Authors: Samiah Moustafa*, Rutgers University, Asa Rennermalm, Rutgers University, Dirk van As, Geological Survey of Denmark and Greenland, Irina Overeem, University of Colorado Boulder, Marco Tedesco, Columbia University, Thomas Mote, University of Georgia, Lora Koenig, University of Colorado Boulder, Laurence Smith, University of California Los Angeles, Birgit Hagedorn, University of Alaska Anchorage, Ronald Sletten, University of Washington, Andreas Mikkelsen, University of Copenhagen, Bent Hasholt, University of Copenhagen, Dorothy Hall, Michigan State University, Xavier Fettweis, University of Liege, Belgium, Lincoln Pitcher, University of California Los Angeles, Alun Hubbard, Centre for Glaciology
Topics: Cryosphere, Water Resources and Hydrology
Keywords: Greenland, hydrology, discharge, runoff
Session Type: Paper
Start / End Time: 1:20 PM / 3:00 PM
Room: Riverview II, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded
Despite the increasing importance of ice sheet surface ablation in Greenland’s sea-level contribution, a quantitative inter-comparison between modeled and measured meltwater discharge across multiple drainage basins is conspicuously lacking. Here we investigate the accuracy of model discharge estimates from the Modèle Atmosphérique Régionale (MAR v3.5.2) regional climate model by comparison with in situ proglacial river discharge measurements at three West Greenland drainage basins – North River (Thule), Watson River (Kangerlussuaq), and Naujat Kuat River (Nuuk). At each basin, we: 1) determine optimal drainage basin delineations; 2) evaluate MAR at daily, 5-, 10- and 20-day time scales; and 3) identify potential sources for model-observation discrepancies by analyzing model bias relative to data from seven automatic weather stations. Our results reveal that daily discharge is best captured by MAR across the Watson River basin, whilst there is lower correspondence between modeled and observed discharge at the Thule and Naujat Kuat River basins (r2 = 0.4-0.5, p < 0.001). Model agreement with observed is reduced during periods of peak discharge, including the exceptional July 2012 peak events, but is improved at the Thule and Nuuk basins over 5-, 10-, and 20-day means (r2 > 0.7, p < 0.001). Our study highlights the importance of reducing MAR overestimation of surface albedo and warm bias in near surface air temperature, underestimation of cloud cover representation, and adding more realistic runoff delay functions to reduce model error and to improve prediction of Greenland’s contribution to global sea level rise.