Preparing a new installation after clearing accumulated snow on the monitoring equipment. Photo: Craig Stewart
Researchers are using custom-built instruments to monitor ice shelf melting in a rapidly melting region of the Ross Ice Shelf, surrounding Ross Island.
Floating ice shelves stabilise the Antarctic ice sheet by slowing the flow of grounded ice. But ocean- driven basal melting (at the bottom) of ice shelves can thin them, reducing this buttressing effect. The meltwater from the ice sheets also influences the ocean, locally and globally. Yet despite the importance of basal melting to both the ice sheet and the ocean, basal melting is rarely observed directly.
This has changed in a joint project led by Dr Craig Stewart, with collaborators Ollie Twigge (NIWA), Keith Nicholls (British Antarctic Survey) and Wolfgang Rack (University of Canterbury). The research is funded by the Antarctic Science Platform and Dr Stewart’s Marsden fast-start project.
The team established 12 monitoring stations on the Ross Ice Shelf. Melting is measured using custom radar instruments called Autonomous phase Sensitive Radio Echo Sounders (ApRES). These radars provide hourly estimates of ice shelf thickness throughout the year, sending data back to New Zealand by satellite.
However, the data collection has had its challenges, with instrument failures during the first year of deployment, causing gaps in the records. Such challenges are to be expected when operating sensitive electronic instruments through the Antarctic winter with temperatures ranging from -25C to -45C.
After a successful 2023/24 Antarctic field season, all instruments are now operational. The team accessed all 12 deployed instruments last summer, with the support of helicopter, a fixed wing aircraft and Scott Base staff. At each site, the team downloaded data, upgraded firmware, installed new memory cards and the power hardware. In addition, at sites intended for longer operation, the full system was excavated and reinstalled nearer to the surface, to allow for future access/recovery.
Each site visit took between 2-7 hours, as the accumulated snow needs to be removed (by hand!) before the equipment can be accessed. (Learn more about what it’s like to be a polar oceanographer here!). The team are using solar panels to power the ApRES radars to minimise the fuel use associated with servicing these sites. The aim is to increase the long-term servicing schedule to only once every 3-4 years as required by snow accumulation.
Snow accumulation at one of the sites. Photo: Craig Stewart
The team must remove the accumulated snow to access the equipment. Photo: Craig Stewart
The data is rolling in, with complete 1-year records already returned from 7 instruments. These data are used to quantify melt rates and seasonal variability in basal melting, and to identify the oceanographic processes driving ice sheet melting.
The plot below shows example data from site RD01, illustrating changes to the ice shelf between the instrument deployment in December 2021 (yellow), and a revisit in December 2023 (light blue). The graph shows the reflected signal strength (vertical axis), vs depth (horizontal axis).
The reflections within the depth range 10-50m are caused by layering within the snowpack and ice shelf. These layers are used as a vertical reference to measure changes in the height of the ice base.
The ice-ocean interface is apparent as the strongest reflector located between 100m-110m depth. This shows a shift left (upwards) by ~ 5.9m over the 2 years. This shift represents the total change in range between the antennas and the ice base over the deployment, and is caused by compaction of the snow in the upper ice shelf, strain thinning of the solid ice, and basal melting. Correcting for the displacement of the internal layers, and ignoring strain this indicates a melt rate of around 2.8m/yr. This is high for the Ross Ice Shelf, but in line with our expectations of this near ice front location which is exposed to warm surface water in summer.
Analysis of the full data record will provide similar information but with daily or better temporal resolution.
Example data from site RD01.
Site map showing all 12 sites. Green dots indicate live instruments.
The data was downloaded from each monitoring station. Photo: Craig Stewart