Nick Holschuh


D1. Improved Marie Byrd Land Topography

data The final gridded topography can be downloaded here. (NetCDF)

The basal topography produced for my EPSL paper published in 2014 was designed to capture the DeVicq glacial trough, which provides a deep passageway for ice retreat into the interior of Marie Byrd Land. Figure (1) shows how the updated topography differs from Bedmap2. Through most of Bedmap2, the bed is artificially shallow, failing to capture the deep ice reserves off the flanks of the mountain ranges in this region. You will note non-zero differences outside of the data collection area, these are the result of differences in the interpolation method used. Our interpolation was designed to produce results true to the data at locations with geophysical coverage, but can diverge from the results of B2's method as you move away from the survey data.

Figure (2) indicates the locations where new data were collected, and Figure (3) shows the data derived bed elevations versus the grid values for B2 and for our updated topography. Plotted in the lower panels are the elevation differences in black, the low-frequency elevation differences in red, and the mean difference in blue. As you can see, our topography does a much better job of capturing the average topographic signal, most notably in the orange and red lines which pass over the DeVicq trough.

This topography was designed for a 5km resolution modeling study, so detailed features have been smoothed out to capture the gross behavior of the system. Small scale errors also exist at the edge of our MBL domain, where the updated topography was stitched into Bedmap2. If attempting to use these data at higher resolution, it is important to keep these features in mind.

© Nick Holschuh - January 2016