In early October the University of Gothenburg, Sweden, played host to a one-day SOOS workshop to discuss the present and future strategy for NECKLACE. A group of 29 researchers from nine countries gathered at the beautifully-located Kristineberg Research Station, a couple of hours north of the city. The meeting was held immediately before the 2016 FRISP meeting
Figure 1: NECKLACE logo
NECKLACE is a SOOS-endorsed programme to assemble a collection of time series of the rates of melting at the base of Antarctic ice shelves, with the aim of providing a product for the ice-ocean modelling community. Specifically, the ambition is to provide a circumpolar set of time series from specific geographical locations (not pan ice-shelf averages) that would be used for validation and improvement of those ocean models that include sub-ice shelf cavities in their domains.
The workshop covered three main areas: technical aspects of the deployment of instruments, including selection of the most technically-appropriate locations; a discussion on the needs of the numerical modelling community; and a review of options for data dissemination. it also included a brief round-up of future modelling and deployment plans.
Keith Nicholls (British Antarctic Survey) led the discussion on the instrument itself – ApRES, a phase-sensitive radar designed for long-term deployments to monitor vertical strain in the ice column and basal melt rates. This was followed by a detailed assessment of potential problems in interpreting data from different deployment sites. One key to a “good” site is in the upstream history of the ice column. Relative vertical motion of the internal reflecting horizons in the ice column is used to determine the vertical straining in the ice, which can then be subtracted from the measurement of ice thinning to give the basal melt rate. A set of internal horizons that has been badly deformed by upstream processes is more difficult to process, as non-nadir returns contaminate the profile.
Another assumption in the data processing is that the local shape of the ice base remains constant. An example from the heavily incised base of Pine Island Glacier was used to show how the form of the basal reflection at one site varies during the year, compared with the reflection from a site only a kilometre away, which remained constant. Such variability ultimately affects the time-resolution that can be obtained when processing the data for melt rates.
Figure 2: Poster session at the FRISP meeting
The modelling session was led by the Hollands: Paul Holland of the British Antarctic Survey, and David Holland of New York University. The strengths of ApRES-derived melt rates were compared and contrasted with those of satellite products, with the conclusion that a valuable contribution will be the calibration and validation of the satellite-based observations. It was noted, though, that additional datasets would be needed to improve the reliability of satellite retrievals of basal melting, primarily to account for the effects of surface processes.
It was clear that models needed to be able to replicate the melt rate patterns determined by satellites, but that a more stringent test of the models was their ability to generate the observed temporal pattern of melting. There was some additional discussion about whether the vertical strain rates recovered from the ApRES datasets would be of value to ice shelf/stream/sheet modellers.
One suggestion that did arise from the group was the idea for a comprehensive, long-term study of an example ice shelf. The possible addition of other techniques available to the community, including sensors deployed in the ocean via boreholes, was identified as potentially adding great value to such a dataset. This would allow a more detailed understanding of the processes, presently parameterised, that mediate the transfer of heat to the ice base.
Pip Bricher, from University of Tasmania, and the SOOS Data Officer, presented potential options for making data from NECKLACE available to the community. Although SOOS is unable to offer any data archival or curation facilities, it is able to assist in finding appropriate homes for data, and to disseminate the data to the community. A key requirement is that the data should be in a common format. The present proposal, broadly accepted by those present at the workshop, was to make monthly-averaged melt rates available.
Pip suggested that NECKLACE might make use of a website, such as SOKI, which was established to be a wiki environment for Southern Ocean researchers to discuss and share field, laboratory, and statistical techniques. Importantly, it has the tools required for version control, data and document sharing, the capability of issuing DOIs etc. It was widely agreed that such a site would be of great benefit as a central resource, especially for documentation associated with deployment and use of the equipment, and as a forum for discussion of problems and solutions.
The workshop closed with statements from representatives outlining plans for deployment of instruments in the 2016-17 field season. The meeting organisers and SOOS are very grateful to the University of Gothenburg for facilitating and hosting the workshop.
Figure 3: Group photo from the FRISP meeting