This paper presents an analysis of oxygen isotopes in modern water isotopes and ancient paleosols from the Great Plains. The authors apply a vapor transport model to see whether this simple 1-D model can explain observed gradients of isotopes (the model does relatively poorly), and also analyze whether these gradients have changed in the past. The abstract and paper text are well written, and it is overall appropriate for the journal. Some conceptual considerations:

Table 1 is helpful, but it would be helpful to have a map figure where sites are color-coded by their lower age and/or upper age.

The mismatch between the reactive transport model and observations is quite dramatic, and therefore I would like a greater discussion of the sources of mismatch between the reactive transport model and the data. Rainout is definitely a factor that can affect isotopic gradients, but what about storm statistics and the changing location of certain types of storms (see papers below)?

Manser et al.

Stable isotope evidence for long-term stability of large-scale hydroclimate in the Neogene North American Great Plains

EGUSphere

Comments of David Fox

Summary

The paper presents new d18O data from Miocene pedogenic carbonates from the Ogallala Group/Formation in the Great Plains region of the USA and, in conjunction with compiled published data, analyses the spatial pattern in relation to patterns predicted from climate data and moisture transport model. The paper is well-conceived and clearly written and is mostly thorough in its treatment of the data, the geological context, and the modern and Miocene climate system and implications of the data for our understanding of Miocene hydroclimate in central North America. I have a few mostly minor comments below and think this paper is ready suitable for publication after minor corrections and additions.

General comments.

1. The lithostratigraphic terminology for the Ogallala is a bit complicated across the region. In some states (e.g., Nebraska) it is a Group with constituent formations and in others (e.g., Kansas) it is an undifferentiated Formation. I think Tedford et al., 2004 (full reference below) is the most authoritative source for the current terminology. I suggest that you address this complexity and explain your use and be consistent throughout (which you are now, but by treating it as a Formation, which is not correct everywhere in the study region).

2. Include all of the data in a supplemental with this paper, both the previously published and your new data, rather than really solely on Dryad for the new data. I could not access the new data on Dryad, though maybe it is not yet accessible or maybe it is pilot error? Regardless, I think including all of the underlying data in a supplemental with this paper is critical even with the data available via Dryad as well.

3. Address explicitly the age assignments and uncertainties. You discuss age uncertainty as a factor in comparisons to published model results, but you do not discuss in any detail the uncertainties in your age assignments beyond listing maximum and minimum ages in Table 1. What is the basis for the assignment of each section and how did you treat the uncertainty? Given that you do not examine the data as time series, this is not so much an issue with your interpretations as a matter of completeness.

4. Do you plan a separate paper for the carbon isotope data? If so, then perhaps including the new carbon isotope data from carbonates here is not necessary. However, if you do not plan a separate paper on the carbon data, then I suggest that you include at least the new carbon data (and probably all of the published carbon data as well) in the supplemental table of data and add a few sentences summarizing how the new carbon data compare to the published data. I recognize the carbon data are not the focus of this paper, and you can qualify some brief comments by saying a detailed discussion is beyond the scope of this paper, but I worry that the carbon data will be orphaned and lost if you do not plan to publish them and do not include them here. The published Miocene data (my papers that you cite and Lukens and Fox (2022. Palaeogeography, Palaeoclimatology, Palaeoecology 586. doi.org/10.1016/j.palaeo.2021.110760) have a strong central tendency (-7.2 to -6.8‰ V-PDB) and low variance, so a simple comparison of your mean and s.d. to the published data would suffice.

Specific comments (indexed by line number)

36. It might be worth noting the influence of longer timescale climate fluctuations here, particularly ENSO. As I understand, the onset of the Dust Bowl coincided with a particularly severe El Nino event, and the little Dust Bowl in the 1950s also coincided with a string El Nino. These longer frequencies in the climate system are obviously not the focus of your paper, but they are relevant in the intro it seems.

39. Is Powell’s work relevant beyond being antecedent? He is a somewhat complicated figure historically, perhaps less so than others, but the mention of him does not do much work here. He is traditionally treated as somewhat of a founder and hero in North American geology and geomorphology, but I am not sure everyone in North America views him so positively. That said, I don't think your treatment is problematic.

48. Check on the regional nomenclature for the Ogallala. It is a Group in Nebraska with multiple constituent Formations, but I think used as an undifferentiated Formation elsewhere in the region, certainly in Kansas (see Ludvigson et al., 2009). My sense is that in Texas, different authors use it as either a formation or a group, but I am not sure which is currently formally correct. You should point this out and establish here a terminology that you will use throughout for the lithostratigraphic unit. For example, you could use Ogallala for the unit and specify “Ogallala Aquifer” when refering to the aquifer. I think Tedford et al. (2004) (see note for line 182) is the most definitive authority on the regional terminology.

51. To this list of references, thanks to my slow review, you can add Korus and Joeckel, 2023. Telescopic Megafans on the High Plains, USA Were Signal Buffers in a Major Source-To-Sink System. The Sedimentary Record 21. https://doi.org/10.2110/001c.89096.

52. The Ogallala is not capped by a single, laterally continuous caprock. This idea was prominent in the early literature on the stratigraphy in the region, but is not correct.

54. In Nebraska, the Ogallala also lies on top of formations of the Arikaree Group.

148. Should this be “between the land surface and the atmosphere”?

161. I am not sure either of these are the best citations for the orographic effects on precipitation amount and d18O (e.g., Rozanski et al., 1993).

182. You should cite Tedford et al. (2004) here as the most recent detailed synthesis of the mammalian biostratigraphy in the region for the study interval, and it includes more or less all of the reliably dated ashes to date. Tedford, R.H., Albright III, L.B., Barnosky, A.D., Ferrusquia-Villafranca, I., Hunt Jr., R. M., Storer, J.E., Swisher III, C.C., Voorhies, M.R., Webb, S.D., Whistler, D.P., 2004. Mammalian Biochronology of the Arikareean through Hemphillian Interval (Late Oligocene through Early Pliocene Epochs): Late Cretaceous and Cenozoic Mammals of North America: Biostratigraphy and Geochronology. Columbia University Press, New York, pp. 169–231.

183. See my earlier comment about lithostratigraphic nomenclature.

186. The citations here could include Tedford et al. (2004), but also more primary literature on each unit.

189. “in Texas” needs to be moved as the Blackwater Draw Formation is only in Texas and only overlies the Ogallala in places there. You need to be clear and specific about this here as the Ogallala is overlain by high energy deposits in places elsewhere (i.e., the Stump Arroyo Mbr of the Crooked Creek Fm in SW Kansas and the Broadwater Fm in W Nebraska).

192. The Ogallala includes multiple stratigraphically distinct cap rocks or mortar beds and not one regionally extensive or continuous one and not only one at the top of the section.

203. As suggested before, you should clarify the stratigraphic nomenclature earlier and make sure it is complete and accurate.

207. This is true almost everywhere and it is well documented that the caprock is not a single unit stratigraphically.

219. Could add Tedford, 1981. Mammalian biochronology of the late Cenozoic basins of New Mexico. Geological Society of America Bulletin 92: 1008-1022.

228. The references for the age assignments for each section should be given in Table 1 so that readers can evaluate the age assignments on their own. You need to state here how you assign a specific age to each section and/or sample given that the sections have age ranges. Are all samples in a section given the same age? Do you assume a sedimentation rate and assign ages in stratigraphic sequence, and, if so, how do you calculate sedimentation rate?

235. Be explicit about how the standards were used…Which was used to correct to the V-SMOW scale and which were used for runtime QA/QC? This needs to be reported clearly. What is the analytical precision and how was that determined?

295. Here and elsewhere, what is the basis for the order of citations? It does not seem to be alphabetical nor chronological.

331. The table of published data should include the original published C and O values and your calculated paleoprecipitation values.

332. I see no reason not to include the full data table as a supplemental to this paper also so that the data are with the interpretation. I cannot access the data using the doi nor by searching on Dryad, though perhaps the data are not yet posted or accessible?

434. Higher is probably a better word choice here than greater.

469. Tedford et al., 2004

537. One factor you do not discuss is the difference in water use efficiencies and evapotraspiration fluxes of woody vegetation vs. C3 grasses vs. C4 grasses. This is embedded in your consideration of land surface characteristics, but there is literature on this could be of use. The phytolith data suggest that grasslands were present throughout the Miocene, and the carbon isotope data have been interpreted as indicating a constant amount of C4 grasses throughout all of or almost all of the Miocene, so these patterns are consistent with your lack of a spatial signal in the d18O data.

545. data have