Assessing Greenland Ice Sheet meltwater losses at the pixel and drainage basin scale
Description
TitleAssessing Greenland Ice Sheet meltwater losses at the pixel and drainage basin scale
Date Created2017
Other Date2017-05 (degree)
Extent1 online resource (xii, 207 p. : ill.)
DescriptionThe Greenland Ice Sheet (GrIS) is expected to increase its contributions to sea level rise with atmospheric warming, and it is important to accurately predict future sea level change. Surface meltwater runoff losses, modulated by surface albedo, are two dominant uncertainties in future GrIS sea level rise estimates. The first component of this study characterizes surface albedo in the lower ablation zone, a key variable controlling the surface energy and mass balance of the GrIS, and an important parameter in regional climate models (RCMs). This analysis is expanded in a second study to evaluate satellite albedo retrievals and assess its ability to resolve sub-pixel spatial variability of ablation area albedo. In situ spectral albedo data collected along a transect, Moderate Resolution Imaging Spectroradiometer (MODIS) daily albedo product, and high spatial resolution WorldView-2 (WV-2) data are utilized in these two studies. The results show that the distribution of dominant ice surface types (e.g., snow, bare ice, light-absorbing impurities, and streams) act as an additional mechanism for controlling ablation zone albedos. This can significantly impact seasonal and inter-annual changes in ablation zone albedo, and subsequent melt. These findings have important implications for current RCMs, which don’t fully integrate a seasonally evolving ice surface type’s albedo scheme. The second study demonstrates over spatially heterogeneous surfaces, such as in the ablation zone, that a multiple ‘point-to-pixel’ comparison, utilizing multiple ground albedo observations coinciding with a satellite pixel, is superior to the frequently used single ‘point-to-pixel’ comparison. This points to the significance of evaluating the spatial representativeness of ground albedo sites (e.g., automatic weather stations) prior to validation of satellite or model-derived albedos. The second component of this study quantifies meltwater runoff losses, a dominant, yet understudied term of GrIS mass loss, at the drainage-basin scale. To do this, the Modèle Atmosphérique Régionale (MAR) RCM discharge estimates are compared with proglacial river discharge observations at three drainage basins – Thule, Watson, and Nuuk – located north-to-south in west Greenland. I find that MAR poorly resolves daily discharge variability in the Nuuk and Thule basins, but is better able to capture variability at longer time averages. Model-observation agreement is reduced during peak discharge events. The model-observation discharge discrepancies are likely due to an underestimation of cloud cover, from an overestimation of downward shortwave radiation. The discrepancies of model and measurements during peak discharge events is important to understand as they are expected to occur more frequently with continued warming. In a fourth study, annual and daily peak river discharge was unprecedented at all basins in the extreme melt season of 2012. Exceptional flows in all three rivers were observed corresponding with two ice sheet wide surface melt episodes in mid- and late-July 2012. These results suggest the need to further study runoff processes at the local-, basin- and continental-scale not fully captured by current RCMs. These four studies collectively contribute information that will allow for better understanding of Greenland’s complex hydrologic system. Finally, these studies provide the framework to improve physical representation of meltwater runoff and albedo components used in RCMs to project changes in Greenland’s mass loss, and subsequent contributions to sea level rise.
NotePh.D.
NoteIncludes bibliographical references
Noteby Samiah Moustafa
Genretheses, ETD doctoral
Languageeng
CollectionGraduate School - New Brunswick Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.
Statistical Profile