loading page

Assessing Satellite-derived Inter-annually Varying Snow/Firn Density Estimates Over the Greenland Ice Sheet during 2003-2009
  • +4
  • Xiaoli Su,
  • C. K. Shum,
  • Kenneth C. Jezek,
  • Jun-Yi Guo,
  • Ian M. Howat,
  • Chung-Yen Kuo,
  • Zhicai Luo
Xiaoli Su
Huazhong University of Science and Technology

Corresponding Author:[email protected]

Author Profile
C. K. Shum
Ohio State University
Author Profile
Kenneth C. Jezek
Ohio State University
Author Profile
Jun-Yi Guo
Ohio State University
Author Profile
Ian M. Howat
Ohio State University
Author Profile
Chung-Yen Kuo
National Cheng Kung University
Author Profile
Zhicai Luo
Huazhong University of Science and Technology
Author Profile

Abstract

Knowledge of snow/firn density is important for deriving ice sheet mass change from satellite altimetry and for surface mass balance modeling. However, snow/firn densities are largely unknown over the Greenland ice sheet (GrIS) away from isolated direct measurements in boreholes. Density assumptions are widely used when converting volume change from satellite altimetry into mass change, which could introduce errors. Here we extract the inter-annual anomalies of mass change from Gravity Recovery And Climate Experiment (GRACE) data and the refined ice elevation change through the repeat-track analysis of Environmental Satellite (Envisat) altimetry data retrieved using both the ICE1 and ICE2 waveform retracking algorithms. By combining these two types of inter-annual anomalies (GRACE+ICE1 and GRACE+ICE2), we investigate the inter-annually changing snow/firn density estimates over the GrIS during the 2003–2009 period. Our results demonstrate that satellite-derived density is relatively greater over Western GrIS, with magnitude falling between 300 “kg” ⁄“m” ^“3” and 917 “kg” ⁄“m” ^“3” occupying more than 71% of the GrIS. At the regional scale, GRACE+ICE1 derived density agrees well with that from density profiles of 9 ice cores at Summit Station, Central GrIS, with relative errors less than 5%. Satellite-derived densities are further compared with that from 110 ice cores mostly along the central ice divide. The percentage of GRACE+ICE1 derived densities with relative errors less than 20% exceeds 84%, as opposed to 41% for GRACE+ICE2. Satellite-derived densities could possibly underestimate the inter-annually varying snow/firn density over Southern GrIS. This study may provide constraints on the currently applied density assumptions for the GrIS.