Drivers of Firn Density on the Greenland Ice Sheet Revealed by Weather Station Observations and Modelling

Recent Arctic atmospheric warming induces more frequent surface melt in the accumulation area of the Greenland ice sheet. This increased melting modifies the near‐surface firn structure and density and may reduce the firn's capacity to retain meltwater. Yet, few long‐term observational records...

Full description

Bibliographic Details
Published in:Journal of Geophysical Research: Earth Surface
Main Authors: Vandecrux, Baptiste Robert Marcel, Fausto, R. S., Langen, P. L., Van As, D., MacFerrin, M., Colgan, W. T., Ingeman-Nielsen, Thomas, Steffen, K., Jensen, N. S., Møller, M. T., Box, J. E.
Format: Article in Journal/Newspaper
Language:English
Published: 2018
Subjects:
Online Access:https://orbit.dtu.dk/en/publications/58cfafb1-3656-40b3-8e7f-66914a10631d
https://doi.org/10.1029/2017JF004597
https://backend.orbit.dtu.dk/ws/files/161703566/Vandecrux_et_al_2018_Journal_of_Geophysical_Research_3A_Earth_Surface.pdf
Description
Summary:Recent Arctic atmospheric warming induces more frequent surface melt in the accumulation area of the Greenland ice sheet. This increased melting modifies the near‐surface firn structure and density and may reduce the firn's capacity to retain meltwater. Yet, few long‐term observational records are available to determine the evolution and drivers of firn density. In this study, we compile and gap‐fill Greenland Climate Network (GC‐Net) automatic weather station data from Crawford Point, Dye‐2, NASA‐SE and Summit between 1998 and 2015. These records then force a coupled surface energy balance and firn evolution model. We find at all sites except Summit that increasing summer turbulent heat fluxes to the surface are compensated by decreasing net radiative fluxes. After evaluating the model against firn cores, we find that, starting from 2006, the density of the top 20 m of firn at Dye‐2 increased by 11%, decreasing the pore volume by ‐18%. Crawford Point and Summit show stable near‐surface firn density over 1998‐2000 and 2000‐2015 respectively while we calculate a ‐4% decrease of firn density at NASA‐SE over 1998‐2015. For each year, the model identifies the drivers of density change in the top 20 m firn and quantifies their contributions. The key driver, snowfall, explains alone 72 to 92% of the variance in day‐to‐day change in firn density while melt explains from 7 to 33 %. Our result indicates that correct estimates of the magnitude and variability of precipitation are necessary to interpret or simulate the evolution of the firn.