Modeling glacier-rock-climate interactions: Moraine deposition, stagnation events, and supraglacial debris

Open Access
Vacco, David Alfred
Graduate Program:
Earth Sciences
Doctor of Philosophy
Document Type:
Date of Defense:
March 09, 2009
Committee Members:
  • Richard B Alley, Dissertation Advisor
  • Richard B Alley, Committee Chair
  • Rudy Slingerland, Committee Member
  • Sridhar Anandakrishnan, Committee Member
  • David Pollard, Committee Member
  • Derek Elsworth, Committee Member
  • glacier
  • glacial
  • moraines
  • modeling
  • supraglacial
  • subglacial
  • debris
  • dynamics
New, quantitative information on past climates can be extracted from observations of glacial deposits through joint interpretation using a new glacier-flow model that simulates debris transport and deposition. Locations of moraine sets in some northern-hemisphere locations including Greenland and North America are consistent with the hypothesis that temperature anomalies associated with millennial cold events occurred primarily in wintertime. Extensive forward simulations show that glacier stagnation is more likely for larger and faster warming, suggesting that glacial deposits can be used to learn the amplitude of abrupt-warming events. Additional data will be required on supraglacial-debris effects before such temperature-change reconstructions are quantitatively accurate for heavily debris-laden glaciers, but existing understanding is sufficient to show that well-defined moraine ridges lacking extensive, homogeneous, hummocky deposits just upglacier represent climatically significant events and not landslide-triggered advances. The first chapter of this work involves the coupling of a dynamic glacier model with a dynamic subglacial sediment package, allowing sediment to be deposited sub- and pro-glacially. The result is a numerical glacier model that deposits moraines based on its terminus position. Experiments were run, varying the climate forcing on the glacier, to observe the resulting moraines and compare them to moraine sets observed in nature. Successful simulation of moraine sets in Greenland driven by ice-core climate records required notable damping of the millennial signal, as required by the hypothesis that millennial temperature fluctuations occurred primarily in wintertime. The second chapter involves testing of the hypothesis that, if the influence of valley hypsometry is controlled for, stagnation of terminal regions of a glacier require warming larger and faster than some threshold, although with some tradeoff between rate and size of warming. Quantitative modeling experiments were run using simplified valley shapes, in order to isolate a cause and effect relationship between hypsometry, climate change, and glacial stagnation. Stagnation was found to be more likely for rougher beds with lower mean slopes, as well as for larger and faster warmings. The results indicate that it may be possible to quantify ancient climate changes based on observations of glacial stagnation deposits. The third chapter expands on the study of glacier stagnation, to include transport of supraglacial debris and its thermal shielding effects on glacial melting. Modern observations and process understanding show that extensive supraglacial debris favors stagnation. The ice-flow model from the first two chapters was expanded to include “point” sources (landslides) and distributed sources of supraglacial debris, transport of the debris, and the thermal effects of the debris. The results show that landslide deposits sufficient to cause notable glacier advance cause stagnation events while remaining widely distributed on the glacier surface, thus producing hummocky topography upglacier of any moraine, such that a single moraine ridge formed without such hummocky deposits represents a climatic event rather than a landslide. Model runs show that if sufficient supraglacial debris is supplied from distributed sources, the glacier will be longer and more-likely to experience stagnation following warming than an equivalent clean glacier. A rich range of behavior is simulated, with sensitivity to some parameters that are not yet well-constrained by field glaciology. Quantitative estimation of the size and rate of warming responsible for the observed pattern of stagnation deposits, or lack thereof, in glaciated valleys in a region remains a realistic possibility, but will require improved understanding of debris dynamics.