Combining satellite altimetry (ERS-2 and ENVISAT) with SAR interferometry and SPOT photogrammetry for studies of Austfonna ice cap (Svalbard)
Alexei Kouraev(1) , Benoit Legresy(1) , and Frederique Remy(1)
14 av Edouard Belin,
Alexei Kouraev1,2, Benoit Legresy1, Frederique Remy1
(1) Laboratoire d’Etudes en Géophysique et Océanographie Spatiales (LEGOS), Toulouse, France,
(2) State Oceanography Institute, St. Petersburg branch, St. Petersburg, Russia,
Glaciers and ice caps are extremely sensitive for detecting global climate variability. We have studied largest ice cap in the Eurasian Arctic - the Austfonna on the Svalbard using combination of various satellite techniques - satellite radar altimetry (ERS-2 and ENVISAT), SAR interferometry (ERS-1 and -2) and SPOT photogrammetry. Though this ice cap has experienced a modest advance during the last 30 years, there are outlet glaciers that have experienced large changes, such as Etonbreen glacier.
We first analyse ERS-2 and ENVISAT altimetry data over the Austfonna since 1995. We present changes of absolute height and height anomalies, as well as backscatter values (depending on the surface roughness, melting and refreezing processes) and the leading edge width (affected by the presence of snow dunes, small-scale ondulations, effects of radar penetration in the snow).
We also present results of the combined use of various Digital Elevation Models (DEM) of the Austfonna with ENVISAT and ERS-2 altimetry data. We use historical DEM from the Norwegian Polar Institute, as well as several DEMs from ERS-1 and -2 SAR interferometry (INSAR) and SPOT photogrammetry. DEMs from INSAR and SPOT stereo pairs are able to resolve topography with high resolution but they lack absolute reference and are prone to long-wavelength errors. Radar altimetry produce absolute measure of height with low spatial resolution (several km) with very high precision reference but they are perturbed by small-scale topographic changes. By combining DEMs with altimetric measures we could improve DEMs precision. The DEMs have been used to first simulate the radar return waveform. By comparing the simulated and measured waveform we then estimate the errors associated with the DEM inclination plane (related to the long-wavelength DEM errors) and correct them using the least square method. Then we again simulate the return waveform and by this iterative approach we assess the application of this particular technique to DEM improvement and to altimeter measurements interpretation.
Using the results obtained, we analyse influence of glacier slope, atmospheric refraction, tides and the state of glacier surface (fresh or wet snow, bare ice, roughness) on the accuracy of altimetry data and estimate necessary corrections. We discuss processing of crossover techniques for processing multi-pass altimetry data obtained over glacier surface and the assessment of expected accuracies.
We have also performed correclation of multitemporal SPOT imagery and INSAR data. While interferometry provide only one component of the speed vector but with high accuracy, while correlation of SPOT imagery (with little time lag and with incidence angles less than 15 degrees) provide estimates of the two component, though with lower precision. These two techniqes are highly complementary and both are necessary to establish the best possible assessment of ice movements.