p
igate the
irectly a
by pixel-
(3). pp.
' Terrain
s, in VII
jeodesy,
(Eds.),
me 137,
jermany:
etti, A.,
netry. A
[magery,
.. 1100 —
F., 2009.
satellites
r Earth
idelberg,
for high-
o bias-
Istanbul,
i-Global
dons on
28-341.
er 20—
nce and
h.
ardt, M.,
Data in
speckle
Surface
Analysis
JItrafine-
s. IEEE
(10), pp.
or DSM
ogramm.
DLR for
his work
‚ed from
v Group
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume XXXIX-B7, 2012
XXII ISPRS Congress, 25 August — 01 September 2012, Melbourne, Australia
GLACIER SURFACE MONITORING BY MAXIMIZING MUTUAL INFORMATION
Esra Erten“, Cristian Rossi”, Irena Hajnsek““
, ,
“ITU, Civil Engineering Faculty, Department of Geomatic Engineering
80626 Maslak Istanbul, Turkey
‘DLR, German Aerospace Center, Remote Sensing Technology Institute,
D-82234 Wessling, Germany
SETH Zurich, Institute of Environmental Engineering, Earth Observation and Remote Sensing Group
CH-8093 Zurich, Switzerland
“DLR, German Aerospace Center, Microwaves and Radar Institute
D-82234 Wessling, Germany
Commission VII/2
KEY WORDS: SAR, polarimetry, information theory, glacier
ABSTRACT:
The contribution of Polarimetric Synthetic Aperture Radar (PolSAR) images compared with the single-channel SAR in terms of tem-
poral scene characterization has been found and described to add valuable information in the literature. However, despite a number of
recent studies focusing on single polarized glacier monitoring, the potential of polarimetry to estimate the surface velocity of glaciers
has not been explored due to the complex mechanism of polarization through glacier/snow. In this paper, a new approach to the prob-
lem of monitoring glacier surface velocity is proposed by means of temporal PoISAR images, using a basic concept from information
theory: Mutual Information (MI). The proposed polarimetric tracking method applies the MI to measure the statistical dependence
between temporal polarimetric images, which is assumed to be maximal if the images are geometrically aligned. Since the proposed
polarimetric tracking method is very powerful and general, it can be implemented into any kind of multivariate remote sensing data
such as multi-spectral optical and single-channel SAR images.
The proposed polarimetric tracking is then used to retrieve surface velocity of Aletsch glacier located in Switzerland and of Inyltshik
glacier in Kyrgyzstan with two different SAR sensors; Envisat C-band (single polarized) and DLR airborne L-band (fully polarimetric)
systems, respectively. The effect of number of channel (polarimetry) into tracking investigations demonstrated that the presence of
snow, as expected, effects the location of the phase center in different polarization, such as glacier tracking with temporal HH compared
to temporal VV channels. Shortly, a change in polarimetric signature of the scatterer can change the phase center, causing a question
of how much of what I am observing is motion then penetration. In this paper, it is shown that considering the multi-channel SAR
statistics, it is possible to optimize the separate these contributions.
1 INTRODUCTION 2 GLACIER SURFACE MONITORING
Polarimetric Synthetic Aperture Radar (PolSAR) data are being This paper expands the ideas firstly presented in (Erten et al.,
used more and more for temporal analysis such as crop inventory, 2012), which uses the MI to characterize the temporal scene in
change detection, land use management and etc. Although there terms of PolSAR images. In this work, the mutual informa-
are interesting works in glacier monitoring with single-channel tion, which is maximum if the temporal images are geometrically
single-polarized SAR images (Erten et al., 2009, Reigber et al., aligned, is used for multi-channel (PoISAR) tracking by exploit-
2008, Fallourd et al., 2011), the potential of PoISAR in glacier ing the second order statistics of the acquisitions.
monitoring has not been fully investigated yet. Indeed, a change
in the polarimetric signature affects the phase center in the pres- 2.1 Theoretical Background
ence of the snow, meaning that tracking with temporal HH chan-
nels may differ from the VV one (Rott and Davis, 1993). To 5 AOT
take into account this polarimetric information into the tracking, A temporal acquisition vector, for each pixel, k = | ka]
instead of single channel tracking, as in common in literature, is a complex vector distributed as a multi-component circular
a new polarimetric approach is proposed in this paper. The ap- Gaussian A/C(0, X) that consists of two target vectors kL ~
proach proposed in this work for tackling this problem is the N€(0, $11) and Eo oo A'€ (0, 3:22) obtained from temporal multi
maximization of the Mutual Information (MI) between tempo- channel SAR images at time t1 and ta, respectively. Here, the
ral polarimetric covariance matrices, yielding the best fit between
à t : number of elements in one of the target vectors k; at time t; is
different displacement (shift) vectors.
represented by m, and hence the temporal target vector k has the
The presentation of the paper is as follows. Section 2 addresses dimension of q = 2 x m. For example, q = 2 corresponds to in-
the theoretical background of the multi-channel MI and express terferometric SAR (InSAR) images whereas q — 6 corresponds
the proposed algorithm how to use in glacier monitoring. Sec- polarimetric interferometric (PolInSAR) images. It can be re-
tion 3 is dedicated to experimental results on for temporal glacier marked that with single channel data (m — 1), only one copolar-
monitoring with DLR airborne E-SAR sensor and Envisat im- ized channel ki = kn or ki = ku, is recorded, and the phase
ages, presenting the performance of the proposed algorithm in a carries no useful information for distributed targets. When multi-
long and short wavelength, respectively. Section 4 concludes the channel (polarimetric) data are available e.g. ki zz [Kuh Kpodtov].
work with discussions. phase differences between channels provide information about
41
