Further improvement and evaluation of nudging in the E3SM Atmosphere Model version 1 (EAMv1): simulations of the mean climate, weather events, and anthropogenic aerosol effects

Zhang, Shixuan; Zhang, Kai; Wan, Hui; Sun, Jian

doi:https://doi.org/10.5194/gmd-15-6787-2022

Articles | Volume 15, issue 17

https://doi.org/10.5194/gmd-15-6787-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/gmd-15-6787-2022

© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 15, issue 17

Development and technical paper

|

08 Sep 2022

Development and technical paper |

| 08 Sep 2022

Further improvement and evaluation of nudging in the E3SM Atmosphere Model version 1 (EAMv1): simulations of the mean climate, weather events, and anthropogenic aerosol effects

Shixuan Zhang, Kai Zhang, Hui Wan, and Jian Sun

Download

Final revised paper (published on 08 Sep 2022)
Supplement to the final revised paper
Preprint (discussion started on 03 Feb 2022)
Supplement to the preprint

Interactive discussion

Status: closed

RC1:
'Comment on gmd-2022-10', Anonymous Referee #1, 20 Mar 2022

Zhang et al. (2022) improved the EAMv1 by revising the sequence of nudging tendency in the model and increasing the nudging frequency. The manuscript has thoroughly discussed the nudging impacts from several sets of model simulations and provided suggestions for the nudged simulations. The manuscript is well organized and has addressed key issues in the nudged model simulations in EAMv1. Below are a few comments.

General comment

The EAMv1 nudging simulations are improved in this work. My major concern is the model resolution used in this work. The current model is configured to be 1 deg. It does not fully take advantage of reanalysis data with higher spatial resolution, which may dilute the impacts from ERAI and ERA5. On the other hand, if the model is configured to a higher spatial resolution (which is the direction in the global climate community), the model accuracy to simulation important small-scale processes could also be improved. In that case, it may require a different nudging strength or frequency.

Specific comments:

Page 9, Figure 3, typo? No magenta box in Figure 4e. Maybe Figure 2e?

Page 21, line 350-351, there is no Table S3 in the Supplement. Also, I’m confused about the sign between line 351 and in Figure 12. Suugest to clarify the discussion here.

Page 21, line 365-369, this is also a typical approach in many other earth system models.

Citation: https://doi.org/10.5194/gmd-2022-10-RC1
- AC1: 'Reply on RC1', Shixuan Zhang, 30 Apr 2022
  
  We greatly appreciate the reviewer’s comments and suggestions. The point-by-point reply to the reviewer's comments is provided in the attached PDF file.
  
  Citation: https://doi.org/10.5194/gmd-2022-10-AC1
RC2:
'Comment on gmd-2022-10', William Collins, 21 Mar 2022

This was a clear study that successfully explained some biases in the previous nudging schemes. The comments on time resolution will be useful to the community which generally used 6-hourly data.

For the aerosol forcing why is 6-hour nudging used when the authors have shown that 3-hour is preferred? This experiment should be repeated with the 3-hour resolution. The authors could explain the issues with the temperature nudging more fully. The effects are attributed to biases. While this could be true for ERA, this should be much less significant for CLIM (fig 2(e)) and would be even less so for 3-hourly (in fig 4(d)). Is this actually removing a meteorological adjustment to the aerosols?

Specific points:

Figure 1: Why is the nudging tendency applied at a different place in the sequence to where it is calculated?

Line 154: Please define how CF is calculated.

Figure 3: This is a useful figure that nicely explains the effects.

Line 345: The differences in emissions between 1850 and present should be detailed here or in the supplement.

364: It is not obvious why there should be biases when nudging to CLIM. However if EAMv1 has notable biases then it is not clear why improving the temperatures by nudging to ERA5 doesn’t give a more physical measure of the aerosol forcing than using the biased temperatures.

Citation: https://doi.org/10.5194/gmd-2022-10-RC2
- AC2: 'Reply on RC2', Shixuan Zhang, 30 Apr 2022
  
  We greatly appreciate the reviewer’s comments and suggestions. The point-by-point reply to the reviewer's comments is provided in the attached PDF file.
  
  Citation: https://doi.org/10.5194/gmd-2022-10-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

AR by Shixuan Zhang on behalf of the Authors (26 Jun 2022) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (26 Jul 2022) by Fiona O'Connor

RR by William Collins (26 Jul 2022)

Suggestions for revision or reasons for rejection

I wish to thank the authors for their careful consideration of my comments, particularly to my question on temperature nudging. Their response and the text lines 400-430 shows that there are two effects – a mean-state correction in ERA5, and a suppression of adjustments in CLIM PD. It would be helpful to explicitly explain it in these terms.

The mean-state correction is presumably a more realistic calculation of the ice-cloud forcing so it is not obvious that ignoring this correction should be the preferred calculation (which line 433 states). Line 22 refers to the temperature-nudged calculation as “a slightly biased estimate”. However figure 13 shows that the ERA5 temperature nudging actually removes a systematic bias, so it could be argued that this is “a less biased estimate”. The problem is that ERA5 temperature nudging also removes the temperature adjustment.

The calculation of the temperature-mediated adjustment is +24% in FSNT (i.e. from -1.862 to -2.453 W/m2 in table 3), but smaller (+8%) in the net. This is a useful scientific point that helps our understanding of the overall radiative effects of aerosols and so should be mentioned in section 5 and also in the conclusions, for instance the adjustment would be absent from an offline calculation and from the classical split into “direct” and “albedo” effects.

I suggest lines 431-435 are rewritten to explain that nudging to ERA5 might remove a mean-state bias, but because it also supresses a temperature-mediated adjustment it is not the preferred estimate. Similar wording changes are needed in the abstract lines 20-22.

Line 125: This needs a short explanation to say that Faer is the Effective Radiative Forcing including all meteorological adjustments (citing for instance AR5 or AR6) defined using fixed-SSTs. You can then refer back to this explanation when you find suppression of adjustments in section 5.

Line 173: Please explain how CF is calculated. Is it a double-call model diagnostic with clear and cloudy-sky, or is it an offline calculation?

Line 388 and table S3. The sign convention for the LW forcing should be explained. I assume it is downward. It might be clearer to use the same direction (outgoing) for both SW and LW.

Hide

ED: Publish subject to technical corrections (26 Jul 2022) by Fiona O'Connor

Dear Shixuan and co-authors,

You will see that I asked one of the reviewers to re-review your manuscript, given that they originally had some major comments. Below, you will see their most recent comments. May I please ask that you revise your manuscript based on these new comments. Once completed, I will be happy to recommend publication.

Kind regards,
Fiona O'Connor

-----
Reviewer comments:

I wish to thank the authors for their careful consideration of my comments, particularly to my question on temperature nudging. Their response and the text lines 400-430 shows that there are two effects – a mean-state correction in ERA5, and a suppression of adjustments in CLIM PD. It would be helpful to explicitly explain it in these terms.

The mean-state correction is presumably a more realistic calculation of the ice-cloud forcing so it is not obvious that ignoring this correction should be the preferred calculation (which line 433 states). Line 22 refers to the temperature-nudged calculation as “a slightly biased estimate”. However figure 13 shows that the ERA5 temperature nudging actually removes a systematic bias, so it could be argued that this is “a less biased estimate”. The problem is that ERA5 temperature nudging also removes the temperature adjustment.

The calculation of the temperature-mediated adjustment is +24% in FSNT (i.e. from -1.862 to -2.453 W/m2 in table 3), but smaller (+8%) in the net. This is a useful scientific point that helps our understanding of the overall radiative effects of aerosols and so should be mentioned in section 5 and also in the conclusions, for instance the adjustment would be absent from an offline calculation and from the classical split into “direct” and “albedo” effects.

I suggest lines 431-435 are rewritten to explain that nudging to ERA5 might remove a mean-state bias, but because it also supresses a temperature-mediated adjustment it is not the preferred estimate. Similar wording changes are needed in the abstract lines 20-22.

Line 125: This needs a short explanation to say that Faer is the Effective Radiative Forcing including all meteorological adjustments (citing for instance AR5 or AR6) defined using fixed-SSTs. You can then refer back to this explanation when you find suppression of adjustments in section 5.

Line 173: Please explain how CF is calculated. Is it a double-call model diagnostic with clear and cloudy-sky, or is it an offline calculation?

Line 388 and table S3. The sign convention for the LW forcing should be explained. I assume it is downward. It might be clearer to use the same direction (outgoing) for both SW and LW.

Hide

AR by Shixuan Zhang on behalf of the Authors (13 Aug 2022) Author's response Manuscript

Short summary

This study investigates the nudging implementation in the EAMv1 model. We find that (1) revising the sequence of calculations and using higher-frequency constraining data to improve the performance of a simulation nudged to EAMv1’s own meteorology, (2) using the relocated nudging tendency and 3-hourly ERA5 reanalysis to obtain a better agreement between nudged simulations and observations, and (3) using wind-only nudging are recommended for the estimates of global mean aerosol effects.