the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Nov 2024
The Tropical Basin Interaction Model Intercomparison Project (TBIMIP)
Abstract. Large-scale interaction among the three tropical ocean basins is an area of intense research that is often conducted through experimentation with numerical models. A common problem is that modelling groups use different experimental setups, which makes it difficult to compare results and to delineate the role of model biases from differences in experimental setups. To address this issue, an experimental protocol for examining interaction among the tropical basins is introduced. The tropical basin interaction model intercomparison project (TBIMIP) consists of experiments in which sea surface temperatures (SSTs) are prescribed to follow observed values in selected basins. There are two types of experiments. One type, called standard pacemaker, consists of simulations in which SSTs are restored to observations in selected basins during a historical simulation. The other type, called pacemaker hindcast, consists of seasonal hindcast simulations in which SSTs are restored to observations during the forecast. TBIMIP is coordinated by the Climate and Ocean - Variability, Predictability, and Change (CLIVAR) Research Focus on Tropical Basin Interaction. The datasets from the model simulations will be made available to the community to facilitate and stimulate research on tropical basin interaction and its role in seasonal-to-decadal variability and climate change.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(2589 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2589 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-3110', Michael Alexander, 21 Dec 2024
This paper describes the i) rational, ii) experiment design/main model simulations, iii) use of observations and a hierarchy of models in addition to the main set of runs, and iv) potential outcomes and issues of a sponsored MIP focused on interactions between the tropical ocean basins. The CMIP protocol encourages (requires?) a paper describing the experiment design and this manuscript well describes the planned simulations. I have just a few suggestions and thus recommend a minor revision.
Comments:
1) The manuscript could include a few figures from observations or previous experiments illustrating potential interactions and hypotheses to be explored, in addition to the schematic shown in Fig. 1.
2) For many in the oceanography community “hindcast” is used to describe long simulations driven by atmospheric reanalysis (and ocean reanalyses) for regional models. (This is called a historical simulation here.) You might choose to use “re-forecasts” instead of “hindcasts” or add a sentence or two explaining how “hindcast” is being used in this context.
3) Will the tapering method as a function of latitude (e.g., linear decrease with latitude) be prescribed to be the same across all experiments?
4) Can an explanation be provided for why the start of the tapering latitude is different in the Atlantic compared with the other two basins.
5) lines 217-218: States: “The technique for initializing the hindcasts (data assimilation etc.) is left to the modelling groups.” This could lead to major differences between the hindcasts (re-forecasts) especially in the first couple of months. Perhaps some tests with a single modeling system could be performed to investigate how much different initialization methods influence the forecast spread and perhaps how long it took for initialization differences not to have a notable influence on the re-forecasts (in a probabilistic sense).
6) Lines 250-260 state:
"The top ocean level interacts with the atmospheric model component through a coupler routine (e.g., Craig et al. 2017), which regulates the exchange of fluxes between the atmosphere and ocean. Another approach for modifying SSTs is therefore through manipulating inside the coupler routine the heat flux that goes into the ocean, which is the method recommended for the TBIMIP experiments. The heat flux in tropical regions consists of four components: net surface shortwave radiation, net surface longwave radiation, latent heat flux, and sensible heat flux. Of these, the sensible heat flux is usually chosen for manipulation (e.g., Kosaka and Xie 2013), and this is the method recommended for TBIMIP. Finally, because the flux coupler controls the SSTs that are “seen” by the atmospheric component, one can modify only this value, thereby “tricking” the atmosphere into reacting to a temperature that is different from the actual ocean SST. This approach leaves the ocean component completely unchanged (Richter and Doi 2019). Furthermore, it allows the SSTs to exactly follow a given distribution (as far as the atmosphere is concerned), rather than approximating it through correction terms. A potential drawback is that this can lead to very unrealistic heat fluxes into the atmosphere (Wang et al. 2005)."
And then on lines 281-282:
"Because the heat flux is absorbed in the top layer first, the immediate temperature response could lead to unrealistic changes in vertical stability"
These two statements seem contradictory, the top implying that you are not actually changing the ocean but just tricking it to see the altered state and the latter indicating an actual change in the ocean. Please clarify.
7) Lines 359-361: State “The curves essentially collapse into one, suggesting that the bias of a given model is mostly time-invariant. We conclude that using a shorter base period should not lead to major imbalances though this should be carefully evaluated for each model.”
It may be worth exploring the results described in the paper:
Beverley, J.D., Newman, M. & Hoell, A. Climate model trend errors are evident in seasonal forecasts at short leads. npj Clim Atmos Sci 7, 285 (2024). https://doi.org/10.1038/s41612-024-00832-w
8) Additional Tier 3 Experiments. The paper discusses a number of potential Tier 3 (optional) experiments using a hierarchy of models. Several of the proposed experiments are interesting and could be run relatively inexpensively. Here are some additional ones the project could consider:
- Use LIM or other methods to remove ENSO’s (or other modes) impact on the observed SST anomalies in the other basins. The SST anomalies that are damped towards would remove this impact on the SST anomalies in the other basins and use those adjusted anomalies in either the historical or hindcast simulations. For example, the impact of ENSO on the tropical Atlantic could be estimated from observations and that part of the anomaly signal removed from the observed SST anomalies that are used in the TBI-PACE-AANOM experiment.
- Specify the observed winds or wind anomalies added to the model’s climatological winds in the forcing regions rather than the SST (or SSTA). Since the oceans are primarily driven by winds in the tropics, both by the surface heat fluxes and dynamics (Ekman, upwelling, etc.). This might reduce or nearly eliminate the heat imbalance by relaxing the heat into the ocean (although other issues might arise). A similar experiment design was used in
Ding, H., R. J. Greatbatch, M. Latif, W. Park, and R. Gerdes, 2013: Hindcast of the 1976/77 and 1998/99 Climate Shifts in the Pacific. J. Climate, 26, 7650–7661, https://doi.org/10.1175/JCLI-D-12-00626.1.
- Base the temperature restoring term on the anomalous heat flux convergence in the ocean obtained from ocean reanalyses to estimate the ocean driven SST variability that is communicated to other basins.
Minor comments:
1) line 46: I suggest not using the colloquial expression “players” on line 46. Perhaps “processes” instead.
2) Lines 150-151: Suggest changing “a wealth of intercomparisons has been performed” to “a wide-range of intercomparisons have been performed”
Citation: https://doi.org/10.5194/egusphere-2024-3110-RC1 -
AC1: 'Reply to reviewer #1', Ingo Richter, 05 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-3110/egusphere-2024-3110-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-3110', Anonymous Referee #2, 21 Dec 2024
The atmosphere has no lateral boundaries, allowing climate variability to be coupled and interact between ocean basins. Such tropical ocean basin interactions (TBI) are recognized to be important for understanding and predicting tropical variability including El Nino/the Southern Oscillation (ENSO). This paper motivates and describes the protocol for a TBI MIP, as an outcome of the CLIVAR TBI Research Focus. The review of relevant literature (sections 1 and 6) is fun to read. The paper is well written and should be published with minor revision.
The proposed pacemaker runs are designed to explore TBI via the atmospheric Walker circulation with SST restoration applied to narrow equatorial oceans (Fig. 3). Here are some slight changes in the domain(s) of SST restoration that could allow other modes of TBI.
- Western North Pacific. Recent studies identified a coupled Indo-western Pacific ocean-atmosphere mode in post-ENSO summer (JJA). Its SST signature is also known as, but more complex than, the IOB mode, while the atmospheric component features a large-scale anomalous anticyclone (AAC) that covers the entire Indo-western Pacific north of the equator. A tropical Pacific domain with a wedge that reaches all the way to the maritime continent on the equator but avoids much of the off-equatorial western Pacific would be able to capture AAC and Asian summer monsoon variability/predictability (P. Zhang et al. 2024, J Climate). Climatically, the South China Sea is in the monsoon westerly regime that includes the North Indian Ocean (e.g., Fig. 2 of Zhang et al. 2024. Thus, the South China Sea should be part of the Indian Ocean, rather than the Pacific, for SST restoring.
- In the Atlantic, the proposed SST restoration is limited to a narrow 10S-10N band. This could miss the subtropical atmospheric Rossby wave pathway connecting the broad tropical North Atlantic with the subtropical Northeast Pacific and then ENSO through WES (Ham et al. 2013a,b, cited in the paper) and low cloud (A. Miyamoto 2025, J Clim) feedback mechanisms.
The TBI team might want to consider these slightly modified configurations as possible extensions of the current TBIMIP.
Minor comments
There are two Chang et al. (2006) references, which may need to be labeled as a, b.
Xie and Carton (2004) not in References.
L289. “subtropical” --> “subpolar”?
Citation: https://doi.org/10.5194/egusphere-2024-3110-RC2 -
AC2: 'Reply to reviewer #2', Ingo Richter, 05 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-3110/egusphere-2024-3110-AC2-supplement.pdf
-
EC1: 'Comment on egusphere-2024-3110', Penelope Maher, 08 Jan 2025
Dear Dr Richter,
Thanks for your submission to GMD. Both reviewers have recommended minor revision and I support their recommendations. In addition to the comments from Reviewer 1 and 2 I would like to add two comment and some simple editing suggestions.
In papers which describe the methodology of a new MIP, all the relevant technical details should be in the manuscript or supplementary. I would recommend you include all relevant technical information in the manuscript, and any additional information which is helpful be added to the supplementary (as it currently stands there is repetition between the full description and the manuscript). This will help people who may wish to contribute to the MIP. Further, this means the full method has undergone peer review, rather than the archive which has not technically been peer reviewed as part of the manuscript.
Tier 3 experiments in Table 1. There was a bit of a disconnect here between the manuscript (which does not explain Tier 3 experiments other than to say they are planned) and the full description in the archive (which suggests possible experiments). Personally I think it would be clearer to only describe Tiers 1-2 in the table and then have a section description of possible directions for further work for the Tier 3 experiments (which would presumably be part of TBIMIP2 or are you planning as part of TRIMIP? ).
Optional editing to the manuscript
- Fig 1: please include units (it is obvious but best practice to include).
- L99: The proxy record are important parts of the observational record and I find the comment "This makes paleo proxies subject to uncertainties and inhomogeneities" a little dismissive. All datasets and models we use are subject to uncertainties so perhaps rewording to ", which can contribute to uncertainties."?
- Table 1: The ESGF database use lowercase for all experiment names. I suggest you follow this approach in the manuscript.
- Take care with statements like the following which may not get updated and are therefore not encouraged: "This is IPRC publication X and SOEST contribution Y."I look forward to receiving your revised manuscript.
Regards,
Penny
Citation: https://doi.org/10.5194/egusphere-2024-3110-EC1 -
AC3: 'Reply to editor', Ingo Richter, 05 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-3110/egusphere-2024-3110-AC3-supplement.pdf
-
AC3: 'Reply to editor', Ingo Richter, 05 Feb 2025
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-3110', Michael Alexander, 21 Dec 2024
This paper describes the i) rational, ii) experiment design/main model simulations, iii) use of observations and a hierarchy of models in addition to the main set of runs, and iv) potential outcomes and issues of a sponsored MIP focused on interactions between the tropical ocean basins. The CMIP protocol encourages (requires?) a paper describing the experiment design and this manuscript well describes the planned simulations. I have just a few suggestions and thus recommend a minor revision.
Comments:
1) The manuscript could include a few figures from observations or previous experiments illustrating potential interactions and hypotheses to be explored, in addition to the schematic shown in Fig. 1.
2) For many in the oceanography community “hindcast” is used to describe long simulations driven by atmospheric reanalysis (and ocean reanalyses) for regional models. (This is called a historical simulation here.) You might choose to use “re-forecasts” instead of “hindcasts” or add a sentence or two explaining how “hindcast” is being used in this context.
3) Will the tapering method as a function of latitude (e.g., linear decrease with latitude) be prescribed to be the same across all experiments?
4) Can an explanation be provided for why the start of the tapering latitude is different in the Atlantic compared with the other two basins.
5) lines 217-218: States: “The technique for initializing the hindcasts (data assimilation etc.) is left to the modelling groups.” This could lead to major differences between the hindcasts (re-forecasts) especially in the first couple of months. Perhaps some tests with a single modeling system could be performed to investigate how much different initialization methods influence the forecast spread and perhaps how long it took for initialization differences not to have a notable influence on the re-forecasts (in a probabilistic sense).
6) Lines 250-260 state:
"The top ocean level interacts with the atmospheric model component through a coupler routine (e.g., Craig et al. 2017), which regulates the exchange of fluxes between the atmosphere and ocean. Another approach for modifying SSTs is therefore through manipulating inside the coupler routine the heat flux that goes into the ocean, which is the method recommended for the TBIMIP experiments. The heat flux in tropical regions consists of four components: net surface shortwave radiation, net surface longwave radiation, latent heat flux, and sensible heat flux. Of these, the sensible heat flux is usually chosen for manipulation (e.g., Kosaka and Xie 2013), and this is the method recommended for TBIMIP. Finally, because the flux coupler controls the SSTs that are “seen” by the atmospheric component, one can modify only this value, thereby “tricking” the atmosphere into reacting to a temperature that is different from the actual ocean SST. This approach leaves the ocean component completely unchanged (Richter and Doi 2019). Furthermore, it allows the SSTs to exactly follow a given distribution (as far as the atmosphere is concerned), rather than approximating it through correction terms. A potential drawback is that this can lead to very unrealistic heat fluxes into the atmosphere (Wang et al. 2005)."
And then on lines 281-282:
"Because the heat flux is absorbed in the top layer first, the immediate temperature response could lead to unrealistic changes in vertical stability"
These two statements seem contradictory, the top implying that you are not actually changing the ocean but just tricking it to see the altered state and the latter indicating an actual change in the ocean. Please clarify.
7) Lines 359-361: State “The curves essentially collapse into one, suggesting that the bias of a given model is mostly time-invariant. We conclude that using a shorter base period should not lead to major imbalances though this should be carefully evaluated for each model.”
It may be worth exploring the results described in the paper:
Beverley, J.D., Newman, M. & Hoell, A. Climate model trend errors are evident in seasonal forecasts at short leads. npj Clim Atmos Sci 7, 285 (2024). https://doi.org/10.1038/s41612-024-00832-w
8) Additional Tier 3 Experiments. The paper discusses a number of potential Tier 3 (optional) experiments using a hierarchy of models. Several of the proposed experiments are interesting and could be run relatively inexpensively. Here are some additional ones the project could consider:
- Use LIM or other methods to remove ENSO’s (or other modes) impact on the observed SST anomalies in the other basins. The SST anomalies that are damped towards would remove this impact on the SST anomalies in the other basins and use those adjusted anomalies in either the historical or hindcast simulations. For example, the impact of ENSO on the tropical Atlantic could be estimated from observations and that part of the anomaly signal removed from the observed SST anomalies that are used in the TBI-PACE-AANOM experiment.
- Specify the observed winds or wind anomalies added to the model’s climatological winds in the forcing regions rather than the SST (or SSTA). Since the oceans are primarily driven by winds in the tropics, both by the surface heat fluxes and dynamics (Ekman, upwelling, etc.). This might reduce or nearly eliminate the heat imbalance by relaxing the heat into the ocean (although other issues might arise). A similar experiment design was used in
Ding, H., R. J. Greatbatch, M. Latif, W. Park, and R. Gerdes, 2013: Hindcast of the 1976/77 and 1998/99 Climate Shifts in the Pacific. J. Climate, 26, 7650–7661, https://doi.org/10.1175/JCLI-D-12-00626.1.
- Base the temperature restoring term on the anomalous heat flux convergence in the ocean obtained from ocean reanalyses to estimate the ocean driven SST variability that is communicated to other basins.
Minor comments:
1) line 46: I suggest not using the colloquial expression “players” on line 46. Perhaps “processes” instead.
2) Lines 150-151: Suggest changing “a wealth of intercomparisons has been performed” to “a wide-range of intercomparisons have been performed”
Citation: https://doi.org/10.5194/egusphere-2024-3110-RC1 -
AC1: 'Reply to reviewer #1', Ingo Richter, 05 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-3110/egusphere-2024-3110-AC1-supplement.pdf
-
RC2: 'Comment on egusphere-2024-3110', Anonymous Referee #2, 21 Dec 2024
The atmosphere has no lateral boundaries, allowing climate variability to be coupled and interact between ocean basins. Such tropical ocean basin interactions (TBI) are recognized to be important for understanding and predicting tropical variability including El Nino/the Southern Oscillation (ENSO). This paper motivates and describes the protocol for a TBI MIP, as an outcome of the CLIVAR TBI Research Focus. The review of relevant literature (sections 1 and 6) is fun to read. The paper is well written and should be published with minor revision.
The proposed pacemaker runs are designed to explore TBI via the atmospheric Walker circulation with SST restoration applied to narrow equatorial oceans (Fig. 3). Here are some slight changes in the domain(s) of SST restoration that could allow other modes of TBI.
- Western North Pacific. Recent studies identified a coupled Indo-western Pacific ocean-atmosphere mode in post-ENSO summer (JJA). Its SST signature is also known as, but more complex than, the IOB mode, while the atmospheric component features a large-scale anomalous anticyclone (AAC) that covers the entire Indo-western Pacific north of the equator. A tropical Pacific domain with a wedge that reaches all the way to the maritime continent on the equator but avoids much of the off-equatorial western Pacific would be able to capture AAC and Asian summer monsoon variability/predictability (P. Zhang et al. 2024, J Climate). Climatically, the South China Sea is in the monsoon westerly regime that includes the North Indian Ocean (e.g., Fig. 2 of Zhang et al. 2024. Thus, the South China Sea should be part of the Indian Ocean, rather than the Pacific, for SST restoring.
- In the Atlantic, the proposed SST restoration is limited to a narrow 10S-10N band. This could miss the subtropical atmospheric Rossby wave pathway connecting the broad tropical North Atlantic with the subtropical Northeast Pacific and then ENSO through WES (Ham et al. 2013a,b, cited in the paper) and low cloud (A. Miyamoto 2025, J Clim) feedback mechanisms.
The TBI team might want to consider these slightly modified configurations as possible extensions of the current TBIMIP.
Minor comments
There are two Chang et al. (2006) references, which may need to be labeled as a, b.
Xie and Carton (2004) not in References.
L289. “subtropical” --> “subpolar”?
Citation: https://doi.org/10.5194/egusphere-2024-3110-RC2 -
AC2: 'Reply to reviewer #2', Ingo Richter, 05 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-3110/egusphere-2024-3110-AC2-supplement.pdf
-
EC1: 'Comment on egusphere-2024-3110', Penelope Maher, 08 Jan 2025
Dear Dr Richter,
Thanks for your submission to GMD. Both reviewers have recommended minor revision and I support their recommendations. In addition to the comments from Reviewer 1 and 2 I would like to add two comment and some simple editing suggestions.
In papers which describe the methodology of a new MIP, all the relevant technical details should be in the manuscript or supplementary. I would recommend you include all relevant technical information in the manuscript, and any additional information which is helpful be added to the supplementary (as it currently stands there is repetition between the full description and the manuscript). This will help people who may wish to contribute to the MIP. Further, this means the full method has undergone peer review, rather than the archive which has not technically been peer reviewed as part of the manuscript.
Tier 3 experiments in Table 1. There was a bit of a disconnect here between the manuscript (which does not explain Tier 3 experiments other than to say they are planned) and the full description in the archive (which suggests possible experiments). Personally I think it would be clearer to only describe Tiers 1-2 in the table and then have a section description of possible directions for further work for the Tier 3 experiments (which would presumably be part of TBIMIP2 or are you planning as part of TRIMIP? ).
Optional editing to the manuscript
- Fig 1: please include units (it is obvious but best practice to include).
- L99: The proxy record are important parts of the observational record and I find the comment "This makes paleo proxies subject to uncertainties and inhomogeneities" a little dismissive. All datasets and models we use are subject to uncertainties so perhaps rewording to ", which can contribute to uncertainties."?
- Table 1: The ESGF database use lowercase for all experiment names. I suggest you follow this approach in the manuscript.
- Take care with statements like the following which may not get updated and are therefore not encouraged: "This is IPRC publication X and SOEST contribution Y."I look forward to receiving your revised manuscript.
Regards,
Penny
Citation: https://doi.org/10.5194/egusphere-2024-3110-EC1 -
AC3: 'Reply to editor', Ingo Richter, 05 Feb 2025
The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2024/egusphere-2024-3110/egusphere-2024-3110-AC3-supplement.pdf
-
AC3: 'Reply to editor', Ingo Richter, 05 Feb 2025
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 459 | 111 | 53 | 623 | 13 | 17 |
- HTML: 459
- PDF: 111
- XML: 53
- Total: 623
- BibTeX: 13
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Ingo Richter
Ping Chang
Gokhan Danabasoglu
Dietmar Dommenget
Guillaume Gastineau
Takahito Kataoka
Noel Keenlyside
Fred Kucharski
Yuko Okumura
Wonsun Park
Malte Stuecker
Andrea Taschetto
Chunzai Wang
Stephen Yeager
Sang-Wook Yeh
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(2589 KB) - Metadata XML