A characterization of Greenland Ice Sheet surface melt and runoff in contemporary reanalyses and a regional climate model

For the Greenland Ice Sheet (GrIS), large-scale melt area has increased in recent years and is detectable via remote sensing, but its relation to runoff is not known. Historical, modeled melt area and runoff from Modern-Era Retrospective Analysis for Research and Applications (MERRA-Replay), the Int...

Full description

Bibliographic Details
Published in:Frontiers in Earth Science
Main Authors: Richard eCullather, Sophie eNowicki, Bin eZhao, Lora eKoenig
Format: Article in Journal/Newspaper
Language:English
Published: Frontiers Media S.A. 2016
Subjects:
Q
Online Access:https://doi.org/10.3389/feart.2016.00010
https://doaj.org/article/3dea6d877af54ebfbeb4fd861ba3cbed
Description
Summary:For the Greenland Ice Sheet (GrIS), large-scale melt area has increased in recent years and is detectable via remote sensing, but its relation to runoff is not known. Historical, modeled melt area and runoff from Modern-Era Retrospective Analysis for Research and Applications (MERRA-Replay), the Interim Re-Analysis of the European Centre for Medium Range Weather Forecasts (ERA-I), the Climate Forecast System Reanalysis (CFSR), the Modèle Atmosphérique Régional (MAR), and the Arctic System Reanalysis (ASR) are examined. These sources compare favorably with satellite-derived estimates of surface melt area for the period 2000-2012. Spatially, the models markedly disagree on the number of melt days in the interior of the southern part of the ice sheet, and on the extent of persistent melt areas in the northeastern GrIS. Temporally, the models agree on the mean seasonality of daily surface melt and on the timing of large-scale melt events in 2012. In contrast, the models disagree on the amount, seasonality, spatial distribution, and temporal variability of runoff. As compared to global reanalyses, time series from MAR indicate a lower correlation between runoff and melt area (r2 = 0.805). Runoff in MAR is much larger in the second half of the melt season for all drainage basins, while the ASR indicates larger runoff in the first half of the year. This difference in seasonality for the MAR and to an extent for the ASR provide a hysteresis in the relation between runoff and melt area, which is not found in the other models. The comparison points to a need for reliable observations of surface runoff.