Minimal impact of model biases on Northern Hemisphere El Niño–Southern Oscillation teleconnections

Tyrrell, Nicholas L.; Karpechko, Alexey Yu.

doi:https://doi.org/10.5194/wcd-2-913-2021

Articles | Volume 2, issue 3

https://doi.org/10.5194/wcd-2-913-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/wcd-2-913-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 2, issue 3

Research article

|

24 Sep 2021

Research article |

| 24 Sep 2021

Minimal impact of model biases on Northern Hemisphere El Niño–Southern Oscillation teleconnections

Nicholas L. Tyrrell and Alexey Yu. Karpechko

Download

Final revised paper (published on 24 Sep 2021)
Supplement to the final revised paper
Preprint (discussion started on 10 Dec 2020)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Review', Anonymous Referee #1, 23 Dec 2020
SC1: 'Comment on wcd-2020-59', Martin King, 05 Jan 2021
- AC2: 'Reply on SC1', Nicholas Tyrrell, 16 Feb 2021
RC2: 'wcd-2020-59 referee comment', Anonymous Referee #2, 10 Jan 2021
AC1: 'Comment on wcd-2020-59', Nicholas Tyrrell, 16 Feb 2021

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Nicholas Tyrrell on behalf of the Authors (16 Feb 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (04 Mar 2021) by Erich Fischer

RR by Anonymous Referee #2 (15 Mar 2021)

RR by Anonymous Referee #1 (07 Apr 2021)

For final publication, the manuscript should be

accepted as is

accepted subject to technical corrections

accepted subject to minor revisions

reconsidered after major revisions

I would be willing to review the revised paper, if the editor considers it necessary.

I am not willing to review the revised paper.

rejected

Suggestions for revision or reasons for rejection

Minimal impact of model biases on northern hemisphere ENSO teleconnections
Nicholas L. Tyrell and Alexey Yu Karpechko

The authors have carried out a set of bias-correction experiments in the ECHAM 6 model investigating the role of basic-state model biases on the teleconnection between ENSO and the NAO. By imposing a climatologically varying set of fixed tendencies to the temperature and divergence fields in the troposphere, stratosphere, or both, they can partially correct model biases in different regions of the model. They then impose tropical pacific SST anomalies modeled after El Nino or La Nina states and analyse the impact on the NAO, considering established pathways for the teleconnection through the stratospheric polar vortex.

They find that while the bias correction does affect the response of the polar vortex to ENSO, the NAO response to ENSO itself is largely unaffected. They conclude from this that, in the model, the teleconnection between ENSO and the NAO does not depend on the stratospheric pathway, and that it is not sensitive to the model biases corrected for by the bias correction methodology employed. Another interesting hypothesis presented is that the response of near-tropopause heat fluxes (indicative of vertically propagating waves) to the Aleutian low may saturate at large amplitudes.

I have a few questions and concerns outlined below; my main criticism is that the authors should discuss their results more quantitatively in the context of the observed teleconnec
tion. The text could also have used another read through to catch typos.

If these concerns are meaningfully addressed the paper should meet the standard for publication.

Major comments

Structure of Biases

I have the same criticism of this revised paper as I did of the discussion paper: the authors discuss the effects of model bias on the ENSO/NAO teleconnection without clearly showing or discussing the detailed structure of the biases. The supplementary figures (1 through 3) are far more meaningful than the profiles of RMSE shown in Figure 1. They should be included in the main text in addition (or in lieu of) the current panels a-f of Figure 1 and the discussion in 3.1 should be rewritten in terms of the structure of the temperature and wind biases that are corrected.

In particular, there are strong cold biases (I assume what is shown is model - reanalyis, though this is not specified) at high latitudes near and above the tropopause evident in suppleental figure 2. This is included in the TropBC run, but the bias may be better thought of as a lower-stratospheric bias. This could be indicative of an extratropical tropopause that is too high in the model. The TropBC run reduces the high latitude temprature anomalies at the expense of introducing cold anomalies in the tropical upper troposphere. The wind biases that are corrected are in thermal wind balance with these anomalies (as expected), but this makes sense of the significant reduction in RMSE found in the mid-latitude stratosphere in TropBC and FullBC.

Structure of ENSO response

The differences between the modeled and observed teleconnections are very substantial, particularly with regards to the stratospheric pathway of the teleconnection. This is material to the central conclusions of the manuscript with regards to the relative importance of the tropospheric pathway in this model: the model is not properly capturing the stratospheric pathway in its basic configuration or in the bias correction runs. This is particularly evident with regards to the response of the heat flux metric (panel 2c), which is highly nonlinear in the observations, with a positive response to either signed tropical Pacific SST anomlies. This leads to a highly nonlinear stratospheric response (panels d and e), and an NAO response with a seasonal structure that more closely resembles the stratospheric seasonal cycle than the seasonal cycle of the tropical divergence response.

Statements made in lines 163 and 207-8 are far too speculative and the hypothesis that this non-linearity is consistent with sampling variability should be quantitatively tested. This can be done by subsampling periods from the model runs as long as the observational record to determine if something like the observational signal could have occurred plausibly by chance within the context of the models variability. Given the large structural differences seen in Fig. 2 this comparison needs to be done more quantitatively.

These differences also need to be explicitly discussed both in section 3 and in the conclusions.

Specific comments:

l 34: This isn't consistent with Figure 2; the NAO response is negative, not positive. Similar non-linearities are seen in HF and stratospheric responses.

l 158: 'almost opposite' this is ambiguous and confusing - the text could be read to mean that the La Nina response is also an increase in HF but with CTRL and StratBC runs showing half as strong a response as Full BC and TropBC.

l 162-3: what is the 'canonical' response here? the observations?

l 273: I find supplemental figure 5 quite helpful in supporting this argument - I think it's worth showing in the main text.

Methodology:

Are the temperature and divergence bias correction tendencies computed from a nudging run in which only temperature and divergence are relaxed, or ar they taken from a run in which all four fields are nudged?

Fig. 4: Since we are comparing magnitudes of responses in different runs, the shading should show a confidence interval (which is relevant for comparison), not a measure of variability (which is not). Typos

l57: affect

l81: nudged

l145 symmetric and actually

l163: response.

l185-187: Please clarify

l190: which effect?

l222: its

l269: closer to observations.

Hide

ED: Publish subject to revisions (further review by editor and referees) (06 May 2021) by Erich Fischer

AR by Nicholas Tyrrell on behalf of the Authors (17 Jun 2021) Author's response Author's tracked changes Manuscript

ED: Referee Nomination & Report Request started (04 Jul 2021) by Erich Fischer

ED: Publish subject to technical corrections (23 Aug 2021) by Erich Fischer

Short summary

Tropical Pacific sea surface temperatures (El Niño) affect the global climate. The Pacific-to-Europe connection relies on interactions of large atmospheric waves with winds and surface pressure. We looked at how mean errors in a climate model affect its ability to simulate the Pacific-to-Europe connection. We found that even large errors in the seasonal winds did not affect the response of the model to an El Niño event, which is good news for seasonal forecasts which rely on these connections.