Comparison of radar derived to model derived annual snow accumulation in North-Central Greenland

Data collected with a FMCW radar operated on a surface traverse along the ice divide in North-Central Greenland in 2007 has been processed to reveal annual internal layers in the ice, and the depths of individual internal annual layers along the traverse has been determined. The two endpoints of the 375 km traverse are NGRIP and NEEM ice camps, where ice cores have been previously recovered. The ice core data from these locations provide firn/ice density profile measurements which are used to calculate the radar signal propagation speed through the ice to increase the range accuracy of the internal layer depths, and the age-depth information is used to confirm the annual chronology of the radar-detected internal layers. At approximately 12 m intervals along the traverse, the difference in depth of adjacent internal layers was determined for each year back to 1958, and the thickness between adjacent internal layers is converted to a water equivalent snow accumulation using an interpolated snow density profile between NGRIP and NEEM.

The radar-derived annual snow accumulation rates between 1958 and 2007 have been compared to the gridded annual accumulation rate data of Burgess et al. (2010), which uses the solid precipitation output from the Polar MM5 and spatially calibrates the model output using ice core and meteorological station data. Burgess et al. provide annual accumulation between 1958 and 2007 on a 1.25 km raster grid covering all of Greenland. We have used ArcGIS to project the radar-derived annual accumulation measurements for each year between 1958 and 2007 onto the same grid used by Burgess et al., and have averaged the radar-derived measurements within each grid cell to allow a comparison between the two data sets. A comparison with the ice core derived annual accumulation at NGRIP and NEEM has also been performed.

Results generally show a good agreement between the Burgess et al. accumulation and the radar-derived data with the exception of a few years. Overall, the Burgess et al. accumulation has a positive bias in comparison to the radar-derived accumulation along the traverse route, and a positive bias at the two ice core sites.