This dissertation presents results from three studies that address major scientific questions in glacial geology and paleoclimatology for the late Pleistocene and Holocene using relatively new geochemical and statistical techniques. Each of the studies attempts to answer a longstanding question in the respective field using geochemical or statistical methods that have not been applied to the problem thus far. A longstanding question in glaciology is the nature and mechanism of the so- called "Heinrich events" of the last ~60 ka. These massive iceberg discharge events into the North Atlantic from the partial breakup of the Laurentide Ice Sheet are identified from distinct ice rafted debris and detrital carbonate layers in marine sediment cores. The mechanism associated with the initiation of these events is commonly thought to be related to internal ice sheet instabilities. However, Heinrich events consistently occur following a long cooling trend that culminates in an extreme cold event, thus suggesting a possible triggering mechanism by climate. Recent modeling work has proposed an oceanic mechanism associated with ocean warming, but no physical evidence has been made available to date. To test this ocean-warming hypothesis, we measured temperature sensitive trace metals and stable isotopes in benthic foraminifera from a sediment core collected in the western North Atlantic that spans the last six Heinrich events and compared our results to climate model simulations using CCSM3. Our results show subsurface warming occurred prior to or coeval with nearly all of the Heinrich events of the last ~60 ka, thus implicating subsurface ocean warming as the main trigger of these rapid breakups of the Laurentide Ice Sheet. In the field of glacial geology a longstanding question has been the timing of alpine glacial advances during the Holocene. A number of studies have interpreted several Holocene glacial advances in western North America, but age control is based largely on relative dating techniques, which have been shown to be in error by up to 10,000 yrs in some cases. Based on 124 ¹⁰Be surface exposure ages from twenty cirque moraines in ten mountain ranges across western North America, glacier were retreating from moraine positions during the latest Pleistocene or earliest Holocene and not throughout the Holocene epoch as previously assumed, thus requiring a refined interpretation of Holocene glacial activity in western North America and the associated climate forcing. In the field of paleoclimatology a question regarding how global temperature varied over the entirety of the Holocene epoch has remained to be answered for some time. While many temperature reconstructions exist for the last 2000 years, a full Holocene temperature stack does not exist, despite its potential utility of putting modern climate change into a full interglacial perspective. Based on a global composite of 73 proxy based temperature record, a Holocene temperature stack was constructed and used to demonstrate that a general cooling of ~1°C has occurred from the early to mid Holocene and that centennial and millennial scale variability is modest. We account for both temperature calibration and chronologic uncertainties using a Monte Carlo based approach. Our results are consistent with prior reconstructions of the last 2000 years and now allow for a full Holocene temperature perspective for evaluation with present and future climate change. / Graduation date: 2011 / Access restricted to the OSU Community, at author's request, from May 5, 2011 - May 5, 2012