This study investigates the silicate weathering and carbon cycle for the Pliocene climate that featured a reduced SST gradient in both zonal and meridional directions (the so-called “permanent El Niño”). The authors first explore the impact of present-day El Nino events on the weathering flux using the present-day observation/analysis. Their analysis suggests a small net impact (~2–3%) of ENSO SST patterns on the weathering flux. Next, using climate model simulations, the authors find a similarly small net impact of Pliocene-like SST patterns on the weathering flux.

The research question of the manuscript is very relevant for understanding the global cooling since the Miocene and Pliocene. The introduction is very well written, reflecting that the authors have very good understanding of the problem and relevant dynamical processes. However, I must admit that, to understand the manuscript (such as the motivation and the complicated procedure for creating the Pliocene slab ocean simulation with altered ocean q-flux), I had to read the manuscript several times. This makes me wonder whether the authors can work out a simpler approach to highlight the key experiments and results. Please see also below for a few other comments. These comments need to be addressed before the publication of the manuscript.

Major comments

1. Title. Since the main conclusion of the manuscript seem to be that the weathering flux change from the Pliocene-like SST pattern is small in global mean, I think the authors could use a better title to clearly deliver the message, such as “A potentially small net effect of the Pliocene temperature pattern on the silicate weathering and the Pliocene-Pleistocene cooling”.
2. Related to 1, the authors conclude that “We find support for this hypothesis…” (Line 416). Given the small net contribution from the SST pattern effect and the abundant uncertainties from SST pattern reconstruction, precipitation bias from model/reanalysis, assumptions built in the silicate weathering model, I wouldn’t conclude that “We find support for this hypothesis”. The authors’ analysis actually lends little support for the hypothesis.
3. The manuscript is too long with too many figures (17 figures!), which may decrease the readability and reduce the focus of the manuscript. One way to improve, I think, is to reduce the number of main figures though combining multiple figures into a single figure with a common theme. For example, Figures 2‒4 could be combined into one figure showing the average chemical weathering rates and the separation into different regions. Similarly, Figures 5 and 6 could be combined with a focus on the visual and quantitative comparison of the weather rates between El Nino and La Nina events. Same for Figures 10 and 11, and for Figures 16 and 17.
4. Table 1 and the discussion on the Pliocene experiments are lengthy and difficult to follow. I am sure the authors have carefully thought about the purpose of these experiments, but it is not necessary to list and show all the experiments. For example, the “control” simulations with CO2 of 1138.8 and 854.1 ppmv are not discussed in the manuscript (854.1-ppmv simulation is shown in Figure 12 but may not be needed); same is true for a few other simulations. I think it will help readers to follow and help highlight the most important findings of the manuscript if the authors could carefully examine their simulations and delete unnecessary ones.
5. The authors’ simulations highly rely on the SST pattern adapted from Burls and Fedorov (2014), but they failed to provide information on how well the SST pattern match the proxy record. I suggest the authors adding values of the meridional and zonal SST gradients from the simulations and the comparison with proxy data. Values of the SST gradients should be provided for COA-ctrl, COA-Plio and all the slab ocean simulations, with Pliocene values compared with proxy reconstruction. This information could be added to Table 1. Also, the global mean surface temperature of each simulation could also be added to Table 1. These key features will help readers better understand each simulation. A short discussion on the implication/caveats should be added if the simulated SST pattern does not match the proxy reconstruction.
6. Please add significance test for many of the analysis, such as the difference map in Figures 1, 5, 8, 10, 14, and 16.
7. At many places, when summarizing the effect of weathering flux changes, the authors used the estimated temperature changes, such as Lines 10–11. The authors need to clarify how the temperature changes are estimated and what is the associated uncertainty.
8. When explaining the SST offset between Pliocene and preindustrial slab ocean simulations (Lines 305–351), the authors could mention the SST pattern effect on the cloud radiative effects (such as Zhou et al., 2017, doi:10.1002/2017MS001096).

Minor comments

Line 13: Spell the “C cycle” fully.

Line 27: Change “net net (upward)” to “net (upward)”.

Line 205: Change “seeks” to “seek”

Line 310: Change “than” to “as”

Line 320: Change “visible” to “shortwave”

This paper uses historical climate reanalysis, paleoclimate simulations and a geochemical weathering model to explore the effect of ENSO on silicate weathering, aiming to test the idea that Pliocene permanent El Nino conditions led to changes in Earth’s surface weatherability, such that removal of these conditions may have contributed to long-term CO2 removal and cooling. This is an interesting idea, and the relatively minor effect that is found seems reasonable.

I think the paper can be published with some minor revision. My main points are:

1. While the text starts brilliantly, the discussion and conclusions need much more proofreading. I found a lot of gramattical issues in this seciton.
2. Section 3.2 which introduces the climate modelling strategy should be explained more clearly. It is really tough to get through at the moment. I would recommend a paragraph at the start of this section clearly setting out what they ultimately want to simulate and how they will do it, before going into the details.

Minor comments:

Given that this is a more interdisciplinary paper which focuses mostly on chemical weathering, the authors might consider a higher level explanation of the walker circulation, Bjerknes feedback and southern oscillation.

Biotic weathering enhancement may play a role here? Plants are very dependent on water. Could the authors acknowledge this? Does their modelling hint at how strong this effect might be?

Line 205: “we apply the method of Burls and Fedorov (2014) to seeks to reproduce Pliocene SST.” A typo here?

First part of section 3.2 and figure 7 could be explained more clearly. It is not clear at the start of this section what the motivation is for producing these different sets of runs with the full or tropical-only SSTs, and what the comparison in figure 7 is really showing. Some introductory text would be useful.

Line 367 “If carbon cycle was left to reach balance”, missing word? E.g.  “the carbon cycle”

And “CO2 would decrease until temperature is back around pre-industrial one” perhaps change to “CO2 would decrease until temperature returned to the pre-industrial value”

Line 377: “meridional and tropical zonal gradient of temperature” change to “meridional and tropical zonal temperature gradients”?

Line 394: “the results agree on a moderate warming effect of permanent El Niño”. It would be useful to re-state the actual temperature increase here. Also, this is described as minor in the abstract and I would agree. Moderate, in the context of Pliocene warmth, would probably be more like 1C?

Line 396: “that is ~ 2.5 C than pre-industrial”, missing word “warmer”?

Line 398: “If flatter meridional temperature gradient”, missing word “the”?

Line 400: “This raise the question”, typo “raises”

Line 418: “the difference of a permanent El Niño climate state on global silicate weathering”. Suggest change “on” for “for”

Line 421: “particularly on tropical weathering hotspots”, change “on” to “in”

The authors proposed an interesting hypothesis that the SST pattern in the warm Pliocene has contributed to weaker weatherability and thus helped to maintain the warmer Pliocene climate. To test their hypothesis, the study first calculated  weathering fluxes in El Nino years versus La Nina years using reanalysis data. They find reduced (silicate) weathering fluxes during El Nino years due (mainly) to a shift in the precipitation pattern. Next, a set of “Pliocene” simulations were created to further explore how SST patterns in the Pliocene might have affected the weatherability. The results show that regional increases and decreases in weathering fluxes largely cancel out one another. Overall, I am not fully convinced that the Pliocene simulations used in this study well reproduced the Pliocene climate. However, the study highlighted the importance of the SST pattern and mean climatic state on silicate weathering. A valuable lesson is that robust global SST reconstructions integrated with climate simulations are critical in evaluating Earth’s thermostat. I propose minor revisions before acceptance.

Major comments:

Line 213-264: “In order to create a Pliocene SST field for the tropics only”. It seems to me this is a scenario with a zonal SST gradient similar to the Pliocene while maintaining a modern meridional SST gradient. But what is the motivation and justification to create an SST field for the tropics only? And why results from this tropical-SST scenario were not included in Figure 15?

Line 352-362:

“to go back down to pre-industrial temperature, with full Pliocene SST, CO2 needs to be lowered to ∼140 ppmv.”

“For this reason, we cannot perform the inversion to compute the equilibrium CO2 where silicate weathering flux balances pre-industrial degassing. We instead analyse the weathering fluxes at fixed CO2.”

This paragraph raises some puzzles. Figure 15 seems to suggest that both global temperature and pCO2 will be much lower in the Pliocene in order to balance the pre-industrial CO2 degassing rates. The inference is that (silicate) chemical weathering fluxes must be higher in the Pliocene given the fact that Pliocene was warmer. However, there is no evidence suggesting that Pliocene CO2 degassing rates were higher than today. Based on Figure 15, if we assume Pliocene has the same pCO2 as today (correspondingly ~2.5C warming), the silicate weathering anomaly is almost ~1 T mol/yr. This is almost a 20% increase from the modern value (Line 153), a very large number for the long-term carbon cycle. If we extrapolate the red curve (full-Pliocene-SST scenario) to the level where weathering flux anomaly = 0, the global temperature probably will be (much) lower.

Thus, the full-Pliocene-SST scenario seems to produce a combination of climate-and-weatherability that is inconsistent with geological evidence. This leads me to suspect that either the simulated Pliocene climate is a poor representation of the Pliocene climate or the sensitivity of GEOCLIM to climate changes needs some revision.

Furthermore, if we extrapolate the red curve to modern global temperature, it seems to suggest a positive anomaly in weathering fluxes. This would indicate a higher weatherability in the Pliocene, as opposed to what Molnar and Cronin proposed? Am I missing something?

My sense is that discussions in lines 380-405 are trying to address some of these puzzles. I would recommend the authors reorganize the discussions in sec 4 a little bit and make a tighter link to Figure 15.

Line 210: Can the authors comment on how well the simulated SST agrees with proxy data, rather than simply citing Burls and Fedorov, 2014? For instance, what is the magnitude of warming in the EEP relative to the proxies?

Some minor points:

The authors may want to mention at the beginning that there is little evidence of changes in CO2 degassing rates since the Pliocene. Thus, a change in weatherability (e.g. Molnar and Cronin) is the most likely explanation for the long-term coolling.

Line 68-69: It seems to be a little bit controversial now to say “El Niño events is a good representation of Pliocene permanent El Niño.”

Line 71-73: “However, one cannot quantitatively estimate this warming … yet, if silicate weathering is disturbed, CO2 will adjust …” This sentence can be confusing to some readers. Maybe reword it a little bit?

Line 106-122: consider to provide a supplementary figure showing the time series of the ENSO index you calculated?

Line 155: probably point out what climate field they used.

Line 366: “This result means that Pliocene SST does not generate any significant increase or decrease of weatherability, save perhaps at low CO2” typo? 

Line 410: “could plan an important role” typo?

Line 345: “This goes against our initial hypothesis, based on the “wet gets dryer, dry gets wetter” feature from Burls and Fedorov (2017).” I think this is the first time you mention this. It is better to front load this in the introduction.