Let me begin by saying that I believe this paper is a worthwhile contribution to this very interesting period in earth history, Middle Miocene Climatic Optimum, especially since this is a terrestrial record and most records focus on marine. But I cannot recommed this for publication in it’s current state to Climates of the Past; however, I would happily rereview.

I have attached line by line comment in a pdf, but below I summarize my major concerns with the manuscript in its current format:

• I have some major concerns about how the carbonates and bulk geochemical samples were collected. These need to be collected from certain depths and it does not appear that they were. Things should also have been trenched and it does not appear that they were at least based on figure 2. I have made specific comments throughout the manuscript regarding these topics.
• None of the methods describe any efforts to identify or avoid diagenesis in the carbonates, such as petrography or cathodoluminescence.
• Significant more attention needs to be paid to the literature about stromatolites and fluvial carbonates and the effects on carbon and oxygen isotopes that are very different than those that affect pedogenic carbonates.
• Many points in the discussion are unsupported by evidence either in the paper or from other literature. I noted areas that require more discussion or specificity to support these interpretations.
• The clumped isotope interpretations and analysis are lacking. More samples are need to be able to make the interpretations in this paper. Some samples only have 1 replicate, which is not acceptable either for stromatolites or soil carbonates. In addition more attention needs to be paid to the literature for soil and stromatlite carbonates – reading some of this literature would really help to improve you interpretations.
• Bulk carbonate should never be used in isotopic interpretations, it is too susceptible to diagenesis. Although it is useful to help determine IF your samples have diagenesis. See additional comments in text on this topic.
• I suggested some additional lines of research that should be investigated using the data you already have from bulk geochemistry PPM1.0, RF-MAP, mass balance, etc that would greatly improve your interpretations.

Sharma et al. presented geochemical records for Middle Eocene Climatic Optimum (MECO) from a fluvial sedimentary succession in Spain. The authors attempted to constrain the regional climatic conditions during MECO with a suite of sedimentary archives and geochemical proxies. This is a valuable contribution as few previous studies has identified MECO and assess its climatic implications in terrestrial records. However, the following issues should be addressed/clarified prior to publication.

1) Uncertainties in age constraints

            Section 2.2 provided a brief description of the age constraints but discussions about the associated uncertainties in the age models are missing. As the preservation of stratigraphic record is almost always incomplete (e.g. Paola et al., 2018), it is important to provide details on the uncertainties in sample ages and discuss how they affect the interpretation of the geochemical proxies and the comparisons with other records.  

2) Statistical/quantitative analyses

            Error bars are missing in many of the plots. Uncertainties estimates were not provided in some of the results. And there’s a general lack of statistical analyses for claimed correlation, trend, and changes. See line by line comments for details.

3) Carbonate diagenesis

            Carbonate minerals are extremely susceptible to diagenetic alteration. The authors interpreted the measured compositions as primary signals but did not provide evidence for the lack of diagenesis. And the possibility of diagenesis is not mentioned/discussed until the very end of the manuscript.

4) CIA/chemical weathering/clays

            The chemical index of alteration (CIA) can be modulated by other factors such as grain size (e.g. von Eynatten et al., 2012) and is not solely indicative of the degree of chemical weathering. And the characterization of lithology/source rock from which the sediments were derived from is important (but currently missing) for the interpretation of the measured CIA and clay mineralogy.

Line by line:

Line 168-169: What defines an excursion? There may be a slight decrease in the seven-point average d13C_org values between 30-50m, but the absolute magnitude of change in d13C_org appear to be the largest between ~80m and 110m – what makes the signal between 30-50m an “excursion” rather than natural variability? Also, the individual d13C_org values during the MECO period appears to be more variable than other part of the section, is this expected?

Line 174-175: Statistical analyses are needed to show robust correlation between the two records, visual similarity is not a very strong argument.

Line 179:  What’s the reason to use a 7-point moving average (as opposed to 3-point or no smoothing at all)?

Line 186: What are the “several factors”? d13C values cannot be interpreted only in terms of MAP if other factors are not constrained.

Line 193-194: why does the lack of closed canopy forest + dry conditions = sparse vegetation? Maybe provide a more detailed explanation.

Line 220-221: what does “a more sensitive proxy for the MECO” mean and what are the implications? The absolute change of d13C maybe larger in the organic carbon isotope record, but the background d13C_org data is also more variable.

Line 229-232: where are the “normal paleosol values” from? Are stromatolites and pedogenic nodules effectively the same endmember? They form via different mechanisms and intuitively they may express different fractionation – references/evidence that support they can be considered the same endmember would be helpful. What’s the uncertainty of the estimated fraction with pedogenic origin?

Line 287-288: what are the uncertainties with the MAP estimates? These appear to be very small variations, statistical analyses is needed to support your claims (i.e. there’re actual changes).

Line 299-310: a summary table with information (e.g. location, age, citation) of all the other records compared here would be helpful.

Line 307-310: I didn’t fully follow the argument here. Why high Ca content results in underestimation of precipitation?

Line 333-334: if the samples are powdered, how come there’s still inhomogeneity?

Line 341: how uncertain is this 400-kyr lag?

References:

Paola, C., Ganti, V., Mohrig, D., Runkel, A. C., & Straub, K. M. (2018). Time not our time: physical controls on the preservation and measurement of geologic time. Annual Review of Earth and Planetary Sciences, 46, 409-438.

von Eynatten, H., Tolosana-Delgado, R., & Karius, V. (2012). Sediment generation in modern glacial settings: Grain-size and source-rock control on sediment composition. Sedimentary Geology, 280, 80-92.