the RR rate coefficient to be obtained from other sources that may not be equally accurate. The unified RR+DR rates using the R-matrix method are for isolated atoms, but users might like to know the difference and the necessity of independent RR and DR treatment. The unified rate coefficients can be employed at densities lower than where high-density effects on DR manifest themselves, and for accuracy check on the density dependent DR rate added to an independently computed RR rate. This issue should be addressed as a possible source of uncertainty.

ACTION: We have rewritten the first paragraph of Section 5.2 and introduced a new paragraph where DR density suppression is discussed. Please see manuscript.

3. Generally, a comprehensive list of data sources is necessary so that users are not left wondering if a particular data source was included or neglected. To wit: the list of databases at the end of Section 2 does not include NORAD, which mostly contains high accuracy R-Matrix data, so it is not clear if and how much of that data was included.

REPLY: The list of atomic data sources listed at the end of Section 2 are those used in the original paper on the XSTAR database by Bautista & Kallman (2001, ApJS, 134, 139). As mentioned in Nahar (2020, Atoms, 8, 68), the NORAD atomic database was established at Ohio State University in 2007. No data from this latter source was included in XSTAR at the time, and it is therefore not included in this list.

4. The authors have demonstrated that high-density effects significantly affect data and models. But the common and necessary use of superlevels close to the continuum in modeling codes is another topic that requires a bit more explanation. The authors explain that superlevels are coupled to the ground state, presumably taking account of angular and spin symmetries for three-electron systems. And I assume the superlevel contribution to H-like Balmer series of recombination-emission lines, and corresponding transitions in He-like ions, are included via detailed cascade matrices. However, any prior error analysis concerning the use of superlevels should be reported, even if it means summarizing information in previous papers, albeit briefly.

Perhaps a section can be added on aforementioned uncertainties in data sources and approximations.

ACTION: An additional paragraph describing superlevels has been added in Section 2.

Minor Comments:

1.  Fig. 1: label the lines z, y, x, w, with corresponding transitions in the caption. This may seem like unnecessary rigor but careful astrophysicists might or should appreciate or familiarize themselves with proper spectrscopic notation. In that spirit, the widely misused epithet "triplet" should be in quotation marks, as in "K-alpha triplet" Section 4.5 on Silicon. This is not just academic and confusing with spectroscopic spin-multiplicity (2S+1), but practically related to identification of levels and transitions throughout the text and use of data in the database, viz. Fig. 2.

ACTION: The lines in Figure 1 have been labeled as requested. However, we do not agree with the recommendation to put the word triplet in inverted commas. In spectroscopy the widely used line multiplicity does not correspond to the spin multiplicity, and it is not our intention in this paper to change this convention.

2. The discussion on K-complex diagnostics is well written and indeed one of the main strengths of XSTAR models. My comment is regarding references to radiative and Auger damping treatment by Hickman and Robichaux. While the authors state that it is "essentially equivalent" to the treatment given earlier by Davies and Seaton (1976), they should add the reference to the seminal paper on DR by Bell and Seaton (1986), which is more general and which also underpinned the unified R-matrix treatment of RR and DR by Nahar and Pradhan (1992). That would serve to point out the fundamental relation between radiation damping and DR, described in detail by Seaton (1983) in his comprehensive review of QDT.

ACTION: The paper by Bell & Seaton (1986) is not about DR. Perhaps the misquoted reference is Bell & Seaton (1985). The latter has been included in the manuscript as requested.

3. Apropos Fig. 7, the authors demonstrate that orbital relaxation is an important consideration, especially for the K-shell. Maybe I overlooked, but are relaxed orbitals included in all HFR and Autoss calculations for XSTAR, and which ions?

REPLY: This query is explained in the last paragraph of Section 3.2. The following additional sentence was added at the end of the paragraph to fully clarify this point: “Due to these difficulties, most large-scale computations of photoabsorption and photoionization cross sections do not include orbital relaxation effects unless otherwise indicated.”

4. Apropos Fig. 8 and Eq. (9) on contimuum lowering due to Debye screening, state the density ne(cm-3) and temperature T(K) corresponding to mu = 0.1 and 0.25. Also, clarify units of Debye length in Eq. (8): atomic unit of length the bohr radius or cm; energy in rydbergs or hartrees; ne and Te on the right side presumably in cm-3 and K.

ACTION: Atomic units have been specified in Eq. (8). The energy unit (eV) in Eq. (9) is already stated. Units have also been specified throughout in the caption of Figure 8.

Reviewer 2 Report

The authors have compiled a very useful data set, having level energies, transition wave lengths, transition rates, resonance widths, and photoionization cross sections. Various quantities have been calculated using configuration interaction wave functions, Breit-Pauli R-matrix code etc. They have used Auto structure and Grasp codes also. For calculating photoionization cross sections simpler approximation like the distorted-wave approximation has been used because of a large number of levels. They have described various processes, given results in tables, and have shown some results in figures as well for a number of ions. Numerous references have been given.

It is a very exhaustive compilation. There are other available data sets. All of them ombined will prove very important for analyzing solar and astrophysical observations. It is impossible to detect any error in their presentation.   

Author Response

We thank Reviewer 2 for this positive report.

 Title: The XSTAR Atomic Database

Manuscript ID: atoms-1044355

Reviewer 2

The authors have compiled a very useful data set, having level energies, transition wave lengths, transition rates, resonance widths, and photoionization cross sections. Various quantities have been calculated using configuration interaction wave functions, Breit-Pauli R-matrix code etc. They have used Auto structure and Grasp codes also. For calculating photoionization cross sections simpler approximation like the distorted-wave approximation has been used because of a large number of levels. They have described various processes, given results in tables, and have shown some results in figures as well for a number of ions. Numerous references have been given.

It is a very exhaustive compilation. There are other available data sets. All of them ombined will prove very important for analyzing solar and astrophysical observations. It is impossible to detect any error in their presentation.

Reviewer 3 Report

The authors present and overview of the XSTAR project, including its database. This includes the updates made to handle modeling of higher density plasma.

The work is well presented and I have very few comments for the authors. The layout is excellent, walking through code purpose, database structure, physics requirements for extending to high density, and details of how this was implemented. in particular, I found the discussion of the O I K-alpha energy enlightening. I recommend the article for publication.

My comments, such as they are, should be considered as suggestions to further improve the paper. It is acceptable for publication without modification.

Line 361 (section 6.2):

  Are the unexpected metastable levels due to excited states with no decay mechanism? If so, could these not be systematically detected by a quick scan through the database for lines lacking any decay route, as opposed to waiting for them to present problematic spectra? Or to phrase it another way, how confident are you that you have purged all of the problematic levels?

Line 395 (end of section 6.5):

  If updating is an involved and error prone process, how are such errors prevented? Is there any methodology, or plan to implement one, beyond being really really careful?

Line 448 (end of section 8):

 The collections are an interesting idea. Is creating these part of the XSTAR development program, or is this present for users to adjust the data themselves?  Are there any obvious use cases?

Typos:

line 57: "intervene the code"

line 460: computed -> compute

Author Response

We thank Reviewer 3 for the suggestions, which have led to improvements of the paper. The actions taken are indicated after each point and in bold face in the text of the manuscript.

The authors present and overview of the XSTAR project, including its database. This includes the updates made to handle modeling of higher density plasma.

The work is well presented and I have very few comments for the authors. The layout is excellent, walking through code purpose, database structure, physics requirements for extending to high density, and details of how this was implemented. in particular, I found the discussion of the O I K-alpha energy enlightening. I recommend the article for publication.

My comments, such as they are, should be considered as suggestions to further improve the paper. It is acceptable for publication without modification.

Line 361 (section 6.2):

  Are the unexpected metastable levels due to excited states with no decay mechanism? If so, could these not be systematically detected by a quick scan through the database for lines lacking any decay route, as opposed to waiting for them to present problematic spectra? Or to phrase it another way, how confident are you that you have purged all of the problematic levels?

ACTION: most of Section 6.2 has been re-written to be more explicit regarding the problem of metastable levels.

Line 395 (end of section 6.5):

  If updating is an involved and error prone process, how are such errors prevented? Is there any methodology, or plan to implement one, beyond being really really careful?

ACTION: the last sentence of Section 6.5,  “This database model therefore can make updating an involved process that is certainly error prone.” has been replaced by the new paragraph “This database model can therefore make updating an involved and error-prone process, which can be mitigated by adopting an SQL-based relational database engine and devising a utility to perform the transcription to the atdb.fits file. Despite the wide variety of data types (see Appendix A) and record lengths, this is possible as shown by the development of the uaDB database (see Section 8). However, this scheme would hamper the generation of new data types, which is central to the XSTAR data curation strategy (see Section 2), and also lead to a more voluminous atdb.fits file. This proposition has not been considered in the XSTAR development roadmap.”

Line 448 (end of section 8):

 The collections are an interesting idea. Is creating these part of the XSTAR development program, or is this present for users to adjust the data themselves?  Are there any obvious use cases?

 ACTION: Thank you for querying an important point. At the end of the last paragraph of Section 8, we have added the following sentence: “Collections are not part of the XSTAR development program, but have been widely used in, for instance, the PyNeb Python package to model emission lines in gaseous nebulae and, more recently, to assess atomic data accuracy [89,90].”

Typos:

line 57: "intervene the code"

ACTION: the phrase "intervene the code" has been replaced with "interfere with the code".

line 460: computed -> compute

ACTION: word has been changed.

Reviewer 4 Report

Dear Authors,

I have read the article "The XSTAR Atomic Database" by Mendoza et al., and overall, it is an excellent article. It includes an excellent description of XSTAR and supporting databases. This paper will be a great source for users, and will surely provide a significant boost in the number of users of XSTAR

I only have a few minor points that I suggest the authors take into account .

1. Line 19-20
The text says "THe computational model determines...at every point...". How dense is the spatial model grid? I want to better understand what is meant by "at every point". Maybe a sentence could be added to elaborate.

2. Line 23
Change "depends" to "depend"

3. line 34
I suggest you add the word "relatively" before "high spectral resolution". In some fields, 5 eV resolution is not that high.

4. I suggest zooming in some more in Figure 1 so the reader can better see the differences between model and data. Maybe the energy range between 6.45 and 6.64 keV.

5. Line 56
I think you mean reference 74 instead of reference 47.

6.Line 144
It might be good for the reader to understand what happens if fewer than 1000 levels of more than 1000 levels are chosen so that the choice of 1000 is better supported.

7. Line 288
Change "round" to "around"

8. Line 375
Change "mayor" to "major"

9. Line 460
Change "computed" to "compute"

Author Response

We thank Reviewer 4 for the suggestions, which have led to improvements of the paper. The actions taken are indicated after each point and in bold face in the text of the manuscript.

Dear Authors,

I have read the article "The XSTAR Atomic Database" by Mendoza et al., and overall, it is an excellent article. It includes an excellent description of XSTAR and supporting databases. This paper will be a great source for users, and will surely provide a significant boost in the number of users of XSTAR

I only have a few minor points that I suggest the authors take into account .

1. Line 19-20
   The text says "THe computational model determines...at every point...". How dense is the spatial model grid? I want to better understand what is meant by "at every point". Maybe a sentence could be added to elaborate.

ACTION: the phrase “at every point from the source.” has been replaced by “at every user-defined spatial zone of the cloud.”

2. Line 23
Change "depends" to "depend"

ACTION: the phrase “which in turn depends” has been replaced by “a procedure that in turn depends”.

3. line 34
   I suggest you add the word "relatively" before "high spectral resolution". In some fields, 5 eV resolution is not that high.

ACTION: the word “relatively” has been added as suggested.

4. I suggest zooming in some more in Figure 1 so the reader can better see the differences between model and data. Maybe the energy range between 6.45 and 6.64 keV.

ACTION: Figure 1 has been replotted taking into consideration the proposed suggestions.

5. Line 56
   I think you mean reference 74 instead of reference 47.

ACTION: The misquoted reference has been corrected.

6.Line 144
It might be good for the reader to understand what happens if fewer than 1000 levels of more than 1000 levels are chosen so that the choice of 1000 is better supported.

ACTION: The phrase “The number of levels in the ion models is never above 1000 to enable practical runtimes when modeling” has been replaced by “The number of levels in the ion models is never above 1000 to enable practical runtimes when modeling (see Section 6.1)”.

7. Line 288
Change "round" to "around"

ACTION: word has been changed.

8. Line 375
Change "mayor" to "major"

ACTION: word has been changed.

9. Line 460
Change "computed" to "compute"

ACTION: word has been changed.

Round 2

Reviewer 1 Report

None