Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier


Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program


Year Degree Awarded


Month Degree Awarded


First Advisor

Robert M. DeConto

Subject Categories

Climate | Geology | Glaciology


Sea level rise is one of the major social and environmental challenges that threatens modern civilization, yet the response of polar ice sheets to future warming is deeply uncertain. Mass loss from the Antarctic Ice Sheet is projected to dominate global sea level rise in the near future, but how much, and when, remains a key unknown. The challenges associated with projecting Antarctica’s future sea level contribution are derived from a knowledge gap of physical ice sheet processes in a world warmer than today, and a lack of understanding of climatic thresholds that drive potentially irreversible retreat.

Future and even modern climatic conditions are unprecedented within the last few million years; therefore, we must look to the geologic record for a glimpse of prospective Earth landscapes and climates. Past ‘warm periods’ (characterized by elevated atmospheric CO2 and surface temperatures) can provide a window into the feedbacks and instabilities that govern ice sheet dynamics under a fundamentally different climatic state. In this work, I integrate process-based ice sheet modeling, climate modeling, and remote sensing observations along with geologic data to explore the stability and behavior of the Antarctic Ice Sheet during past warm periods.

In Chapter 3, I investigate Antarctic ice sheet and climate evolution during the mid-Miocene, a time period about 17 to 14 million years ago characterized by an epoch of peak global warmth followed by glacial expansion. Coupled ice sheet and climate model scenarios under varying boundary conditions provide continent-wide context for localized geologic paleoclimate and vegetation records. I combine model simulations with geologic constraints to make inferences about past CO2, tectonic uplift, and ice sheet fluctuations across this key time period. Chapter 3 has been published in EPSL (Halberstadt et al., 2021), with coauthors H. K. Chorley, R. H. Levy, T. Naish, R. M. DeConto, E. Gasson, and D. E. Kowalewski.

In Chapter 4, I employ a similar modeling approach to address a long-standing data-based discrepancy regarding the stability of the Antarctic Ice Sheet during past warm periods. Marine data reconstruct periodic large-scale marine ice sheet fluctuations since the mid-Miocene (suggesting a dynamic ice sheet response to past increases in temperature and atmospheric CO2) while preserved terrestrial landforms reflect persistent cold conditions (implying that the Antarctic Ice Sheet was largely insensitive during past warm periods). I use high-resolution climate modeling under warm interglacial boundary conditions to reconcile terrestrial persistent cold conditions with receded or collapsed ice sheets during past warm periods. Chapter 4 will be published in Geology, with coauthors D. E. Kowalewski and R. M. DeConto.

In Chapter 5, I focus on the modern ‘warm period’ using the satellite observational record. Increased surface meltwater generated on ice shelves fringing the Antarctic Ice Sheet can drive ice shelf collapse, so it is crucial to quantify the historical and current evolution of surface melt to better understand vulnerability of ice shelves. Because in situ observations of surface melt are very sparse given the vast size of Antarctica, I use remotely sensed data to map surface meltwater features from the satellite record. I developed a new methodology to identify meltwater from multispectral satellite images, using Google Earth Engine to train supervised image classification algorithms and identify surface lakes. This work paves the way for automating lake identification at a continental scale throughout the satellite observational record. Chapter 5 is published in Remote Sensing (Halberstadt et al., 2020), with coauthors C. J. Gleason, M. S. Moussavi, A. Pope, L. D. Trusel, and R. M. DeConto.


Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.