Controlled Interferometric Modelling of Glacier Changes in South Svalbard

Aleksey Sharov(1)

(1) Joanneum Research, Wastiangasse 6, A-8010 Graz, Austria


The present paper describes main outcomes from the INTEGRAL (EC FP6) and SIGMA (ESA AO No.2611) research projects devoted to enhanced interferometric modelling of glacier dynamics in Svalbard. The underlying concept of the research is to compensate the lack of reliable basic control and accurate reference models, which are needed for precise interferometric modelling of active glaciers, with spaceborne altimetry data. The main test sites comprise Austfonna in northeast Svalbard and several large tidewater glaciers in the national park Sör-Spitsbergen, south Svalbard. Our paper deals with joint processing of the ERS-1/2 SAR interferograms and ICESat-GLAS altimetric transects obtained over test glaciers in 1995-1996 and 2003. Norwegian topographic maps showing glacier fronts from 1990 and glacier elevations from 1936, two hydrographic charts issued in 2002, several ASTER-VNIR spaceborne images taken in 2004, and 3-D coordinates of control points from D-GPS surveys in 2004 laid the groundwork for detecting, measuring and interpreting glacier changes in linear, areal, volumetric and fluxometric terms.

Straightforward and rigorous procedure for the absolute orientation of SAR interferograms with altimetry data including operations for co-registering altimetry data with standard interferograms, determining and removing phase offsets, height ambiguity control and spatial baseline refinement, geocoding and mosaicking of INSAR models was designed and tested. Further interpretation of geocoded INSAR products allowed the location of ice-flow divides to be determined and topographic heights of main ice divides, ice coasts and specific locations on the glacier surface to be precisely measured. The arbitrary height and velocity of each specific point on the glacier surface characterised by a high coherence value was derived from single geocoded SAR interferograms without complex processing artifices such as phase unwrapping and differential interferometry. In geocoded INSAR products, all foreshortening effects at precipitous glacier fronts can be accounted for, thus allowing precise glacier change detection in multitemporal data sets.

The largest glacier retreat was detected in the southern part of Svalbard, where the fronts of Hambergbreen, Hornbreen and Storbreen tidewater glaciers terminating in Hornsund receded 3.3 (6.5), 2.4 (8.8) and 2.6 (9.5) km respectively in the course of past 14 (68) years. Ostrogradskifjella (622 m a.s.l.) and Arrheniusfjellet (882 m a.s.l.) mountains became capes. The width of icy isthmus connecting Sörkapp Land with the Spitsbergen main island narrowed from 25 km in 1936 to 14.4 km in 1990 and 8.8 km in 2004. The elevation of this narrow “ice bridge” with a relatively flat, smooth surface fell by 100-130 m. Presently it does not exceed 150 meters a.s.l. Careful analysis of multitemporal interferograms revealed that quite large seaward parts of the ice bridge undergo both horizontal motion due to the glacier flow and local vertical movements caused by tidal effects that allowed us to assume the presence of a strait and a separate island with an area of ca. 1270 km² in that region. Apart from essential retreat of Bråsvellbreen (350-500 m), relatively insignificant marginal changes were detected on Austfonna. Significant positive height changes up to 30 – 60 m were determined in the accumulation area of this ice cap. Most test glaciers have steepened their sides. The results obtained were compared with those published by other investigators. The study revealed that the high glacier strain rate values manifested in our value-added INSAR products can reliably indicate future changes at maritime glacier margins. Glacier changes and related processes were represented in the form of controlled interferometric composites, satellite image maps and animated graphics. Our tests proved the high elevation accuracy of controlled interferometric composites, and typical height errors were estimated as being nearly 5 times smaller than those in standard INSAR products. Further ground-control surveys, reference observations and map content reviews will be carried out during two field campaigns in 2006.


Workshop presentation

Full paper


  Higher level                 Last modified: 07.10.03