ABSTRACT ARCTIC AND NORTH ATLANTIC PALEO-ENVIRONMENTAL RECONSTRUCTIONS FROM LAKE SEDIMENTS MAY 2017 GREGORY A. DE WET, B.Sc., BATES COLLEGE M.Sc., UNIVERSITY OF MASSSCHUSETTS, AMHERST Ph.D., UNIVERSITY OF MASSACHUSETTS, AMHERST Directed by: Drs. Raymond S. Bradley and Isla S. Castañeda There are few fields in the discipline of Earth Science that hold more relevancy in 2017 than studies of earth's climate. Called the "perfect problem" considering its complexity and magnitude, climate change will continue to be one of the greatest challenges humanity will face in the 21st century. And while numerical models provide valuable information on conditions in the future, the results from these simulations must be contextualized by the past. Climate reconstructions from paleo-environmental archives, even from periods colder or different from what we are experiencing today, provide that context. Every piece of information gleaned from these studies informs our collective knowledge of the climate system. In some cases, environmental reconstructions may include proxies for anthropogenic as well as climatological information, directly addressing one of the most important questions in climate science: how does changing climate affect humans? The following chapters of this dissertation are exercises in trying to understand climate change in one of the most climatically sensitive regions on earth - the high northern latitudes. While my doctoral studies cover a wide range of timescales, it is broadly unified by the focus on the Arctic. In some cases, my research spans multiple glacial/interglacial cycles, in others the concentration is on the past few thousand years. In all cases, however, the goal is to utilize lacustrine sedimentary archives to inform our knowledge of climatic change in this important region. One of the most rewarding aspects of this Ph.D. has been the creativity I have been afforded in working towards that goal. Chapter One of this dissertation involves the analysis of organic molecules, specifically bacterial membrane lipids called branched glycerol dialkyl glycerol tetraethers (brGDGTs), to sediments from Lake El'gygytgyn. These biomarkers allow for a quantitative reconstruction of temperature from multiple interglacial periods over a million years ago. Our data suggests that "super interglacial" Marine Isotope Stage 31 may have in fact lasted much longer than previously thought in the Arctic, with implications for Antarctic ice sheet extent and CO2 concentrations highly relevant to our future. Though the extraction and analysis of these biomarkers is expensive and time-consuming, the data is highly valuable and informative. Conversely in some cases more "quick and dirty" techniques are sufficient to reconstruct important processes or factors back through time (e.g. the presence of a glacier within a lake catchment) at relatively low cost, and therefore are highly useful. The chapters dealing with Lake Gjøavatnet in Svalbard (Chapter Two) or the application of Fourier Transform Infrared (FTIR) spectroscopy to lake sediments (Chapter Five) exhibit how such techniques can also be highly informative. In Chapter Two, using a combination of core-scanning data and relatively simple destructive analyses (%loss-on-ignition, bulk carbon isotopes), we reconstructed past fluctuations in glacier extent over the Holocene, as well as identified intervals that may have been related to freshwater pulses in Fram Strait. Chapter Five is focused mainly on expanding and confirming the use of FTIR spectroscopy to Arctic lake sediments to reconstruct biogenic silica and organic matter concentrations through time. Though potentially a less direct climate proxy than paleotemperature from brGDGTs, this new technique allows for more rapid analyses using less sediment than previous methods, a valuable advance. Chapters Three and Four are in many ways the confluence of these earlier stages of my Ph.D., where we apply a wide range of proxies to answer questions related to climate and human population dynamics. The use of biomarkers is expanded in these projects, where a broad suite of organic molecules are used to reconstruct both climate and other paleo-environmental conditions, including vegetation changes, variations in pH, and potentially anthropogenic influences. Coupled with some of the more "basic" techniques described above, we characterize an environmental disturbance in the Faroe Islands ~2200 years ago that may be evidence for the first appearance of humans in the archipelago (Chapter Three). In Chapter Four, we explore the application of brGDGTs (among other proxies) to reconstruct temperature change in southwest Greenland during the period of Norse settlement and subsequent abandonment. This work is part of an ongoing investigation into the efficacy and calibration of this promising proxy in a locale where climate change likely had a dramatic impact on the fragile communities living there. In summary, I have not attempted to unify these chapters into a single climatological context (though some of my work, such as in the Faroes and Greenland, is highly related). Instead I present them as they are, individual projects that each have their own goals and merits within the broad framework of paleoclimatology. As I mentioned above, one of my favorite things about this field is the creativity we are afforded in our attempts to answer questions about the past. This Ph.D. has been an exercise in that creativity, focused in the high northern latitudes, and centered around the archive of lake sediments.

Future impacts from climate change can be better understood by placing modern climate trends into perspective through extension of the short instrumental records of climate variability. This is especially true for extreme climatic events, such as extreme precipitation and wildfires, as the period of instrumental records provides only a few examples and these have likely have been influenced by anthropogenic warming. Multi-parameter records showing the past range of climate variability can be obtained from lakes. Lakes are particularly good recorders of climate variability because sediment from the surrounding environment accumulates in lakes, making them sensitive recorders of climate variability and providing high-resolution histories of local environmental conditions in the past. In some cases, such as at Basin Pond, sediment is persevered efficiently enough to produce distinguishable annual laminations (varves) in the sedimentary record. The varved record at Basin Pond was used to construct an accurate, highly-resolved age-to-depth model over the past 300 years. Using a multi-proxy analysis, including organic biomarker analysis of molecular compounds and sedimentological features preserved in the sediment record, a history of environmental and climatic change at Basin Pond was constructed. These analyses were compared with the record of known extreme events (from instrumental measurements and historical documents), including 129 years of high-resolution precipitation and temperature meteorological data, 19 tropical systems over the past 145 years, and two known wildfire events over the past 200 years. Long-term trends in precipitation, including the increase in precipitation seen throughout the last half of the 20th century and the drought of the 1940's, were captured in the analysis of long-chain n-alkane distributions and through varve thickness measurements obtained through X-Ray Fluorescence analysis. Furthermore, Polycyclic Aromatic Hydrocarbons (PAHs), a class of organic compounds that can be used to trace combustion activity, were found in abundance in the Basin Pond sedimentary record. Peaks in the abundances of two PAHs (retene and chrysene) and the ratio retene/(retene + chrysene) were found to be highly correlated with the known wildfire events occurring in the historical period, giving promise as using these compounds and ratio as a robust proxy for regional wildfire events in the northeastern U.S.

The last glacial maximum, ca. 21,000 years ago, caused a fall in eustatic sea level of ca. 120 m below present. The low-gradient, shallow Sunda Shelf, Southeast Asia was subaerially exposed during this sea-level lowstand and experienced rising sea level thereafter. Sea level rose to a +1.3––5 m highstand ca. 6,500 cal yr BP, and then fell to modern sea level. The objective of this research is to characterize environmental change on the Sunda Shelf in response to the post-glacial rising eustatic sea level. To address this objective, six gravity cores were collected along a transect crossing the paleo-Chao Phraya incised river valley complex between peninsular Malaysia and southern Vietnam. Thirteen AMS radiocarbon samples, 130 bulk sediment magnetic susceptibility samples (BMS), 66 X-ray fluorescence spectrometry (XRF) samples, and 54 samples for the analysis of foraminiferal assemblages were used to characterize change in the depositional environments of the cored sediments. BMS, XRF, and foraminiferal analysis distinguish two main units. Unit 1 is found in the lower part of the cores and typically contains more terrestrial material than sediments further up-core as shown by higher BMS values, higher % Ti, % Al, % Fe, and lower indicators of marine influence, for example, lower % Ca, % planktonic foraminifera, and percentages of deeper water benthic foraminifera such as Heterolepa dutemplei. Unit 2 is characterized by a significant increase in % Ca, % Heterolepa dutemplei and % planktonic foraminifera ca. 6,500 cal yr BP. Unit 1 is consistent with shallower water depths and is part of a transgressive systems tract (TST, ca. 1 m thick) that terminates ca. 6,500 cal yr BP. Unit 2 represents the overlying highstand systems tract (HST, ca. 1 m thick) and is characterized by an increase in % planktonics, % Ca, and shifts in benthic foraminiferal assemblages, indicating deeper water conditions than the sediments below. This shift from a TST to a HST ca. 6500 cal yr BP is consistent with the Sunda Shelf sea-level record.