Methodology:

In 2004, as part of a NERC Centre for Polar Observation and Modelling (CPOM) initiative led by Professor Martin Siegert, then based in the Bristol Glaciology Centre, photographic copies of the raw 35 mm film records were scanned, and each scan was reformatted to form a single electronic copy of a RES transect (i.e. a 'jpeg' or 'tiff' file). Each transect was loaded into an image analysis package (i.e. ERDAS ImagineTM) in which internal layers were traced. The traced layers, in addition to the ice surface and bed, were then digitised at each navigational control point. The resulting dataset was then standardised with respect to the ice surface, such that the data relate to depths below the surface of the ice sheet. A CPOM follow-up to this work, conducted between 2005 and 2006 by Dr. Robert Bingham, then also based at the Bristol Glaciology Centre, was concerned with emplacing these data into a 3-D environment. This website provides a repository for both the 2-D traced internal layers and their visualisations within a 3-D environment.

Data collection:

RES data were collected using a 60 MHz system with a 250 ns pulse length. Navigation proceeded through the use of an inertial aircraft system, and although not as accurate as that which can be achieved using modern GPS, navigational accuracy was typically attained to within 1 km, an acceptable resolution for most ice-sheet modelling applications. Because the survey programme preceded the 'digital age', data recording was analogue only. Data were recorded onto 35 mm photographic film as 'Z-scope' 'radargrams', in which single RES pulses are stacked next to one another in a time-dependent manner, and navigational tick-marks added to these records enabled tying-in of the records with geographical waypoints. These data represent pseudo-cross-sections of the ice sheet from which internal layers can be viewed, traced and extracted. Nevertheless, prior to 2004, the analogue records pertaining to internal layers remained largely unused in glaciological applications: this was primarily because the data were not in digital form thus it was difficult and time-consuming to extract information relevant to modelling.