Svalbard glacier products
Revision as of 18:51, 28 March 2014 by MaxK
Two products for the glaciers of Svalbard are available. The Glacier Area Outline (GAO) product displays changes in glacier extent over the years. Glacier Surface Type (GST) classifies glaciers in glacier ice, superimposed ice and firn, thus allowing to see retreat or advance of the firn are.
The Glacier Area Outlines (GAO) were produced through a collaboration between the Department of Geoscience, University of Oslo, and the Norwegian Polar Institute, supported by the European Space Agency (ESA) through the projects Glaciers_CCI (4000101778/10/I-AM) and Cryoclim, which is also supported by the Norwegian Space Centre.
Glacier Area Outline (GAO)
The archipelago of Svalbard presently contains approximately 33,200 km2 of glaciers, with a large number of small valley glaciers as well as large areas of contiguous ice fields and ice caps. While a first glacier inventory was compiled in 1993, there has not been a readily available digital version. Here we present a new digital glacier database, which will be available through the GLIMS project. Glacier outlines have been created for the years 1936, 1966-71, 1990, and 2001-2010. For most glaciers, outlines are available from more than one of these years. A complete coverage of Svalbard is available for the 2001-2010 dataset. Glacier outlines were created using cartographic data from the original Norwegian Polar Institute topographic map series of Svalbard as basis by delineating individual glaciers and ice streams, assigning unique identification codes relating to the hydrological watersheds, digitizing center-lines, and providing a number of attributes for each glacier mask. The 2001-2010 glacier outlines are derived from orthorectified satellite images acquired from the SPOT-5 and ASTER satellite sensors. In areas where coverage for all time periods is available, the overwhelming majority of glaciers are observed to be in sustained retreat over the period from 1936-2010.
The following figure shows the GAO products for Kongsfjorden, Svalbard, on top of a SPOT image. Glacier retreat is clearly visible in the image
The GAO product is available for three different time periods, covering all of Svalbard in the period 2001-10:
The following screenshot displays a screenshot of http://svalbardkartet.npolar.no showing a selected glacier in Kongsfjorden with its attributes. Again, glacier retreat since 1936 is clearly visible:
More documentation on this data is available at data.npolar.no/dataset/89f430f8-862f-11e2-8036-005056ad0004
Glacier Surface Type (GST)
On many glaciers in Svalbard, three surface types are visible on SAR images, the dark glacier ice at the glacier's lower end, the brighter superimposed ice in the middle, and the white firn at the higher elevations. Surface classification of these types is valuable especially since the retreat or advance of the firn area provides information on the status of the glacier. While the snowline reacts immediately to annual changes, the firn area smoothes out these short-term changes and shows, similar to the glacier front, longer-term changes of the glaciers.
GST uses a Otsu three-category algorithm to separate the image into these three surface types for selected Svalbard glaciers. The method works very well on glacier with distinct surface types, the main weakness is crevasses and rough areas being classifed as superimposed ice. A quality number indicates if an individual classification is ideal (1), good (2) or of medium quality (3). The quality number mainly indicates how much crevasses are classified as superimposed ice. The firn area should be displayed correctly for all selected glaciers.
The clear division of surface types is present on many Svalbard glaciers, as seen in this ideal example from Kongsvegen glacier
A close-up of the GST product for Kongsvegen shows, how the retreat of the firn line can be seen over the years
The available glaciers classified for surface type can be seen at willem.npolar.no/cryoclim
More documentation on this data is available at data.npolar.no/dataset/d756f766-de33-11e2-8993-005056ad0004
Glacier Balance Area (GBA)
The GBA product was intended to be validated within the CryoClim project. Inital Results suggested that classifying the glacier surface on a C-band SAR image with a k-means algorithm gives an area, the yellow area in the image below, whose extent varies with mass balance. The following images show a glacier surface classified with this method:
The following Figure compares the annual variation of the extent of the yellow area with in-situ mass balance measurements. The Correlation between both is very good:
However new data showed, that this correlation does not exist in all years and vanishes in 2005-2009. The reason for this is not understood yet:
GBA was therefore not included in the CryoClim products.
König, M., J. Wadham, J-G. Winther, J. Kohler, and A-M. Nuttall. 2002. Detection of superimposed ice on the glaciers Kongsvegen and midre Lovénbreen, Svalbard, using SAR satellite imagery. Annals of Glaciology, 34, 335-342.www.ingentaconnect.com/content/igsoc/agl/2002/00000034/00000001/art00050
König, M., J-G. Winther, J. Kohler, and F. König. 2004. Two methods for firn area and mass balance monitoring of Svalbard glaciers with synthetic aperture radar (SAR) satellite images. Journal of Glaciology, 50(168).www.ingentaconnect.com/content/igsoc/jog/2004/00000050/00000168/art00012
König, M. C. Nuth, J. Kohler, G. Moholdt and R. Pettersen, 2013, A New Digital Glacier Database for Svalbard, AGU poster available at public.data.npolar.no/cryoclim/CryoClimGAO.pdf
König, M. C. Nuth, J. Kohler, G. Moholdt and R. Pettersen, 2013, A Digital Glacier Database for Svalbard, in press, GLIMS publication. www.springer.com/earth+sciences+and+geography/remote+sensing/book/978-3-540-79817-0
Nuth, C., Kohler, J., König, M., von Deschwanden, A., Hagen, J. O., Kääb, A., Moholdt, G., and Pettersson, R. , 2013. Decadal changes from a multi-temporal glacier inventory of Svalbard, The Cryosphere Discuss., 7, 2489-2532, doi:10.5194/tcd-7-2489-2013 available at www.the-cryosphere-discuss.net/7/2489/2013/tcd-7-2489-2013.html