Date of Award

Spring 5-2021

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Master of Science (MS)


Quaternary and Climate Studies


Karl Kreutz

Second Committee Member

Dominic Winski

Third Committee Member

Kristin Schild


Temperature changes in glaciated regions are of immediate concern for estimates of future sea level rise. Alaska and the surrounding region contain over 40 mm of potential sea level rise in its many alpine glaciers, which are experiencing some of the highest rates of mass loss globally. However, records of both past and present temperatures in the region’s alpine sectors are sparse and limited in temporal and spatial extent. Here I examine the application of MODIS land surface temperatures and layers of refrozen melt in ice cores as temperature indicators in the St. Elias and Alaska Ranges. First, I find that a previously observed cold bias in MODIS LSTs relative to in situ temperatures is likely due to a discrepancy between surface and air temperatures over glaciated surfaces. The bias is not a result of MODIS’ large footprint (1 km2), nor is it introduced by poorly constrained snow emissivity values used in the LST calculation, although the role of emissivity in its amplification remains unknown. Although MODIS LSTs may be used to supplement in situ temperatures, factors affecting the relationship between surface and air temperatures must be accounted for. Second, I find that melt layers since 2000 in an ice core from Eclipse Icefield (St. Elias Range) do not correspond to years of high temperatures. However, years of high surface ablation do, suggesting that complex percolation dynamics, rather than the surface environment, control the preservation of an annual melt record in ice cores. Although modern melt layers do not reflect temperature, Eclipse may transition between percolation regimes with modest temperature changes and melt layers from the pre-industrial Holocene may yet provide a valuable record of past temperatures. Lastly, I find that a novel analytical technique for melt layer identification using bubble number density agrees with established methods, validating their continued use. Although bubble number density cannot be used to unequivocally identify highly thinned melt layers at depth, the method shows promise with improved accuracy in depth measurements and signal-to-noise ratio.


As of 2002, Degree of Master of Science (MS) Quaternary and Climate Studies published under the auspices of the Climate Change Institute.

Files over 10MB may be slow to open. For best results, right-click and select "save as..."