Combining space-borne SAR data and digital camera images to

Submission to the 2011 IEEE International Geoscience and Remote Sensing Symposium
Combining space-borne SAR data and digital camera images
to monitor glacier flow by remote and proximal sensing
F. Vernier1 , R. Fallourd1,2 , Y. Yan1,3 , D. Rosu1 , E. Trouvé1 , J.-M. Nicolas2 , J.-M. Friedt4 , L. Moreau5
1
: Laboratoire d’Informatique, Systèmes, Traitement de l’Information et de la Connaissance
Université de Savoie, Polytech Annecy-chambéry, BP 80439, F-74944 Annecy-le-Vieux Cedex, France
2
3
: LTCI, CNRS, Télécom ParisTech, 46 rue Barrault, 75013 Paris, France
: ISTerre, CNRS, IRD, Université de Savoie, Campus Scientifique, 73376 Le Bourget du Lac Cedex, France
4
5
: Institut FEMTO-ST, Département LPMO, 25044 Besançon, France.
: EDYTEM, CNRS, Université de Savoie, F-73376, Le Bourget du Lac, France.
Corresponding author: Emmanuel Trouvé1 , emmanuel.trouve@univ-savoie.fr
Extended abstract:
In the last decades, a spectacular retreat of most of the monitored temperate glaciers has been observed [1].
If confirmed in the coming years, this evolution will have important consequences in terms of water resources,
economical development and risk management in the surrounding areas. To monitor glacier displacements
and surface evolutions, two main sources of information are available:
• in-situ data collected for instance by using accumulation/ablation stakes, GPS stations, or proximal
sensing such as digital cameras installed near the glaciers to acquire regular images of specific areas:
serac falls, unstable moraines...
• remote sensing data acquired by different kind of sensors such as air-borne photography, space-borne
multispectral images or Synthetic Aperture Radar (SAR) data.
Both sources are complementary: proximal sensing data are usually more accurate and specifically produced for the observed glacier, but very few temperate glaciers are monitored by ground measurements
because of the cost, the access difficulty and the risks associated to ground missions in high mountain areas
where alpine glaciers are usually located. Remotely sensed data have the advantage of a more global observation potential (30 × 50 km2 with TerraSAR-X stripmap images for instance), but they are more dependent
on the sensor/satellite availability: several days often separate repeat passes and conflicts between requested
acquisitions may occur. Space-borne SAR data and especially the recently lunched High Resolution satellites such as TerraSAR-X, COSMO-SkyMed or Radarsat-2, allow global evolution monitoring and provide
regular measurements thanks to the all-weather capabilities of SAR imagery. Both amplitude and phase
information can be used to derive surface changes and velocity fields [2], or to detect and track rocks and
crevasses [3].
To monitor geophysical phenomena such as glaciers or volcanoes, it is necessary to acquire series of images
at different time (along the year for slow process, up to several images per day for fast moving areas). Digital
cameras installed near a glacier have the advantage to produce regular optical images which allow a specific
area to be monitored by very high resolution images acquired every day or even every hour if necessary, the
main constrain being the storage or transmission capacity [4]. Since each source of information has different
potential and limitations, it is interesting to develop processing techniques in order to combine them to
increase measurement accuracy and availability.
In this paper, an original approach is presented to measure serac fall displacement fields by combining
daily optical images acquired by a digital camera installed near the Argentière glacier (Chamonix, France)
and 11-day repeat pass SAR images acquired by the high resolution TerraSAR-X satellite over the MontBlanc area. Each sensor is able to observe only projections of the displacement vector in its own geometry.
The correlation technique provides for the SAR amplitude images 2 projections, respectively in the line of
sight (LOS) and Azimuth directions, whereas for the camera, the two projections depend on the image plan
orientation and on the distance to the sensor. An accurate digital elevation model (DEM) is necessary to
geocode these measurements and a short term stationarity hypothesis is necessary to combine the 11-day
SAR measurements with the 1-day measurements provided by the camera.
To apply the proposed approach, a pre-processing stage is necessary for each information source, followed
by a fusion stage to derive the final 3D (East, North, Up) displacement vectors. Accordingly, the processing
chain is divided into 8 steps:
1. For the camera, the initial JPEG images acquired in Red-Green-Blue color format, are converted into
luminance images to obtain mono-band images.
2. An initial co-registration between the images is made on the motion-free parts of the images. In
practice, the motion-free parts, i.e. mountains on the background, are used to perform it. This initial
image co-registration on motion-free areas is realized by a translation without applying sub-pixel
offsets.
3. A fast correlation technique developed in the EFIDIR Tools [5] is applied on the image pair with a
search window of 51 × 51 pixels, corresponding to a maximum offset of 10 pixels in each direction. On
motion-free areas, the sub-pixel offsets provide an accurate estimation of the remaining offset due to
the camera instability. On the glacier, the measured offset is the sum of the displacement offset and
the geometrical offset which has not been compensated for at step 2.
4. For the SAR images, an initial co-registration by a simple translation (without resampling) is applied
by matching an area of the image located at an intermediate elevation of about 2000 m ASL.
5. The fast correlation technique is applied on the whole image with a search window of 77 × 77 pixels,
corresponding to an offset of ±16 m in each direction. On motion-free areas, the sub-pixel offsets
provide an accurate estimation of the remaining offset due to the SAR geometry. On the moving
2
glaciers, the measured offset is the sum of the displacement offset and the geometrical offset which has
not been compensated for at step 4.
6. Depending on the variations of the geometrical offset along the glaciers, a post-processing step can be
necessary to deduce the offsets only due to the glacier movement. The remaining geometrical offset
can be subtracted by using either the predictions from the DEM and the orbits, or the results of the
sub-pixel correlation around the glaciers.
7. The results obtained by the two sensors are geocoded on the DEM grid (4 meter spacing) and a mask
of pixels where the 4 displacement projections are available with sufficient confidence is built.
8. Finally the Weighted Least Square inversion technique presented in [6] is applied to derive the 3D
vectors over the moving area visible by the two sensors.
Figure 1: Argentière glacier, TerraSAR-X ascending image, 2009/10/10
The experimental results presented in this paper are obtained on a series of TerraSAR-X images acquired
during summer 2009 over the Chamonix Mont-Blanc test site of the EFIDIR project [5]. They cover several
instrumented glaciers, including Argentière glacier (see Fig. 1) where a digital camera has been installed
in front of the Lognan serac fall (see Fig. 2). The correlation level obtained with the camera 4224 × 2376pixel images is rather high (above 0.9) and provides dense 2D displacement fields, expect on the areas where
seracs fell between the two image acquisitions. The correlation level obtained with the SAR High Resolution
images is lower due to the speckle effect and the backscattering changes occured during the 11-day evolution
of the ice surface and of the crevasse and serac shapes. However, the first results obtained on ascending or
descending acquisitions seem to agree with the expected glacier velocity over this area during the summer
3
season (up to 50 cm per day) and GPS measurements performed nearby [4]. Some difficulties remain due
to different sources of uncertainty: the camera orientation which has to be estimated, the low correlation of
SAR data in some part of the glacier surface and the DEM accuracy in this fast changing area. Future work
directions include performance assessment for the proposed data fusion approach and comparison with the
results obtained when ascending and descending SAR images or two digital cameras are available.
Figure 2: Argentière glacier, digital camera installed since summer 2008 in front of the Lognan serac fall
Acknowledgments
This work was supported by the French Research Agency (ANR) through Hydro-Sensor-FLOWS project and the
EFIDIR project (ANR-2007-MCDC0-04, http://www.efidir.fr). The authors wish to thank the German Space Agency
(DLR) for the TerraSAR-X SAR data (project MTH0232) over the Mont-Blanc test-site and the Société d’éléctricité
Emosson SA for their logistic support.
References
[1] C. Vincent, A. Soruco, D. Six, and E. Le Meur. Glacier thickening and decay analysis from 50 years of glaciological
observations performed on glacier d’Argentière, Mont Blanc area, France. Annals of Glaciology, 50:73–79, 2009.
[2] E. Trouvé, G. Vasile, M. Gay, L. Bombrun, P. Grussenmeyer, T. Landes, J.M. Nicolas, P. Bolon, I. Petillot, A. Julea,
L. Valet, J. Chanussot, and M. Koehl. Combining airborne photographs and spaceborne SAR data to monitor temperate
glaciers. Potentials and limits. IEEE Transactions on Geoscience and Remote Sensing, 45(4):905–923, 2007.
[3] R. Fallourd, O. Harant, E. Trouvé, J.-M. Nicolas, M. Gay, A. Walpersdorf, J.-L. Mugnier, J. Serafini, D. Rosu, L. Bombrun,
G. Vasile, N. Cotte, F. Vernier, F. Tupin, L. Moreau, and Ph Bolon. Monitoring temperate glacier displacement by
multi-temporal TerraSAR-X images and continuous GPS measurements. IEEE Journal of Selected Topics in Applied Earth
Observations and Remote Sensing, 2011 - to appear.
[4] R. Fallourd, F. Vernier, J.-M. Friedt, G. Martin, E. Trouvé, L. Moreau, and J.-M. Nicolas. Monitoring temperate glacier
with high resolution automated digital cameras - application to the argentière glacier. In PCV 2010, ISPRS Commission
III Symposium, pages CDROM, 5 pages, Paris, France, September 2010.
[5] French national research (ANR) project. Extraction and Fusion of Information for measuring ground displacements with
Radar I magery (EFIDIR) project. online. http://www.efidir.fr.
[6] Y. Yan, V. Pinel, E. Trouvé, E. Pathier, S. Galichet, G. Mauris, and A. Bisserier. Combination of sub-pixel image correlation
and differential interferometry in measurement of the 2005 Kashmir earthquake displacement field. In Fringe 2009, pages
CDROM, 8 pages, Frascati, Italy, 2009.
4