the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 07 Oct 2025
Comparing the seasonal predictability of the Tropical Pacific variability in EC-Earth3 at two horizontal resolutions
Abstract. Seasonal predictability is an active field of research given its strong potential to guide decision-making in many societal and economic sectors. In this study, we compare the predictive skill of the climate model EC-Earth3 at two different horizontal resolutions. The standard resolution – SR – (high resolution – HR) is of around 70 (40) km in the atmosphere and 100 (25) km in the ocean. Both forecast systems are initialised in the same way in May and cover the period 1990–2015, with a forecast period of 8 months. We focus on the Tropics, and particularly on El Nino Southern Oscillation (ENSO), the main source of predictability at seasonal timescales. Statistically significant improvements are found in HR with respect to SR for predicting ENSO. However, the predictive skill drops quickly in the Western Equatorial Pacific (WEP) in both configurations, more pronouncedly in SR. The poor skill in the WEP is directly linked to a misrepresentation of its relationship with the ENSO region, which is ultimately associated with an overly strong westward extension of ENSO-related variability, a model error more pronounced in SR. This erroneous spatial simulation of ENSO is related to the mean cold bias of the cold tongue, which progressively extends westwards with the forecast time. We show that an overly weak air-sea coupling, more pronounced in SR, prevents the model from simulating the correct ENSO development. We also show that a better simulation of the Atlantic Nino teleconnection with the tropical Pacific in HR compared to SR leads to better ENSO prediction. Improving model resolution can increase the predictive skill of forecast systems by improving the simulation of the mean state and atmospheric teleconnections. However, ENSO simulation errors and mean state biases need to be better understood to improve forecasts, in particular in the WEP, a region of convection particularly important for teleconnections to extratropics.
- Preprint
(13131 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 13 Dec 2025)
- RC1: 'Comment on egusphere-2025-4658', Anonymous Referee #1, 09 Nov 2025 reply
-
RC2: 'Comment on egusphere-2025-4658', Anonymous Referee #2, 01 Dec 2025
reply
Review of "Comparing the seasonal predictability of the tropical pacific variability in EC-Earth3 at two horizonal resolution" by Carreric et al. (2015)
General Assessment:
This manuscript investigates the impact of increased horizontal resolution in the EC-Earth3 model on the seasonal predictability of tropical Pacific sea surface temperatures (SSTs). The authors report an improvement in the SST anomaly correlation coefficients (ACC) in the western Pacific, comparing the standard resolution (SR) and the high resolution (HR) configuration. This enhancement is attributed to a better representation of the mean state and ENSO teleconnections. They also found the improvement of ENSO predictability results from the better presentation of Atlantic teleconnection. The findings are generally sound and contribute meaningfully to the ENSO and seasonal prediction communities. However, several points regarding the interpretation of results and discussion require modification prior to publication.
Major Comments:
-
Significance of ENSO Predictability Improvement: The assertion that ENSO predictability is significantly improved in the HR configuration is not strongly supported by the presented data. In Figure 1, the difference in SST ACC values between the SR and HR configurations in the central Pacific is marginal, ranging between 0.05 and 0.1. As shown in Figure 2, the ENSO index ACC already reaches 0.7 in the SR configuration. The observed increase to 0.75 or 0.8 in the HR configuration does not represent a substantial enhancement in practical forecast utility. Therefore, the conclusions regarding the magnitude of improvement should be tempered.
-
Discussion of the Spring Predictability Barrier: The manuscript should more explicitly address the influence of the spring predictability barrier, a known limitation in ENSO forecasting where skill scores typically drop below 0.6 after spring. The current analysis utilizes hindcast data initialized in May, which crosses this barrier. Although lines 470-475 mention that November-initialized hindcasts show less noticeable differences in ACC between resolutions, it is recommended that these results be presented, particularly in a format similar to Figure 2. This would allow for a direct assessment of whether increasing model resolution mitigates the impact of the spring predictability barrier.
Recommendation for Revision:
Based on the points above, it is suggested that the authors revise the abstract and conclusions to more accurately reflect the findings. A more appropriate characterization would be: "ENSO predictability shows a slight improvement with increased model resolution" and further analysis is required to determine the extent to which the spring predictability barrier is affected by this resolution increase.
Minor Comments:
-
Figure Presentation: In Figures 2, 3, and 9, the use of more distinct color palettes for the ECE-HR and ECE-SR timeseries is recommended to enhance clarity.
Typo:
-
Line 45: The term "important model" is ambiguous.
Citation: https://doi.org/10.5194/egusphere-2025-4658-RC2 -
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 247 | 84 | 18 | 349 | 19 | 19 |
- HTML: 247
- PDF: 84
- XML: 18
- Total: 349
- BibTeX: 19
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The manuscript "Comparing the seasonal predictability of the Tropical Pacific variability in EC-Earth3 at two horizontal resolutions" by Carréric et al. explores the forecast skills in ENSO in the EC-Earth model in two different horizontal resolutions. They explore what biases the models have that could explain the limitations in forecast skill. This study is of interest to the research community. I find the manuscript generally well worked out, but I do have a number of comments that should be addressed before publication. Although, none of my comments require any major work, I do find some important discussion need to be addressed. I therefore recommend major revisions.
main points:
(*) Higher resolution = better: the study suggests that HR simulations are better, and it is indeed clear from the analysis that the forecast skill of the HR simulations is better than that of the SR simulation. The study also points out that both model simulations have similar biases and that these biases are likely the cause of the degraded forecast skills. The study also points out that the higher resolution is much more expensive to run. So, is the HR model really better, all things considered?
It is worth discussing that the lower resolution may not be as good, but it is cheaper and may therefore allow faster model development by improving model parameterisations. It will also allow to run more ensembles in diffrent configurations, which could reduce model biases.
(*) Causality: The authors argue several times in the manuscript that certain biases in the model "cause" the forecast biases (.e.g, line 440). While the statements are plausible, no evidence is presented about the causality. The authors should not assume causality, when no analysis or experiments are presented that allow them to do so.
other points (in order as they appear in the text):
---------
line 52 "... and that take overly low values ...": Not clear what this means.
---------
line 87-90 "... only differ in the horizontal grid spacing ... .Differences between systems can be attributed to the change in process parametrization adapted for each resolution,": This is contradicting itself. The two models do not just different in resolution, but also differ in some of the parameterisations.
---------
line 154 "... with g10 bias corrections ...": What is that?
---------
"Drift correction": Would it make sense to analysis this? Is it different between the two simulations? Less correction for the HR model?
---------
figure 1: The labels of the rows could add the lead time in months (e.g., 2,4 and 6 mon).
---------
Figure 6 discussion: The text seem to suggest that at the start of the simulation all biases are zero? Then, why does the east Pacific SST biases are already present at start in May?
---------
line 318 "... during the development of El Niño ...": Not clear how one can see the development of ENSO in Fig. 7.
---------
"3.2.4 Summary": This is only the summary of the subsection, not the final summary. This is confusing; better rename.
---------
line 333 "The westward shift in ENSO SST and wind zones of influence directly impacts precipitation, ...": It can also be the other way around: The biases in precipitation cause the wind and SST to shift.
---- end -----