The middle Eocene is a transitional period from the greenhouse early Eocene to the ice-house Eocene-Oligocene Transition. Due to lack of high-resolution records, the details of this long-term cooling trend are not clear. Low-resolution atmospheric CO2 records suggest declining, but highly variable levels during the middle Eocene which might have driven transient climate variations. However, high-resolution stable isotope records are lacking to document the true climate patterns over this period. Moreover, orbital scale climate reconstructions are essential for reconciling how the sensitivity to orbital forcing changed, if at all, as the greenhouse boundary conditions changes from the early to late Eocene. High-resolution isotope and trace metal records are also essential for constraining the evolution of northern hemispheric sea-ice and continental glaciation as well as North Atlantic Deep Water in the middle Eocene.

To address these shortcomings, I have generated high-resolution stable isotope and trace element data for a site in the northern Atlantic on Newfoundland Ridge (IODP, Site 1408). The records span Chron20r (46.4~43.8 Ma), and are the first orbital-scale stable isotope records for the early middle Eocene. With the exception of the uppermost portion of the section, the δ18O data indicate gradual cooling of 0.8~1.2 ⁰C/Myr over much of the section from 46.4Ma to 44.3 Ma which is consistent with the long-term global cooling trend. This is accompanied by a decrease in Mg/Ca records, while the paired δ18O-Mg/Ca analysis shows no pronounced shift in δ18Osw implying no variation in ice volume. In the upper part of the section (44.3~43.8 Ma), the δ18O shows a slight reversal consistent with warming, while Mg/Ca proxy indicates a continuous cooling. One hypothesis for the apparent between the two records is that the water mass reflects changes in the salinity of local bottom water and a possible shift toward a higher-latitude, cooler and less salty water source. In addition, clear astronomically paced cycles are observed in the stable isotopic and lithological records, and spectral analysis demonstrate that the cycles are dominated by 41-kyr obliquity cycles, a unique feature of Paleocene-Eocene records which are typically dominated by precession and eccentricity. This could reflect either shift in orbital configuration (low eccentricity signals or change in Earth’s tilt), or gradual climatic variation toward the ice-house world in the middle Eocene. Interbasin δ13C gradient reveals relatively low δ13C values at this northern Atlantic site compared with records from the Southern Ocean, indicating the meridional circulation was northward and/or North Atlantic Deep Water formation was absent or weak during this time.

Environmental Changes on the North American Mid-Atlantic Shelf During the Paleocene Eocene Thermal Maximum

Edward A. Ballaron

The Paleocene-Eocene Thermal Maximum (PETM), a large hyperthermal during the Cenozoic (~56 Ma) is characterized by a massive injection of depleted carbon into global reservoirs as indicated by a large negative carbon isotope excursion (CIE) of 3‰-5‰ in marine and terrestrial sedimentary records. Previously studied open ocean (i.e., pelagic) cores have provided evidence for abrupt sea surface warming (5-10°C) and bottom water warming (4-5°C) as well as ocean acidification at the onset of the CIE. However, due to low sedimentation rates and truncation of pelagic sections via dissolution at the CIE, pelagic cores appear condensed or incomplete. Sites on the North American mid-Atlantic margin, provide a unique opportunity to examine both marine and terrestrial responses to a large magnitude climatic perturbation. Rapid and abrupt increases in the flux of siliciclastics including kaolinite to the shelf at the onset of the PETM, indicative of an enhanced hydrologic cycle, result in expanded PETM intervals allowing for higher resolution isotopic sampling. Stable isotopic and trace metal data from New Jersey sections (Millville, Ancora, Bass River and Wilson Lake) have also documented rapid warming, coastal ocean acidification and a freshening event at the onset of the CIE in agreement with evidence for an enhanced hydrologic cycle during the PETM. Maryland sections to the south (South Dover Bridge and Cambridge-Dorchester) studied here, are more proximal to the main drainage system in the Salisbury Embayment and therefore should have experienced environmental changes (salinity) to a greater degree than New Jersey sites. Here we present coupled stable isotopic (carbon and oxygen) and trace metal data (Mg/Ca) to constrain both changes in temperature as well as salinity on the Maryland mid-Atlantic shelf during the PETM. Data presented here shows a temperature increase at SDB and Cam-Dor of ~4-5°C for planktonic foraminifera and ~5-9°C for benthic foraminifera with oxygen derived temperatures being consistently larger than Mg/Ca derived temperatures. This indicates some salinity contribution to δ18O, artificially lowering oxygen derived temperatures. This is further supported by salinity calculations which display a freshening event at the onset of the CIE due to increased runoff. Also observed in this data set is a slight weakening in δ13C and δ18O depth gradients on the shelf. Changes in δ18O gradients can perhaps be attributed to greater warming at depth and a destabilization of the thermal stratification of the water column, similar to observations from New Jersey sites Millivlle and Ancora. The breakdown in vertical δ13C gradients was perhaps due to the vertical migration of mixed layer foraminifera taxa seeking refuge from warming surface waters. Another possible explanation for the change in vertical δ13C gradients is a change in upwelling patterns driven by hyperpycnal flow or changes in wind patterns driving coastal upwelling and bringing depleted δ13C to the surface.

The early Eocene is punctuated by a series of rapid warming events, known as hyperthermals. The first among these, the Paleocene-Eocene Thermal Maximum, or PETM, has been argued to be one of the closest approximations to modern-day, anthropogenic warming in the geologic record. The PETM, and the subsequent hyperthermals of smaller magnitude, such as Eocene Thermal Maximum 2, or ETM2, can provide insight into our future warmer world. In addition to temperature changes, these hyperthermals drove significant changes in regional precipitation patterns. This study seeks to better understand these changes, focusing on the mid-latitude North Atlantic coasts, utilizing a suite of techniques including high-resolution climate models and geochemical, sedimentological, and paleoecological proxies. Our models show a significant increase in the seasonality and extremes of precipitation in response to elevated CO2 levels, with significant differences regionally depending upon orbital configuration. In the western Mid-Atlantic we find a dramatic weathering response to these precipitation changes, with an increase both in chemical alteration and physical transport of sediment. During ETM2, in contrast to studies in other regions, the climatic response was much more muted, with temperatures decreasing, a moderate increase in chemical weathering, and little response in the marine palynology, highlighting the importance climatic forcings apart from CO2 such as orbital forcing and dynamic feedbacks.

This dissertation reconstructs the response of marine carbonate chemistry to rapid carbon injection during the Paleocene-Eocene Thermal Maximum (PETM). Chapter 1 uses boron-based proxies to reconstruct ocean acidification during the PETM for the first time, concluding that surface seawater pH declined by ~0.3 units at the onset of the event, and remained acidified for at least 70 thousand years before recovering in step with temperature and the carbon isotopic signature of the PETM. Chapter 2 describes a new sedimentary record of the PETM from the deep North Atlantic which provides the first evidence for a hypothesized "overshoot" of carbonate saturation in the aftermath of the PETM in response to long-term weathering feedbacks on climate. Stable isotope records from that section demonstrate that this overshoot occurred during the recovery of the PETM, ~70 thousand years after its onset, which provides novel constraints on the evolution of the carbonate compensation depth over the event. Finally, Chapter 3 uses these new (and previous) records to constrain carbon cycle model simulations of the PETM. Consistent runs require both a large initial release as well as a protracted release over tens of thousands of years (possibly representing a slow positive feedback to warming) and the removal of isotopically light carbon to accelerate the PETM recovery, likely representing organic carbon burial. No consistent scenarios feature under-saturated conditions in the surface ocean during the PETM, consistent with the lack of calcifier extinctions during the event. Comparison of the most consistent PETM scenarios with forecasts of anthropogenic carbon emissions demonstrate that carbonate chemistry change during the PETM was less severe and far more gradual than what might be expected in coming centuries.

The late Paleocene and early Eocene (LPEE; ca. 59 to 52 Ma) was an interval of warming superposed on the abrupt carbon release events (or hyperthermals) of the early Eocene. The LPEE provides an opportunity to test the coupling of carbon release (and associated long- and short-term warming), to other climate-carbon cycle feedbacks (e.g., ocean acidification and shifts in the hydrologic cycle). Much work has focused on determining the sensitivities of sea surface pH, climate and the hydrologic cycle to carbon release during the largest hyperthermal of the Cenozoic, the Paleocene-Eocene Thermal Maximum (PETM; ca. 56 Ma). However, Eocene Thermal Maximum 2 (ETM-2; ca. 54 Ma) provides a secondary calibration point on assessing the sensitivity of carbon cycle-climate feedbacks to carbon release. Additionally, it has been hypothesized that the long-term carbon storage and release mechanisms (e.g., marine clathrates and organic carbon) are linked to the hyperthermal events, however, constraints on long-term changes in carbonate chemistry are lacking for the interval.

This thesis aims to characterize the nature of changes in the climate and carbon cycle associated with the onset of the LPEE warming, and subsequent cyclicity up to ETM-2 with the overarching goal of testing hypotheses for the origin of both the LPEE and ETM-2. To this end, a multiproxy approach utilizing boron- based (δ11B and B/Ca) and paleothermometry proxies (Mg/Ca and δ18O) is applied to determine the magnitude of shifts in climate, marine carbonate chemistry, and, on shorter timescales, the hydrologic cycle using materials collected from ODP Sites 1209, 1210, 1262 and 1265 (Legs 198 and 208; subtropical Pacific and Atlantic, respectively). Further, I present the first LOSCAR carbon cycle/climate numerical simulations to model the release of carbon to the atmosphere during ETM-2, constrained by the CCD, δ13C, and proxy-based observations of changes in temperature and pH. Additionally, changes in long-term Pacific sea surface pH are used to assess the validity of previously published carbon cycle simulations of the LPEE.

Hydrogenetic Ferromanganese (Fe-Mn) crusts grow from seawater and in doing so sequester elements of economic interest and serve as archives of past seawater chemistry. Ferromanganese crusts have been extensively studied in open-ocean environments. However, few studies have examined continent-proximal Fe-Mn crusts especially from the northeast Pacific. This thesis addresses Fe-Mn crusts within the northeast Pacific California continental margin (CCM), which is a dynamic geological and oceanographic environment. In the first of three studies, I analyzed the chemical and mineralogical composition of Fe-Mn crusts and show that continental-proximal processes greatly influence the chemistry and mineralogy of CCM Fe-Mn crusts. When compared to global open-ocean Fe-Mn crusts, CCM crusts have higher concentrations of iron, silica, and thorium with lower concentrations of many elements of economic interest including manganese, cobalt, and tellurium, among other elements. The mineralogy of CCM Fe-Mn crusts is also unique with more birnessite and todorokite present than found in open-ocean samples. Unlike open-ocean Fe-Mn crusts, carbonate-fluorapatite is not present in CCM crusts. This lack of phosphatization makes CCM Fe-Mn crusts excellent candidates for robust paleoceanography records. The second and third studies in this thesis use isotope geochemistry on select CCM Fe-Mn crusts from four seamounts in the CCM to study past terrestrial inputs into the CCM and sources and behavior of Pb and Nd isotopes over the past 7 million years along the northeast Pacific margin. The second study focuses on riverine inputs into the Monterey Submarine Canyon System and sources of the continental material. Osmium isotopes in the crusts are compared to the Cenozoic Os seawater curve to develop an age model for the samples that show the crusts range in age of initiation of crust growth from approximately 20 to 6 Myr. Lead and neodymium isotopes measured in select Fe-Mn crusts show that large amounts of terrestrial material entered the CCM via the Monterey Canyon from prior to 6.8±0.5 until 4.5 ±0.5 Myr ago. These data combined with reconstructions of the paleo-coastline indicate that incision of the modern Monterey Canyon started around 7 Myr ago. Isotope plots of potential source regions indicate that the source of the material is the border of the southern Sierra Nevada and western Basin and Range. This answers a long-standing and fundamental question about the timing and formation of the Monterey Canyon, the dominant feature of the Monterey Bay. The third study presented here uses the differences in lead and neodymium isotopic values in CCM Fe-Mn crusts over time compared to open-ocean Pacific, North Pacific, and Arctic Ocean Fe-Mn crusts to identify regional time-series trends and sources for these important oceanographic tracers. I found that sediment fluxes and inputs of terrestrial material from North American rivers effects the lead and neodymium isotope composition of regional seawater.