Atomic Data for Plasma Spectroscopy: The CHIANTI Database, Improvements and Challenges
Round 1
Reviewer 1 Report
Review report:
Overall: Interesting clearly written Review paper of the development and current status of the well-known CHIANTI database. The paper also provides a very useful list of relevant literature references.
Atomic data for plasma spectroscopy: the CHIANTI database, improvements and challenges
G. Del Zanna and P. Young
Some requests for improvement:
1. ¨CHIANTI basic atomic rates have also been integrated within VAMDC 2 , a general portal for atomic and molecular data (see Albert et al., this issue)¨
VAMDC only offers the CHIANTI v7 data. Why is the data from the more recent versions not provided through the VAMDC access infrastructure?
2. ¨Within our experience, we had countless cases, even very recently, when original data associated with a publication have been lost.¨
Remarkable finding. But was the original data lost by the journal or could it not be retrieved from the authors of the paper?
3. ¨Also, the CE rates are directly included at the calculated temperatures, instead of being spline fitted, to preserve the original rates."
For how many temperatures are CE rates provided in v8?
4. ¨...the MCDHF calculations have been used to assess the reliability of line identifications, finding errors in literature values and in NIST, as described in the next section.¨
Section 2.4 suggests that identification errors in NIST have been found from MCDHF calculations. But the provided SoHO reference [47] dates back to 2007. For which ions is CHIANTI more accurate than NIST?
5. Sect. 3.2: The references
[57], [58], and [59] are rather old. But hopefully still relevant and useful. Have any of their PI cross-sections also been included in CHIANTI?
6. ¨The [67] formulae have been introduced in CHIANTI version 10 to improve the modeling of high density...¨
Is there a (perhaps projected) reference for v10 yet? The Dere et al. 2019 paper [62] appears to be the latest v9.
7. Please improve [84] about KAPPA with its actual journal reference: ApJS, 217, 14 (2015).
8. ¨The real challenge in the future is to build models where self-consistently all these physical effects, plus of course including PE, PI, CX effects.¨
Something should be improved/added in this sentence...
9. Was [86] actually published in A&A as their preprint states in arXiv (accepted sep 2019)? It is by the ´SPICE Consortium´, next alphabetically Anderson, A., with 88 co-authors. ADS only provides arXiv numbers.
10. Fig. 2 is interesting to estimate the differences with the zero-density results. Should the yellow line to the far right not be labeled as well in this graph? But if the curves represent the charge state distributions (instead of spectroscopic ionization stages) should they not be labelled with O^+, O^2+, O^3+, etc. ?
Author Response
We thank the referees for their comments. We have fixed a few typos and updated some of the references in the end with journal numbers.
We have modified a few parts, in bold. Here are our replies to the points
raised by the referee:
1. ¨CHIANTI basic atomic rates have also been integrated within VAMDC 2 , a general portal for atomic and molecular data (see Albert et al., this issue)¨
VAMDC only offers the CHIANTI v7 data. Why is the data from the more recent versions not provided through the VAMDC access infrastructure?
Reply
-----
It is mentioned in the Albert+ paper that the CHIANTI database format was significantly
changed in version 8 and again in version 9, the current one. As we have very limited resources, our first priority
has been to try and keep the database updated. The inclusion into VAMDC is not as straightforward as one might think.
2. ¨Within our experience, we had countless cases, even very recently, when original data associated with a publication have been lost.¨
Remarkable finding. But was the original data lost by the journal or could it not be retrieved from the authors of the paper?
Reply
-----
No, lost from the authors. Unfortunately not all the Journals require that
atomic data are made available on line.
3. ¨Also, the CE rates are directly included at the calculated temperatures, instead of being spline fitted, to preserve the original rates."
For how many temperatures are CE rates provided in v8?
Reply
-----
We have now clarified that
we provide the rates at the scaled temperatures corresponding to the original tabulated rates,
plus two (one for threshold and one for the high-energy limit), so users can
obtain extrapolated rates outside the original temperature range.
4. ¨...the MCDHF calculations have been used to assess the reliability of line identifications, finding errors in literature values and in NIST, as described in the next section.¨
Section 2.4 suggests that identification errors in NIST have been found from MCDHF calculations. But the provided SoHO reference [47] dates back to 2007. For which ions is CHIANTI more accurate than NIST?
Reply
-----
The baseline is always NIST energies. But for many states and ions we try to improve them, reviewing
recent literature. We have revised this section, to clarify this point.
All the details are provided in the energy files, and cannot be summarized here.
5. Sect. 3.2: The references
[57], [58], and [59] are rather old. But hopefully still relevant and useful. Have any of their PI cross-sections also been included in CHIANTI?
Reply
-----
So far, the hydrogenic Kramers approximation with the Gaunt factors from Karzas and Latter
is available, as well as the total PI from the ground states from Verner
These comments have been added to this section.
6. ¨The [67] formulae have been introduced in CHIANTI version 10 to improve the modeling of high density...¨
Is there a (perhaps projected) reference for v10 yet? The Dere et al. 2019 paper [62] appears to be the latest v9.
Reply
-----
Yes, it was mentioned somewhere else: Del Zanna+2020
7. Please improve [84] about KAPPA with its actual journal reference: ApJS, 217, 14 (2015).
Reply
-----
Thanks for spotting this. We have also put the Journals in other recently published papers.
8. ¨The real challenge in the future is to build models where self-consistently all these physical effects, plus of course including PE, PI, CX effects.¨
Something should be improved/added in this sentence...
Reply
-----
Of course.
The real challenge in the future is to build models where self-consistently all the above-mentioned physical effects
are included.
9. Was [86] actually published in A&A as their preprint states in arXiv (accepted sep 2019)? It is by the ´SPICE Consortium´, next alphabetically Anderson, A., with 88 co-authors. ADS only provides arXiv numbers.
Reply
-----
Yes ADS only provides the arXiv number. A&A lists the paper as published within a
special volume with a DOI, but with no volume or page number. We have now included the DOI.
10. Fig. 2 is interesting to estimate the differences with the zero-density results. Should the yellow line to the far right not be labeled as well in this graph? But if the curves represent the charge state distributions (instead of spectroscopic ionization stages) should they not be labelled with O^+, O^2+, O^3+, etc. ?
Reply
-----
We have followed the CHIANTI convention to label the spectroscopic ionization stages.
We have modified the caption.
Reviewer 2 Report
The present paper is well written and well comprehensive in the description of CHIANTI database. The paper is quite useful for the readers who are considering the use of CHIANTI database as well as the current users. The list of references is quite exhaustive.
The present paper deserves for publication in Atoms in its present form.
Typo:
line 182: ionisation should read ionization.
line 189: the the DW code FAC code should read the FAC code.
line 263: gauges should read forms
line 278: for for should read for.
Author Response
We thank the two referees for their comments. Aside from fixing the few typos
and updating a few references with the actual Journal number, we have modified the text (in bold) following the suggestions from the first referee.
Below are our replies:
1. ¨CHIANTI basic atomic rates have also been integrated within VAMDC 2 , a general portal for atomic and molecular data (see Albert et al., this issue)¨
VAMDC only offers the CHIANTI v7 data. Why is the data from the more recent versions not provided through the VAMDC access infrastructure?
Reply
-----
It is mentioned in the Albert+ paper that the CHIANTI database format was significantly
changed in version 8 and again in version 9, the current one. As we have very limited resources, our first priority
has been to try and keep the database updated. The inclusion into VAMDC is not as straightforward as one might think.
2. ¨Within our experience, we had countless cases, even very recently, when original data associated with a publication have been lost.¨
Remarkable finding. But was the original data lost by the journal or could it not be retrieved from the authors of the paper?
Reply
-----
No, lost from the authors. Unfortunately not all the Journals require that
atomic data are made available on line.
3. ¨Also, the CE rates are directly included at the calculated temperatures, instead of being spline fitted, to preserve the original rates."
For how many temperatures are CE rates provided in v8?
Reply
-----
We have now clarified that
we provide the rates at the scaled temperatures corresponding to the original tabulated rates,
plus two (one for threshold and one for the high-energy limit), so users can
obtain extrapolated rates outside the original temperature range.
4. ¨...the MCDHF calculations have been used to assess the reliability of line identifications, finding errors in literature values and in NIST, as described in the next section.¨
Section 2.4 suggests that identification errors in NIST have been found from MCDHF calculations. But the provided SoHO reference [47] dates back to 2007. For which ions is CHIANTI more accurate than NIST?
Reply
-----
The baseline is always NIST energies. But for many states and ions we try to improve them, reviewing
recent literature. We have revised this section, to clarify this point.
All the details are provided in the energy files, and cannot be summarized here.
5. Sect. 3.2: The references
[57], [58], and [59] are rather old. But hopefully still relevant and useful. Have any of their PI cross-sections also been included in CHIANTI?
Reply
-----
So far, the hydrogenic Kramers approximation with the Gaunt factors from Karzas and Latter
is available, as well as the total PI from the ground states from Verner
These comments have been added to this section.
6. ¨The [67] formulae have been introduced in CHIANTI version 10 to improve the modeling of high density...¨
Is there a (perhaps projected) reference for v10 yet? The Dere et al. 2019 paper [62] appears to be the latest v9.
Reply
-----
Yes, it was mentioned somewhere else: Del Zanna+2020
7. Please improve [84] about KAPPA with its actual journal reference: ApJS, 217, 14 (2015).
Reply
-----
Thanks for spotting this. We have also put the Journals in other recently published papers.
8. ¨The real challenge in the future is to build models where self-consistently all these physical effects, plus of course including PE, PI, CX effects.¨
Something should be improved/added in this sentence...
Reply
-----
Of course.
The real challenge in the future is to build models where self-consistently all the above-mentioned physical effects
are included.
9. Was [86] actually published in A&A as their preprint states in arXiv (accepted sep 2019)? It is by the ´SPICE Consortium´, next alphabetically Anderson, A., with 88 co-authors. ADS only provides arXiv numbers.
Reply
-----
Yes ADS only provides the arXiv number. A&A lists the paper as published within a
special volume with a DOI, but with no volume or page number. We have now included the DOI.
10. Fig. 2 is interesting to estimate the differences with the zero-density results. Should the yellow line to the far right not be labeled as well in this graph? But if the curves represent the charge state distributions (instead of spectroscopic ionization stages) should they not be labelled with O^+, O^2+, O^3+, etc. ?
Reply
-----
We have followed the CHIANTI convention to label the spectroscopic ionization stages.
We have modified the caption.
