Virkisjökull, in south-east Iceland, is retreating rapidly, like most glaciers in Iceland. Since 1996, the glacier margin has retreated nearly 500 m and it appears that this rate has accelerated over the last five years. We operate an observatory site at Virkisjökull, studying the evolution of the glacier (Icelandic: jökull) and the surrounding landscape and their responses to regional climate (Figure 1).
Sensors at the site are constantly collecting climate and seismic data. Repeated high-resolution surveys study how the glacier, land surface and the deposits beneath change over time. Cutting-edge technologies, not used in such a combination anywhere else in the world, are used to give unique insights into processes of landscape formation and responses of glacial systems to climate forcing.
In the 2009, 2010 and 2011 surveys, the scanner used was the Riegl LPM-i800HA. In the 2012 survey it was the Riegl VZ-1000. Both scanners have a high-resolution digital camera associated with them (Figure 2).
Multiple instrument set-ups at a series of locations around the glacier were required to obtain a complete surface model, as not all the subject area was visible from any one location. It is essential that these different datasets are accurately referenced to a common coordinate system. Multiple recording locations also minimised the number of ‘shadow areas’ (parts of the subject obscured from the scanners’ view) within the final compiled surface model, with each component scan containing at least three common points to assist with orientation and significant overlap, typically around ten per cent.
The terrestrial LiDAR data produced by the oriented laser scan and GNSS survey were processed to develop a virtual outcrop model (VOM) of the glacier and glacial margin (Figure 3). The raw data produced by the RiProfile program, used in 2009 to 2011, consisted of point clouds comprising 70 million x-y-z points for the September 2009 survey, 40 million x-y-z points for the September 2010 survey and 70 million x-y-z points for the September 2011 survey. The raw data produced by the RiScanPro program, used in 2012, consisted of a point-cloud containing 300 million x-y-z points.
These data were oriented using the relative differential GPS positions of both the scanner and the back-sights and output as an ASCII file, made up of x-y-z intensity and RGB colour values. For the 2009 to 2011 surveys, the data were imported into the ImAlign package within Polyworks to align individual scans and to check for errors in orientation. In 2012 this was carried out in RiScanPro. Surface 3D digital elevation and change models between epochs (Figure 4) were created using Surfer and IMSurvey, and 3D coloured point-cloud models and video fly-throughs were created in QTModeler and Pointools.
In April 2012 six single-frequency GNSS units were deployed on Virkisjökull and Falljökull (Figure 5) and a dual-frequency GNSS unit was set up as a permanently recording base-station near the car park. In September 2012, the first 140 days of data from these on-ice GNSS units and the base-station unit were downloaded. These data were processed in Leica Geo-Office and positional change and velocities were calculated for the glacial flow in Excel (Figure 6).
The main issues with the LiDAR surveys carried out at the Virkisjökull Ice Observatory were with the locations of the scan sites themselves and, in 2009, problems with GNSS availability. The logistics of transporting 60 kg worth of surveying kit across the sandur and glacial foreland proved quite difficult; transporting the same amount of kit onto the glacier itself (wearing crampons) was hazardous.
The System 1200 GNSS units used in 2009 struggled with satellite availability at such a northern latitude, but from 2010 onwards the Viva GNSS unit, which utilises both GPS and GLONAS, coped much better. The 2009 scans were able to be correctly orientated by re-aligning them using ImAlign.
If you want to discover more then please contact Lee Jones.