the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 May 2026
Long-Term Decoupling of Precipitation Extremes from Mean Annual Precipitation During Repeated Early Paleogene Hyperthermals in the North American Mid-Latitudes
Abstract. The early Paleogene hyperthermals, including the Paleocene-Eocene Thermal Maximum and the hyperthermals of the Early Eocene Climatic Optimum, were the warmest periods of the Cenozoic Era. Due to the similar continental configuration and drivers of extreme warmth, this period serves as an analogue for how precipitation is altered by extreme warming driven by greenhouse gases. Through high resolution geochronology and construction of a bulk organic carbon isotope curve, we identify up to 11 different hyperthermals in the Uinta Basin, Utah, adding to the small number of terrestrial sites where the lower magnitude Paleocene and Eocene hyperthermals have been recognized. We use paleosol bulk geochemistry methods to quantify changes in precipitation during these extreme warming events. We find no significant changes in mean annual precipitation during the warming events. However, paleosol mass balance results track increased clay illuviation, accumulation of redox-sensitive elements, and carbonate leaching during many of these events. These results, along with shifts in fluvial stratigraphy, provide evidence for increased intensity and seasonality or intermittency of precipitation that may be related to poleward shifts in global circulation. These results are compared to the state-of-the-art DeepMIP model ensemble, composed of the same models used for future climate simulations. The model ensemble overestimates mean annual precipitation and underestimates the seasonality or intermittency of precipitation compared to this proxy record. These differences may be a function of the coarse model resolution, missing processes, or incorrect boundary conditions that should be investigated further.
- Preprint
(5017 KB) - Metadata XML
-
Supplement
(92 KB) - BibTeX
- EndNote
Status: open (until 02 Aug 2026)
-
CC1: 'Comment on egusphere-2026-1804', Giacomo Medici, 11 May 2026
reply
-
AC1: 'Reply on CC1', Jacob Slawson, 11 May 2026
reply
Thank you for your interest and thoughtful comments! We will review the suggested citations and be sure to make the other changes in the next version of the manuscript.
Citation: https://doi.org/10.5194/egusphere-2026-1804-AC1
-
AC1: 'Reply on CC1', Jacob Slawson, 11 May 2026
reply
-
RC1: 'Comment on egusphere-2026-1804', William Lukens, 26 Jun 2026
reply
General comments:
The goal of this work is to use geochemical analyses of paleosols in the Uinta Basin to assess hydroclimate changes in response to multiple Paleocene through early Eocene hyperthermal events. This work is of potentially high significance, as hyperthermals are frequently associated with gaps in terrestrial records or retain only partial preservation; accordingly, they are most commonly studied in marine sedimentary successions. The authors seek to characterize not only mean annual precipitation but also the dynamics of precipitation (seasonality, intensity, and intermittency) using bulk geochemical proxies and mass-balance analysis.
The major strength of this paper is the stratigraphic context and number of high-quality geochemical analyses paired with paleosol descriptions. The interpretations of paleo-rainfall dynamics are compelling within the context of the MAP proxy reconstruction and other, prior work on fluvial sedimentology.
The major flaws of this work surround the paper’s central interpretations, that is, the mass-balance calculations and their interpretations. I applaud the authors for their creativity in these analyses, but justification of many interpretations is lacking. I think there is a tenable path forward to final publication if interpretations are either bolstered or tempered, and if some of the text is reframed to reflect what is known from modern soil science and what can be (or cannot be known) given constrains while working in deep-time systems.
My review will focus on the elements of this work that would better allow researchers to clearly understand the links between what is being measured, how such data are most appropriately interpreted, and what it may mean for the larger story. I also provide some editorial comments regarding figures and text.
Specific comments:
Mass-balance geochemistry
I am concerned with the interpretation of mass-balance calculations. The authors assessed multiple possible parent materials, on which the entirety of mass-balance analysis rests. Each of the pedotypes presented in Fig 4 show fining-upward sequences, which are ubiquitous in fluvial deposits. The authors correctly sought candidate ‘proxy’ parent materials that may best represent the B horizon sediment prior to pedogenesis, as the underlying, coarser parent materials would offer results dominated by gran size effects. However, even after careful assessment, the null hypothesis in any comparison should be that the primary reason elemental constituents show gain or loss would be due to either 1) differences in grain size, or 2) unrepresentative parent material selection. If those can be ruled out logically, then the inferences on pedogenic processes should be pursued. I believe that these two null hypothesis cases could best explain some of the trends and values observed in the section. My assertion is based on the fact that most of the paleosols are mudstones and the parent material choice is a very fine sandstone, that there appear to be secular changes in sedimentology and/or paleopedology up-section (in contrast to text on L328-329), and that some of the mass-balance results do not make sense with pedogenic features observed.
Clay illuviation is not the only process that removes Na, K, Al, and Si.
- Hydrolysis of feldspars removes Na, but Na very rarely fits into authigenic minerals in soils (e.g., Na-montmorillonite or analcime). Rather, Na usually is dissolved and flushed into groundwater. The same goes for K, depending on the clay mineralogy (which was not measured, but inferred to be smectite, which doesn’t contain K). I would instead interpret the Na and K mass-balance patterns to indicate a mismatch in parent material to paleosol mineralogy (i.e., fewer feldspars in the mudstones).
- Further, clay illuviation should be identified using morphology - did you see clay skins in hand sample (macromorpholgy) or in thin sections (micromorphology)? Lacking physical evidence, there is no basis for inferring clay illuviation from mass-balance results alone (Fig. 8a).
- Nearly all samples show SiO2 losses (L308-309) – this is probably because there is less quartz in the finer mudstone paleosols compared to the very fine sandstone proxy parent material. Given that, it’s not surprising.
- Nearly all samples show Al2O3 gains; this is unlikely to be from clay translocation alone, and is most likely an artifact from higher clay content in the mudstones compared to the proxy parent material.
Carbonate translocation is not the only process that could cause lower CaO, SrO, and MgO in the paleosols compared to parent material (e.g., Fig 8C, lines ~430 and 463).
- Was the very fine sandstone that was used for parent material possibly cemented with carbonate? If not, the CaO losses, for example, would need to be attributed to dissolution of detrital carbonate (is there detrital carbonate in these deposits?) or dissolution of Ca-silicates like apatite, anorthite, or pyroxenes (are there Ca-silicates in the sandstones?).
- Nearly all samples show CaO losses (Fig 8C), but pedotypes C and E contain carbonate nodules; why would there not be any CaO gains in these profiles? On L426-434, it is stated that fluctuating water tables in a semi-arid environment would lead to carbonate dissolution-reprecipitation. If so, why would the presence of carbonate nodules in a paleosol profile result in CaO loss relative to parent material? I would assume you would see CaO gains, depending on the source of the Ca (is it detrital and just remobilized, or is from dust deposition, etc.?). After all, carbonate nodules are secondary accumulations of carbonate and bulk CaO should reflect that, no?
Redoximorphy results seem sound, but only because there is clear, independent evidence of oxidizing conditions from the macroscopic profile descriptions.
Linking of mass-balance to rainfall dynamics is not supported.
The mass-balance trends are linked to rainfall intensity and intermittency. This is new to me. What work in modern pedology supports this? This should be clearly established in the introduction and defended in the Discussion. It is addressed very briefly around L450 but far too insufficiently, as the most impactful interpretations are drawn from the linkage between mass-balance results and rainfall dynamics.
Paleosol Profile Descriptions and Sampling
The pedotypes are shown as B-C-R horizon successions.
- What kinds of B horizons are present? There is plenty of information provided to infer that some are most likely Bw, Bss, Bk, Bg, Bkg, etc. This should be clarified, particularly because the pedotypes are interpreted to be Inceptisols or Vertisols. The authors should consider the new paleosol taxonomy published by Nordt et al. (2025), noting that some paleosols may not meet the classification of “Vertisol” and instead may key out as vertic Inceptisols, or similar things. This is true for paleosols with wedge peds but lack true master slickensides (e.g., Pedotype G on L226-228). [Nordt, L., Stinchcomb, G., McCarthy, P., & Driese, S. (2025). Soil Taxonomy adapted to buried paleosols: First approximation. Earth-Science Reviews, 266, 105141.]
- “R” horizons are bedrock zones underneath the weathering front that progresses into residuum (e.g., Shoeneberger et al., 2024, p. 2-3: “bedrock, strongly cemented to indurated”). It is impossible that each of the parent material deposits as the base of the fluvial fining-upward sequences were lithified prior to pedogenesis. They are rocks now, but that doesn’t mean they were R horizons – they would be C horizons.
Composite and cumulative paleosols were not sampled due to lack of equilibration with climate and multiple climate state influences, respectively (L133-136). Was this observed with your data, or was this a decision made a priori, and therefore sampling was not performed on such profiles? I’d be curious to know how many profiles were composite or cumulative in the succession – are you missing a fair amount of paleosols by excluding these?
I may have missed this somewhere, but it seems that there is a secular change in pedotypes up-section (Fig. 5). Is this explained anywhere in the text?
Stable Isotope analysis
It is stated that samples were acidified and then washed with Milli-Q water to remove carbonates prior to stable isotope analysis (~L160-165). Was the supernatant removed by pipette? Was centrifugation used? Was the supernatant evaporated off? There is debate regarding acidification methods (we all have preferred approaches), so be as specific as possible, particularly because your d13C data are offset relative to the Birgenheier et al. (2020) data in Fig. 9.
Also, please provide the d13C values and number of in-house standards that were used for correcting your data. This is minimum-reproducible information required these days, particularly given the statement on L545-547 regarding differences between your data and those of Birgenheier et al. (2020).
Stable Isotope interpretations
Figure 5B shows the d13Corg data generated in this study. I found two components of your data to be noteworthy but are not discussed in the text.
- First, most of the values are on the higher end of d13C values for most modern C3 plants once you account for differences in d13C value of atmospheric CO2 (d13Ca). According to the Tipple et al. (2010) reconstruction, d13Ca values mirror those of the CENOGRID stack (Figure 5a) and range from about -4 at the base of your section up to -6.0 to -5.5 at the top. The difference of about 2.5 per mil (at the base of the section) and 0.5 per mil (at the top of the section) relative to a preindustrial atmosphere implies that your values of about -22 are like a ‘preindustrial’ -24.5 at the base and a -22.5 toward the top. These values compare favorably with your MAP estimates that suggest semi-arid conditions.
- Second, whereas the benthic d13C records show a secular, negative shift from 58 to 54 Ma, your data do the opposite, albeit in a noisy way. Is there an overall tendency toward higher d13C values in your section (after applying atmospheric corrections), and does that tell us anything about MAP values and/or vegetation community turnover? Or something else (or nothing)?
Comments on Figures:
All Figures should have clear y-axis tick marks and labels, including minor ticks. It’s difficult to match strat levels between some panels (Fig. 6, 8, 9 especially). It would also be more helpful to show on Figure 6 the inferred positions of hyperthermals with some shading or other identifier. Without these, it is difficult for a reader to ascertain whether or not trends in the data correspond with hyperthermals (e.g., validating comments on and about L435-440). It would also be useful to annotate the pedotypes in Figure 8 along the y-axis.
Minor and technical comments:
- One could argue that the Early Eocene decidedly lacks similar continental configurations to today (L9-10; L45): differences in ocean gateways comes to mind, and also different topographies in many mountain belts (particularly compared to, e.g., the Miocene).
- L69: “eastern terminus of the Sevier Orogeny” is odd phrasing. Are you referring to the eastern margin of deformation associated with the Sevier Orogeny? (Orogenies are events, not special domains)
- L94: “…is a mixed fluvial-lacustrine deposits” – rephrase
- L155: “Aluminum…is mobile under certain soil forming conditions…”. Be specific here – Al is mobilized via dissolution under acidic conditions or mobilized physically via translocation of clays in well-drained conditions.
- L246: “climofunctions” is not the best word choice. A climofunction is a single equation, usually a regression line. That might be true for the CIA-K proxy, but the RF-MAP and PPM are better described as models, proxies, or if you’d like to be most specific, “statistical model” to differentiate these from actual climate models (like those used in DeepMIP).
- L307: The Sheldon et al. (2002) reference is not appropriate for this statement. I would suggest a soil chemistry textbook or mineralogy reference. I believe Moore and Reynolds (1997) even have a chunk of text on this, though I cannot remember where (and I no longer have a copy).
- L320: I found this to be a jarring start to the Discussion. Could you provide a starting paragraph that frames where we are going, at a high level, as we transition into the discussion?
Citation: https://doi.org/10.5194/egusphere-2026-1804-RC1
Data sets
Supporting information for: Long-Term Decoupling of Precipitation Extremes from Mean Annual Precipitation During Repeated Early Paleogene Hyperthermals in the North American Mid-Latitudes Jacob S. Slawson, Piret Plink-Bjorklund, and Emily J. Beverly https://doi.org/10.5281/zenodo.19408468
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 180 | 34 | 19 | 233 | 17 | 18 | 19 |
- HTML: 180
- PDF: 34
- XML: 19
- Total: 233
- Supplement: 17
- BibTeX: 18
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comments
Very good research on the Paleogene paleoclimate. Please, follow my feedback to improve your manuscript.
Specific comments
Lines 33-34. “Given that dramatic increases in temperature are expected under even intermediate climate change scenarios, there is an increasing need to turn to the past to test models and inform about future climate”. Insert references on recent literature on past climates (with a focus on Paleogene) to better unravel the present.
- Medici, G., Marianelli, D., Cornacchia, I., Gori, F., Brandano, M. 2026. Multi-disciplinary approach to paleokarst occurrence in the Eocene–Oligocene succession of the Apulia Carbonate Platform (Salento, Italy). Facies, 72, doi: 10.1007/s10347-026-00729-5
- Quan, C., Liu, Z., Utescher, T., Jin, J., Shu, J., Li, Y., & Liu, Y. S. C. (2014). Revisiting the Paleogene climate pattern of East Asia: A synthetic review. Earth Science Reviews, 139, 213-230.
Line 64. Describe the 3 to 4 specific objectives of your research by using numbers (e.g., i, ii, and iii) at the end of the introduction.
Lines 65-110. Insert background information on tectonics.
Lines 65-110. Insert more detail on the sedimentary structures of your paleosoil, floodplain and channel deposits.
Lines 116-175. The number of samples for each type of analysis is not clear.
Figures and tables
Figure 1a. Location in US very difficult to see.
Figure 1. There is room for a larger figure to make the stratigraphy more visible.
Figure 3a. The sedimentary deposits on the uppermost part look floodplain. Do you need to change the caption?
Figure 4. Spatial scale unclear.
Figure 5. Increase the graphic resolution. Letters and numbers are difficult to see.
Figure 9. Same here. Increase the graphic resolution. Letters and numbers are very difficult to be read.