Georadar‐derived estimates of firn density in the percolation zone, western Greenland ice sheet

Abstract

[1] Greater understanding of variations in firn densification is needed to distinguish between dynamic and melt-driven elevation changes on the Greenland ice sheet. This is especially true in Greenland's percolation zone, where firn density profiles are poorly documented because few ice cores are extracted in regions with surface melt. We used georadar to investigate firn density variations with depth along a $70 km transect through a portion of the accumulation area in western Greenland that partially melts. We estimated electromagnetic wave velocity by inverting reflection traveltimes picked from common midpoint gathers. We followed a procedure designed to find the simplest velocity versus depth model that describes the data within estimated uncertainty. On the basis of the velocities, we estimated 13 depth-density profiles of the upper 80 m using a petrophysical model based on the complex refractive index method equation. At the highest elevation site, our density profile is consistent with nearby core data acquired in the same year. Our profiles at the six highest elevation sites match an empirically based densification model for dry firn, indicating relatively minor amounts of water infiltration and densification by melt and refreeze in this higher region of the percolation zone. At the four lowest elevation sites our profiles reach ice densities at substantially shallower depths, implying considerable meltwater infiltration and ice layer development in this lower region of the percolation zone. The separation between these two regions is 8 km and spans 60 m of elevation, which suggests that the balance between dry-firn and melt-induced densification processes is sensitive to minor changes in melt.

Constraining variable density of ice shelves using wide-angle radar measurements

“…A complementary approach is to exploit the density dependence of radio-wave propagation speed. The principle underlying the technique involves illuminating a reflector with different ray paths such that both the reflector depth and the radiowave propagation speed may be calculated using methods such as the Dix inversion (Dix, 1955), semblance analysis (e.g., Booth et al, 2010Booth et al, , 2011, interferometry (Arthern et al, 2013), or traveltime inversion based on ray tracing (Zelt and Smith, 1992;Brown et al, 2012).…”

“…1 is common in seismic investigations, and multiple techniques exist for deriving the velocities from recorded traveltimes (Yilmaz, 1987). Similar to what has been done for wide-angle radar measurements in Greenland (Brown et al, 2012), we follow a variation of the approach delineated by Zelt and Smith (1992). Brown et al (2012) measured common midpoint returns with a 100 MHz radar.…”

“…Similar to what has been done for wide-angle radar measurements in Greenland (Brown et al, 2012), we follow a variation of the approach delineated by Zelt and Smith (1992). Brown et al (2012) measured common midpoint returns with a 100 MHz radar. They used a ray tracing forward model and inferred bulk densities of individual intervals (hereafter interval densities) by inverting reflector depths and interval velocities for single reflectors from top to bottom (a.k.a.…”

“…The traveltime is a non-linear function of the model parameters (and hence the inversion results may be non-unique) because L depends both on the initially unknown radio-wave propagation speed and the reflector depth. The velocity between two radar reflectors is often represented as piecewise constant or piecewise linear (Brown et al, 2012), making the model parameters m v either the interval velocities or the interval velocity gradients, respectively. Here, we introduce additional constraints from Hubbard et al (2013) who fit a depth profile of density of the form:…”

Multi-Channel Ground-Penetrating Radar for the Continuous Quantification of Snow and Firn Density, Depth, and Accumulation

“…I consider these estimates to be continuous because the shot or CMP interval is dense compared to the length of radar transects; rather than sparse CMP sampling which have intervals on the tens-kilometer scale as in the work by Brown et al (2012). The continuous acquisition strategy allows for estimates which capture subtle to large changes in the snow accumulation and firn density gradients.…”