PCLM-D-22-00161

Pangeo-Enabled ESM Pattern Scaling (PEEPS):  A customizable dataset of emulated Earth System Model output

PLOS Climate

Dear Dr. Kravitz,

Thank you for submitting your manuscript to PLOS Climate. After careful consideration, we feel that it has merit but does not fully meet PLOS Climate’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

I would request you to pay special attention to the point raised by the first reviewer, who says that your manuscript needs improvements from the point of view of the scientific discussion and content.

Please submit your revised manuscript by Jan 02 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at climate@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pclm/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Juan A. Añel

Academic Editor

PLOS Climate

Journal Requirements:

1. Please amend your detailed Financial Disclosure statement. This is published with the article. It must therefore be completed in full sentences and contain the exact wording you wish to be published.

a. Please clarify all sources of funding (financial or material support) for your study. List the grants (with grant number) or organizations (with url) that supported your study, including funding received from your institution. 

b. State the initials, alongside each funding source, of each author to receive each grant.

c. State what role the funders took in the study. If the funders had no role in your study, please state: “The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.”

d. If any authors received a salary from any of your funders, please state which authors and which funders.

2.  Please indicate the full and correct funding information for your study and confirm the order in which funding contributions should appear. Please be sure to indicate whether the funders played any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

3. We ask that a manuscript source file is provided at Revision. Please upload your manuscript file as a .doc, .docx, .rtf or .tex.

4. Please provide separate figure files in .tif or .eps format.

For more information about figure files please see our guidelines:  LINK

https://journals.plos.org/climate/s/figures 

https://journals.plos.org/climate/s/figures#loc-file-requirements 

5. Some material included in your submission may be copyrighted. According to PLOS’s copyright policy, authors who use figures or other material (e.g., graphics, clipart, maps) from another author or copyright holder must demonstrate or obtain permission to publish this material under the Creative Commons Attribution 4.0 International (CC BY 4.0) License used by PLOS journals. Please closely review the details of PLOS’s copyright requirements here: PLOS Licenses and Copyright. If you need to request permissions from a copyright holder, you may use PLOS's Copyright Content Permission form.

Please respond directly to this email or email the journal office and provide any known details concerning your material's license terms and permissions required for reuse, even if you have not yet obtained copyright permissions or are unsure of your material's copyright compatibility. 

Potential Copyright Issues:

Figure 1-22: please (a) provide a direct link to the base layer of the map (i.e., the country or region border shape) and ensure this is also included in the figure legend; and (b) provide a link to the terms of use / license information for the base layer image or shapefile. We cannot publish proprietary or copyrighted maps (e.g. Google Maps, Mapquest) and the terms of use for your map base layer must be compatible with our CC-BY 4.0 license. 

Note: if you created the map in a software program like R or ArcGIS, please locate and indicate the source of the basemap shapefile onto which data has been plotted.

If your map was obtained from a copyrighted source please amend the figure so that the base map used is from an openly available source. Alternatively, please provide explicit written permission from the copyright holder granting you the right to publish the material under our CC-BY 4.0 license.

Please note that the following CC BY licenses are compatible with PLOS license: CC BY 4.0, CC BY 2.0 and CC BY 3.0, meanwhile such licenses as CC BY-ND 3.0 and others are not compatible due to additional restrictions. 

If you are unsure whether you can use a map or not, please do reach out and we will be able to help you. The following websites are good examples of where you can source open access or public domain maps: 

* U.S. Geological Survey (USGS) - All maps are in the public domain. (http://www.usgs.gov) 

* PlaniGlobe - All maps are published under a Creative Commons license so please cite “PlaniGlobe, http://www.planiglobe.com, CC BY 2.0” in the image credit after the caption. (http://www.planiglobe.com/?lang=enl) 

* Natural Earth - All maps are public domain. (http://www.naturalearthdata.com/about/terms-of-use/)

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Does this manuscript meet PLOS Climate’s publication criteria? Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe methodologically and ethically rigorous research with conclusions that are appropriately drawn based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available (please refer to the Data Availability Statement at the start of the manuscript PDF file)?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS Climate does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The authors present a demonstration of linear pattern scaling, with an open code-base which downloads scenario information from the CMIP6 archive and performs a regression-based pattern scaling to predict the annual and monthly climatologies for a number of variables. A performance assessment is conducted, comparing predictive skill for different variables and scenarios in the context of internal variability.

The paper is reasonably clear, and the effort to provide open code is appreciated. However, given that there are few actual science advances in this paper, I'm not sure that PLOS climate is the right choice of journal - and a more dedicated code publication journal such as JAMES might be more appropriate. Furthermore - if the primary goal is to provide an open source, cloud-enabled alternatives to existing pattern scaling packages such as MESMER - then the interface to the current codebase should probably be improved.

The assessment of skill of conventional pattern scaling is appreciated, but not novel - as detailed below. In short - my recommendation for PLOS climate that the scope of the review of different pattern scaling approaches is expanded, to provide a science OR the authors submit the paper to a model development journal with a bit more work on the codebase.

1. The science presented here - though not incorrect, is not novel - and has been covered in a number of other studies [Telbadi 2014,2018, Herger 2015] and dedicated software packages [Osborn 2016, Beusch 2020]. Given this - an expansion beyond the current literature is needed. This could be achieved from the current study through two routes:

a) a more developed software interface: A Jupyter notebook is a good way to explore a topic individually, but a bad basis for a community tool -it's not portable, or easily installable on another machine and it can't be easily applied or called from other code. I recommend - if the authors want to develop a community tool exploiting the PanGEO infrastructure (a good goal) - then the codebase should be split into modules and functions which can then be called from a minimal script (and thus usable easily by others.

OR

b) a more extensive scientific assessment of different approaches to pattern scaling. The authors consider only a regression of pixel level change to global mean temperature here - the standard approach used in packages such as MESMER. It would be useful to compare the skill of this approach to others such as timeshifting (Tebaldi 2018), PCA approaches (Herger 2015) and machine learning approaches (Kasim 2021, Watson-Parris 2022) - but clearly this would imply an expansion of scope.

Minor issues:

- It's unclear in the current document what the training data for the regression predictions is, and to what degree the results are conditional on the choice of training data (single scenarios or multiple scenarios).

- It would be useful to provide additional context for accuracy - not just internal standard deviation in the model concerned, but also multi-model spread.

Page 5, reasons for error in pattern scaling - explicitly include the potential for slowly responding elements of the climate system which result, for example, in different land-ocean warming contrasts in transient and equilibrium scenarios.

Watson‐Parris, Duncan, Yuhan Rao, Dirk Olivié, Øyvind Seland, Peer Nowack, Gustau Camps‐Valls, Philip Stier et al. "ClimateBench v1. 0: A benchmark for data‐driven climate projections." Journal of Advances in Modeling Earth Systems (2022): e2021MS002954.

Kasim, M. F., D. Watson-Parris, L. Deaconu, S. Oliver, P. Hatfield, D. H. Froula, G. Gregori et al. "Building high accuracy emulators for scientific simulations with deep neural architecture search." Machine Learning: Science and Technology 3, no. 1 (2021): 015013.

Herger, N., Sanderson, B. M., & Knutti, R. (2015). Improved pattern scaling approaches for the use in climate impact studies. Geophysical Research Letters, 42(9), 3486-3494.

Tebaldi, C., & Knutti, R. (2018). Evaluating the accuracy of climate change pattern emulation for low warming targets. Environmental Research Letters, 13(5), 055006.

Tebaldi, Claudia, and Julie M. Arblaster. "Pattern scaling: Its strengths and limitations, and an update on the latest model simulations." Climatic Change 122, no. 3 (2014): 459-471.

Osborn, Timothy J., Craig J. Wallace, Ian C. Harris, and Thomas M. Melvin. "Pattern scaling using ClimGen: monthly-resolution future climate scenarios including changes in the variability of precipitation." Climatic Change 134, no. 3 (2016): 353-369.

Beusch, L., Gudmundsson, L., & Seneviratne, S. I. (2020). Emulating Earth system model temperatures with MESMER: from global mean temperature trajectories to grid-point-level realizations on land. Earth System Dynamics, 11(1), 139-159.

Reviewer #2: ### General comments

The article reads well, and is easy to follow, including good illustration and support of statements via figures. Structurally, it follows well, and the reader doesn't get lost in a myriad of sections and sub-sections; references are well places when they are needed, as well as pointers to available code on various code-sharing websites; figures are well captioned.

However, my view is that, even if it describes well the experimental setup and its conditions, together with the results (figures and tables), the article needs to demonstrate a bit more how PEEPS, the computational tool that is the main focus of the article, integrates and relates to other such tools and software packages, available elsewhere. I'd suggest this be addressed by adding a subsection where the authors demonstrate that:

- PEEPS is comparable to other similar tools performing pattern scaling, both in terms of performance and results;

- PEEPS offers net performance benefits compared to running a number of metrics on the actual model (ESM) data (which is clearly much more resource-intensive, but the authors don't tell us by how much, in rough numbers of computing resources)

I am aware this is more of a methods paper, than it is a systematic review of available tools, but even so, as a reader, I would like to be reassured that PEEPS is a solid tool for my work, and I shouldn't look elsewhere for another tool that does the same type of analysis. If these comparison results are available elsewhere, then simply referencing that source would be fine.

As a final point, I'd suggest the authors include a quick example on how to use the Jupyter notebook (configuration, guidelines for a basic used case, not anything too detailed) or, alternatively, a pointer to the technical documentation, and mention if they intend to build a Python package for ease of deployment, and installation by the users.

Specific questions and minor suggestions below -

### Tech/sci questions

- is T(x, t) in fact a 3-dim space-time vector? Then it should be expressed as T(x, y, t);

- are you setting a threshold for the lowest percentage that the residuals are within 1-sigma?

- why choose a 20 year averaging period (please explain briefly) - variability can still be detected if averaging period is set to, say 2-4 years, for monthly means data

- lines 174-176: worse/better performance is a vague term if a performance metric is not quoted, are we still looking at the deviation from the baseline? If so please give some numbers that support your performance statements

- line 226 onward (code and output): please specify the Python version (major version, to the very least); please specify the type of netCDF file the code outputs (netCDF4, 3, or classic); are the netCDF output files CF-compliant and/or CMOR compliant? (I assume CMOR, if CMIP6 data was used from ESGF)

- line 239 - performance: please specify clock speed of CPU of said "laptop", rough estimation of memory i.e. are you loading all the data in memory, or are you using memory-efficient tools like Dask or xarray+Dask?

### Typos and suggestions

- typo: line 51: python packages -> Python packages

- suggestion: line 53: ESM output -> ESM output data

- suggestion: line 54: "library" can be confused with its software meaning, suggest "database" instead

- suggestion: line 59: "Because of Pangeo" -> briefly explain (e.g. cloud-based analysis platform etc)

- suggestion: line 59: "the dataset and the PEEPS Jupyter notebook are effectively the same thing" - colloquial and incorrect,

a netCDF is a file type, whereas a Jupyter notebook is a web-based Python runtime ecosystem, please rephrase;

- typo: line 114: error -> errors

- suggestion: line 188: "There is strong reason to suspect that it would" -> "There is strong reason to believe that it would" (suspect bears a negative connotation)

- typo: line 226: python -> Python

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

Do you want your identity to be public for this peer review? If you choose “no”, your identity will remain anonymous but your review may still be made public.

For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: Yes: Dr Valeriu Predoi

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

PCLM-D-22-00161R1

Pangeo-Enabled ESM Pattern Scaling (PEEPS):  A customizable dataset of emulated Earth System Model output

PLOS Climate

Dear Dr. Kravitz,

Thank you for submitting your manuscript to PLOS Climate. After careful consideration, we feel that it has merit but does not fully meet PLOS Climate’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

==============================

Please submit your revised manuscript by Apr 22 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at climate@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pclm/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Juan A. Añel

Academic Editor

PLOS Climate

Journal Requirements:

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #2: All comments have been addressed

Reviewer #3: (No Response)

**********

2. Does this manuscript meet PLOS Climate’s publication criteria? Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe methodologically and ethically rigorous research with conclusions that are appropriately drawn based on the data presented.

Reviewer #2: Yes

Reviewer #3: Partly

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #2: Yes

Reviewer #3: No

**********

4. Have the authors made all data underlying the findings in their manuscript fully available (please refer to the Data Availability Statement at the start of the manuscript PDF file)?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #2: Yes

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS Climate does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #2: Yes

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #2: ### General comments

As a whole, I find this revised version of the manuscript in much better a shape than the previous version. However, there is one point I'd like to stress upon, and with the risk of sounding too technical, I believe the authors should address it - either in the form of a statement in the manuscript, or as actual code: I reckon the use of a Jupyter notebook as the only form of software for running PEEPS is limiting both the outreach, and the users that want to run the tool. A Python package (together with documentation, a streamlined way to create a virtual environment where PEEPS runs, tests, deployment to Python package indices etc.) would be the desired, and probably the most outreaching/portable way of shipping PEEPS, but for now, an equivalent Python script, that can be run by a user outside the Jupyter environment, would be most desirable. I am OK if the authors just include this as a TODO item and mention it in the manuscript.

Comments marked with ">>>...<<<" were initial comments for the original version (from November 2022);

I am mostly looking at these being addressed.

>>> The article reads well, and is easy to follow, including good illustration and support of statements via figures. Structurally, it follows well, and the reader doesn't get lost in a myriad of sections and sub-sections; references are well places when they are needed, as well as pointers to available code on various code-sharing websites; figures are well captioned.

<<<

VP-R1: still stands

>>> However, my view is that, even if it describes well

the experimental setup and its conditions, together with the results (figures and tables), the article needs to demonstrate a bit more how PEEPS, the computational toolthat is the main focus of the article, integrates and relates to other such tools and software packages, available elsewhere. I'd suggest this be addressed by adding a subsection where the authors demonstrate that:

- PEEPS is comparable to other similar tools performing pattern scaling, both in terms of performance and results;

- PEEPS offers net performance benefits compared to running a number of metrics on the actual model (ESM) data (which is clearly much more resource-intensive, but the authors don't tell us by how much, in rough numbers of computing resources)

I am aware this is more of a methods paper, than it is

a systematic review of available tools, but even so, as a reader, I would like to be reassured that PEEPS is a solid tool for my work, and I shouldn't look elsewhere for another tool that does the same type of analysis. If these comparison results are available elsewhere, then simply referencing that source would be fine.

<<<

VP-R1: this has been addressed and I could find a decent amount of comparisons to other similar tools/approaches; performance was also touched upon, not exhaustively but with a fair amount of info so the user has some idea what to expect in terms of resources - runtimes and memory

>>> As a final point, I'd suggest the authors include a quick example on how to use the Jupyter notebook (configuration, guidelines for a basic used case, not anything too detailed) or, alternatively, a pointer to the technical documentation.

<<<

VP-R1: instructions and documentation are still lacking a tad (I just had a look at the GitHub repo as well, which would benefit from a doc repo), but provided this is work in progress, and the submitted article's purpose is not to document the data producing runs, but rather, to document why one should run them at all, I am going to let this one go, provided the authors promise to improve he code's documentation :)

Specific questions and minor suggestions below

----------------------------------------------------------------

- GitHub repository that holds a frozen copy of the code: `https://github.com/JGCRI/linear_pattern_scaling` - great to have it on GitHub!

- couple more cases of lower case "python" instead of "Python"

- consider updating base Python to Python >= 3.9, 3.7 is already obsolete, and barely supported at all

- consider using a dependency/package manager like e.g. Miniconda with the Mamba solver for the environment where the Jupyter notebook runs

- consider implementing automated tests

- these last points are very technical, so no need to address them in the manuscript, just recommending them to improve PEEPS' quality as a piece of software (I can help with any of these, should authors need assistance)

Reviewer #3: In this manuscript, pattern scaling is used on the CMIP6 ensemble to compute a dataset of patterns for annual & monthly mean temperature, mean precipitation changes and relative humidity. This work is a relatively straight-forward extension of former works on pattern scaling and would ease several applications. The code used is provided, an appreciated effort for the sake of replication of data. The code is designed to avoid local storage of the CMIP6 ensemble, a well-suited approach.

However, there are two major drawbacks in this work. First, the assessment of the performances is not good enough to justify the use of such a database for future applications. Then, some sentences are shockingly wrong ranging from blatant misinterpretations on pattern scaling to simply wrong affirmations on simple climate models. I was so shocked that I almost rejected this article, though I consider that these two very serious issues may be fixed by the authors. Only for this reason, I recommend a major revision.

Major revisions:

A.Uses of pattern scaling

This manuscript confuses two uses of pattern scaling. Pattern scaling allows to replicate the patterns of Earth System Models (ESMs), allowing for the replication of the run by the ESM. Because of the complexity of the Earth system, several timescales are at play for local climate responses to global warming. Under a SSP1-2.6 or a SSP5-8.5, the local patterns calculated for these specific patterns will be different, because these scenarios will favor different warming modes. These effects will be more or less pronounced depending on the region and the variable.

Thus, the applications of these patterns depend how you calculate them. Either the pattern is trained on a specific scenario, like here for historical and SSP scenarios, or the pattern is trained on a subset of scenarios. Emulators based on pattern scaling are actually based on the second approach in the vast majority of cases.

Lines 22-41, you remind about the modeling chain making use of Simple Climate Models (SCMs) and pattern scaling for impacts modelers or integrated assessment models. In this framework, the global mean temperature (GMT) from the SCM is not necessarily the GMT corresponding to the scenario used for the pattern. Because SCMs are often run in probabilistic setups, representing the uncertainty in modelling, it will actually almost never be the case. Doing so forces the user to try to replicate a prepared scenario and make strong assumption by weighted averages on responses from pattern scaling. And it prevents the user to run unprepared scenarios. As mentioned before, the alternative is to use pattern scaling trained on a subset of scenarios. It averages the different modes of the climate responses but effectively allows to be applied to scenarios within the range of the subset.

The reference to Wells et al., 2022 (https://egusphere.copernicus.org/preprints/2022/egusphere-2022-914/) made in this manuscript is actually a good example of how such errors may be made. For instance, its Figure 3 exhibits these differences in patterns. Besides, applying patterns from SSP5-8.5 to a SSP1-1.9 makes use of the wrong warming modes. Even worse, applying patterns from SSP1-1.9 is a gross extrapolation, basically using the fit out of its domain of validity.

For this reason, I recommend to reconsider your approach on these spatial patterns depending on which envisioned usage for this database.

a.These patterns are meant to replicate specific scenarios, and no others. In this case, I recommend to save along the GMT timeseries. It does indeed save storage space and computation time for impact modelers, although it restrains the range of applications.

b.They are meant to emulate a whole range of scenarios. Then I recommend not to train these patterns on specific scenarios but on subsets. It corresponds indeed to the average response, thus better applicability but lower overall performances (e.g. SSP1-2.6 pattern applied on SSP1-2.6 GMT performs better than average pattern on SSP1-2.6 GMT).

c.I acknowledge that it represents an extra work, you may choose not to do so, but this has to be mentioned. You may suggest methods on how to use the appropriate(s) pattern(s) for unprepared GMT pathway, although it requires some careful thinking (e.g. taking the two patterns of the two SSP that are the closest in GMT then averaging).

Regardless of the decision, it needs to be highlighted in the introduction and emphasized in the usage notes and probably even reminded in the method. It needs also to be corrected in the conclusion (lines 329-341).

B.Shocking affirmations

Lines 11-13:

This is not about “precision” or “accuracy” of climate emulators, but about their capacity to mimic ESMs. Of course, they are simplified versions of ESMs, but it neglects several aspects. I illustrate here with SCM as case for climate emulator.

For instance, biases in CMIP6 ESMs may advocate for ESMs to be less precises than SCMs rescaled on observations. Besides, thanks to their simplicity, SCMs may include processes still in development for some ESMs (e.g. fully interactive atmospheric chemistry for N2O and CH4), also advocating for more accurate representation.

Depending on the SCM, the emulator is trained not only on ESM data, but also on Dynamic Vegetation Models, Atmospheric Chemistry Models, etc., models of higher complexity in a nutshell. Differences of SCMs to ESMs may be caused by different modeling structure or assumptions, missing processes from averaged responses or extra processes by combining models.

The reason for lower computational cost of climate model emulators isn’t a lower accuracy or precision of their outputs, but the trade-off in their spatial & temporal resolutions.

Please reformulate entirely this paragraph, this is not about comparing accuracy of models.

Lines 26-29:

If mentioning simple climate models, Hector is not the only one, and by far. I recommend citing the RCMIP phase 1 and phase 2 papers:

-https://doi.org/10.5194/gmd-13-5175-2020

-https://doi.org/10.1029/2020EF001900

Lines 206-209:

1.There is nothing such as “the amount of climate change”. One would refer to changes in specific variables.

2. “so small in this scenario “ (ssp126): are you affirming that climate change under ssp126 is so small that it doesn’t require modeling? That the literature is wrong when showing that a mere increase from 1.5 to 2 C in GMT would significantly increase climate impacts? Please refrain from writing such sentences, that may be misinterpreted in a very serious manner.

3.“internal variability is likely a larger source of uncertainty than climate change”: climate change itself is an ensemble of physical process, the range that we see around is natural variability, and this is not exactly an uncertainty per se. What is however is the uncertainty from modeling of climate change, which is the one that you refer to. Please, avoid such confusions.

4.“pattern scaling is not even necessary for this scenario”: no, it depends on the climate variable and on the region. Even with a change in by +1.5 or +2C, the regional changes can be up to twice this global increase for the annual mean temperature or precipitations, or even more for some months. One may not simply recycle the historical data for a ssp126.

Lines 318-319:

No, pattern scaling cannot directly be used for diagnosis, this tool aggregates too much the different processes to pinpoint precisely sources for discrepancies.

C.Performances of the patterns

Figures 6 & 7:

They are the main figure to me. It truly shows the performances of pattern scaling. Though, the regional aggregation is too coarse, with the risk of having errors compensating each others. I highly recommend either representing this for AR6 regions instead, or to compute local deviations and average them. But not the deviations on aggregated results.

Minor revisions:

Line 18:

“all four CMIP5 realizations run”: these are not runs, but scenarios. To avoid confusion, a run is the output of an ESM for a scenario and an ensemble member.

Lines 78-104:

Please explicitly say which algorithm is used for fitting.

Line 287:

Could also remind here Tebaldi et al. (2022) (https://doi.org/10.5194/esd-13-1557-2022) and eventually mention Quilcaille et al. (https://doi.org/10.1029/2022GL099012).

Line 313-315:

A fair description of the existing literature would also include efforts on monthly emulations by Nath et al. 2022 (https://doi.org/10.5194/esd-13-851-2022), given monthly patterns trained here

All figures:

Please put grids and (a), (b) … for panels of figures.

Figure 6: please make this figure less painful to read. Separate variables in groups, colors for latitudes, etc.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

Do you want your identity to be public for this peer review? If you choose “no”, your identity will remain anonymous but your review may still be made public.

For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: Yes: Valeriu Predoi

Reviewer #3: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Attachments

Attachment

Submitted filename: review_Kravitz2023.pdf

PCLM-D-22-00161R2

Pangeo-Enabled ESM Pattern Scaling (PEEPS):  A customizable dataset of emulated Earth System Model output

PLOS Climate

Dear Dr. Kravitz,

Thank you for submitting your manuscript to PLOS Climate. After careful consideration, we feel that it has merit but does not fully meet PLOS Climate’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by Oct 06 2023 11:59PM. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at climate@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pclm/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

• A rebuttal letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
• A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
• An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Yangyang Xu

Academic Editor

PLOS Climate

Journal Requirements:

1. Please provide separate figure files in .tif or .eps format.

For more information about figure files please see our guidelines:

https://journals.plos.org/climate/s/figures 

https://journals.plos.org/climate/s/figures#loc-file-requirements

2. We notice that your supplementary figures are uploaded with the file type 'Figure'. Please amend the file type to 'Supporting Information'. Please ensure that each Supporting Information file has a legend listed in the manuscript after the references list.

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: (No Response)

**********

2. Does this manuscript meet PLOS Climate’s publication criteria? Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe methodologically and ethically rigorous research with conclusions that are appropriately drawn based on the data presented.

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: N/A

**********

4. Have the authors made all data underlying the findings in their manuscript fully available (please refer to the Data Availability Statement at the start of the manuscript PDF file)?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS Climate does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: I acknowledge that this manuscript has improved since the last version, but there are still two major problems, on the appropriate uses of these patterns and on the actual performances of the patterns. The point on the appropriate uses was already highlighted in the former round, and was not correctly accounted for. The importance of these points force me to recommend a major revision.

Please find more details in the file attached.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

Do you want your identity to be public for this peer review? If you choose “no”, your identity will remain anonymous but your review may still be made public.

For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

Attachments

Attachment

Submitted filename: AR2_2.pdf

Pangeo-Enabled ESM Pattern Scaling (PEEPS):  A customizable dataset of emulated Earth System Model output

PCLM-D-22-00161R3

Dear Dr. Kravitz,

We are pleased to inform you that your manuscript 'Pangeo-Enabled ESM Pattern Scaling (PEEPS):  A customizable dataset of emulated Earth System Model output' has been provisionally accepted for publication in PLOS Climate.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow-up email from a member of our team. 

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they'll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact climate@plos.org.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Climate.

Best regards,

Yangyang Xu

Academic Editor

PLOS Climate

***********************************************************

Reviewer Comments (if any, and for reference):

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #3: All comments have been addressed

**********

2. Does this manuscript meet PLOS Climate’s publication criteria? Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe methodologically and ethically rigorous research with conclusions that are appropriately drawn based on the data presented.

Reviewer #3: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #3: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available (please refer to the Data Availability Statement at the start of the manuscript PDF file)?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #3: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS Climate does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #3: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #3: This new version of this manuscript has improved further. I note that the authors have correctly accounted for the previous comments. Therefore I consider that this manuscript may be published.

For the sake of discussion on the first main point, the extrapolation of spatial patterns from a scenario to another, I acknowledge that Kravitz et al, 2017 shows that patterns are very similar for 1pctCO2 and RCP8.5, and for precipitations, which isn’t the most straightforward variable. However, it also shows that these differences are stronger between RCP8.5 and RCP2.6. The suggested explanation is the share of non-CO2 forcing. It may also be due to differences between transient and stationary regimes. Both RCP8.5 and 1pctCO2 mostly describing the transient regime, while RCP2.6 describes a more stationary regime on its latter part. Regardless of the reason for these differences, applying patterns on a scenario trained on another scenario could work in CMIP5 if both scenarios were “not too different”. In CMIP6, the issue is the same, but the range in scenarios is even stronger.

During the editing processes, I recommend fixing the references. All of them appeared in my version as “?”. I also note that figures 6 & 7 seem reversed.

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

Do you want your identity to be public for this peer review? If you choose “no”, your identity will remain anonymous but your review may still be made public.

For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #3: No

**********