Improved ENSO and PDO Prediction Skill Resulting from Finer Parameterization Schemes in a CGCM

Round 1

Reviewer 1 Report

The Paper “Improved ENSO and PDO Prediction Skill Resulting from Finer Parameterization Schemes in FIO-CPS v2.0” studies the prediction skill of ENSO and PDO by the modified FIO-CPS v2.0 which includes four new parameterization schemes. This study is very important for the improving of CGCM and very useful for better prediction of the ENSO and PDO. This paper is written and organized properly, and the results will be beneficial to future relevant research. Overall, I think the manuscript is well written and recommend acceptance after some minor changes.

1.      In the “data and Methods”, the presentation of PDO index is not clear, more detail should be given. For example, the low pass filter is used in this study or not? The analysis of PDO is used the data of monthly mean or annual mean?

2.      In the Figure 1, the shading in the figure can not have been separated in the tropical region, it should be modified.

3.      Figure capture can’t be completely accurate, for example, there is no describe of unit in Figure .2, 3, 4, 5.

4.      Some English expressions are repeated and unclear. Please revise the paper. For example, Line 300-302.

5.      In the discussion of possible reasons for prediction skill improvement, the prediction of SST is investigated. So what about the mix layer?

Author Response

Response to Reviewer 1 Comments

We thank the reviewers for their careful and thorough evaluation of our manuscript and for their helpful comments and constructive criticism, which have helped improve the quality of this manuscript.

The Paper “Improved ENSO and PDO Prediction Skill Resulting from Finer Parameterization Schemes in FIO-CPS v2.0” studies the prediction skill of ENSO and PDO by the modified FIO-CPS v2.0 which includes four new parameterization schemes. This study is very important for the improving of CGCM and very useful for better prediction of the ENSO and PDO. This paper is written and organized properly, and the results will be beneficial to future relevant research. Overall, I think the manuscript is well written and recommend acceptance after some minor changes.

1. In the “data and Methods”, the presentation of PDO index is not clear, more detail should be given. For example, the low pass filter is used in this study or not? The analysis of PDO is used the data of monthly mean or annual mean?

Response: We have added the definition of PDO in modes in detail. It is defined by the first mode of Empirical orthogonal function decomposition in CGCMs. The low pass filter has been done on PDO index.

2. In the Figure 1, the shading in the figure can not have been separated in the tropical region, it should be modified.

Response: We have modified all the figures in high resolution, especially the Fig. 1.

3. Figure capture can’t be completely accurate, for example, there is no describe of unit in Figure .2, 3, 4, 5.

Response: We have rewritten the figure captions, the units have been added to the Figure .2.

4. Some English expressions are repeated and unclear. Please revise the paper. For example, Line 300-302.

Response: The sentence in Line 300-302 “With the increasing lead time, prediction errors grow rapidly, and this is especially true within the tropics as both original and modified models show errors in the tropical Pacific Ocean” has been changed into “With the increasing lead time, prediction errors grow rapidly, especially in the tropical Pacific Ocean”.

5. In the discussion of possible reasons for prediction skill improvement, the prediction of SST is investigated. So what about the mix layer?

Response: We have added the results of mixed layer in this part. Please see the detail in Paper in the part of “3.3 Possible Reasons for Prediction Skill Improvement”, last paragrph.

Author Response File: Author Response.doc

Reviewer 2 Report

Four model parameterization schemes in FIO-CPS v2.0 are described and evaluated for the prediction skills of ENSO and PDO in lead time from month to 1 year.  The methods and results of the manuscript are clearly presented in the manuscript. The model improvement by added parameterization is always a welcome, although it is small, not consistent on all time scales and location, and less understanding of mechanisms.  I recommend its publication with some revision.

The authors could add discussion on the improvement by each of the four parameterizations separately and collectively. This might be important because the aim of the manuscript is to show readers how the parameterizations are beneficial to the prediction skills of ENSO and PDO, and also how the different parameterization schemes might contribute and interact in a sophisticated climate model.

The improvements are different with different lead times, it is useful to add discussion on which time scale is most important and mostly used in the actual forecast of ENSO and PDO?

In Section 3, The reasoning of the skill improvement is relatively simple and weak, and more in-depth analysis is expected, especially combined with discussion of each parameterization separately.

English needs to be polished and check on typo. Some are pointed out here:

L54 maybe: which increase the prediction skill of ENSO

L271 reads not right

L296, which line represents the observation?

L310 significantly smaller

Resolution of Figure 1 and Figure 7 can be improved.

Author Response

Response to Reviewer 2 Comments

Firstly, we greatly appreciate your constructive suggestions and careful comments on our manuscript, and it is very valuable for improving our manuscripts. We have carefully revised our manuscript according to your suggestions. A point-to-point response to your major comments is provided below.

Four model parameterization schemes in FIO-CPS v2.0 are described and evaluated for the prediction skills of ENSO and PDO in lead time from month to 1 year.  The methods and results of the manuscript are clearly presented in the manuscript. The model improvement by added parameterization is always a welcome, although it is small, not consistent on all time scales and location, and less understanding of mechanisms.  I recommend its publication with some revision.

The authors could add discussion on the improvement by each of the four parameterizations separately and collectively. This might be important because the aim of the manuscript is to show readers how the parameterizations are beneficial to the prediction skills of ENSO and PDO, and also how the different parameterization schemes might contribute and interact in a sophisticated climate model.

Response: We have added discussion on the four paramererizations being beneficial to the prediction skills of ENSO and PDO. These parameterization schemes involves two important energy sources (wind and tide) and four physical processes for mixing, help to improve the simulation of mean state by describing the mixing in surface and deep ocean more realistically. Accurate simulation of SST and MLD is key for prediction skill of ENSO and PDO. At the air-sea interface, ocean surface waves play an important role in momentum, energy and gas exchange. The wave-induced mixing improves the excessively high surface temperature simulated in summer and reduces the underestimation of the mixed layer depth in winter as demonstrated by Wang et al. (2022). For improving the SST and MLD simulation, another important submesoscale parameterization scheme (SI parameterization) is included, the primary effect of symmetric instability is the shoaling of the mixed layer. The mixing between mixing layer and ocean interior is driven by internal waves. So, both internal tide and Lee-wave parameterization is considered in present study. The revised internal tidal mixing parameterization introduc a correction function related to the Coriolis effect on internal tidal local dissipation in the deep ocean based on original version developed by St. Laurent et al. [57], which is more resonalable via demonstration in laboratory. The Lee-wave parameterization mainly improves the simulations of deep sea water temperature and salinity in seamount-rich areas, and it is a weak contributor for improvements of ENSO and PDO prediction skill. In general, the four parameterization schemes separately present the four important physical process from surface to interior, which has improved the simulation of ocean state to varying degrees, respectively, as demonstrated by published literatures (Wang et al.2022, Dong et al. 2020, Cao et al., 2022, and Tian et al., 2022). The wave-induced mixing and SI parameterization directly improve the simulation of SST and MLD, and revised internal tide and Lee wave parameterizaiton have constriburion for improving the backeground state simulation. 

The improvements are different with different lead times, it is useful to add discussion on which time scale is most important and mostly used in the actual forecast of ENSO and PDO?

Response: we have added some discussion about the improvement with different lead time. For the forecasting of ENSO, with the longer lead time, the improvement of RSME is more significant, especially at 4-, 5- and 6-month lead time. As we know, it is very important that the improvement at longer lead time. One is because with longer lead time, the forecasting is more difficult, the other is because at 6-month lead time forecasting can be benefit for acrossing the SPB which is a main barrier of seasonal prediction of ENSO. For the forecasting of PDO, with the increasing of leading time, the prediction skill is de-ceasing very quickly. That may be because the longer lead time prediction is related to the initialization and external radiative forcing [69].

69. Choi, J.; Son, S.-W. Seasonal-to-decadal prediction of El Niño–Southern Oscillation and Pacific Decadal Oscillation. npj Climate and Atmospheric Science 2022 5:29 ; https://doi.org/10.1038/s41612-022-00251-9

In Section 3, The reasoning of the skill improvement is relatively simple and weak, and more in-depth analysis is expected, especially combined with discussion of each parameterization separately.

Response: Good suggestion! Four parameterizaitons are combined together to insert into model when predic model runing. So it is difficult for analysing the direct effection of each parameterization sepatately from the present model output. As stated above, we have added the discussion on the four paramererizations being beneficial to the prediction skills of ENSO and PDO in Section 4. We hope more experiments have been done for evaluated the effection of each parameterizaiton separately in the furture.

English needs to be polished and check on typo. Some are pointed out here:

L54 maybe: which increase the prediction skill of ENSO

Response: this sentence has been change to “Although there are significant improvements of lead time and accuracy in prediction of ENSO, several challenges still exist”.

L271 reads not right

Response: The sentence in this line has been changed to “Figure 4 shows the forecast skill that is calculated the RMSE and ACC of Niño3.4 index of the predicted target month and lead time”.

L296, which line represents the observation?

Response: the figure shows the correlation coefficient root mean square error and between observation and models.

L310 significantly smaller

Response: In this line, the “significant smaller” has been change to “significantly smaller”.

Resolution of Figure 1 and Figure 7 can be improved.

Response: The resolution of all figures has been improved.

Author Response File: Author Response.docx