Quantifying the spatial representativeness of ice core surface mass balance records using ground-penetrating radar data in Antarctica
Journal Article
Overview
abstract
<p>The future contributions of the Antarctic Ice Sheet to sea level rise will be highly dependent on the evolution of its surface mass balance (SMB), which can offset increased ice discharge at the grounding line. In-situ SMB constraints over annual to multi-decadal timescales come mostly from firn and ice cores. However, although they have a high temporal resolution, ice cores are local measurements of SMB with a surface footprint on the order of cm<sup>2</sup>. Post depositional processes (e.g. wind driven redistribution) can change the initial snowfall record locally and therefore affect our interpretation of the SMB signal recovered. On the other hand, regional climate models have a high temporal resolution but may miss some of the processes at work as a result of their large spatial footprint, on the order of km<sup>2</sup>. Comparisons of ice core and model SMB records often show large discrepancies in terms of trends and variability.</p><p>We investigate the representativeness of a single shallow core record of SMB of the area surrounding it. For this, we use ice-penetrating radar data, co-located with the ice core records examined, to obtain a multi-annual to decadal radar-derived SMB record. We then compare the radar-derived SMB records to the ice core SMB records to determine the surface area that the ice core record is representative of, in terms of mean SMB as well as SMB temporal variability on historical timescales. We examine ice core records situated over the coastal ice rises of East Antarctica, where SMB is high and spatially heterogeneous, as well as over the interior of the West Antarctic Ice Sheet, where SMB is more uniform spatially. By comparing these two contrasting regions in terms of SMB, we will determine whether a general rule of thumb can be obtained to determine the spatial representativeness of an ice core SMB record.</p>
