ul 2004
and the
e LSM
| plane
ss (D»)
re and
t in the
ne with
t space.
| image
of the
epipolar
or each
into (6)
s. The
pseudo-
T (0)
Pi
nstraints
a multi-
y vector
n/length
gravity,
lirection
1e rough
the flow
rate, the
and Y-
1 respect
can be
n:
t will be
er. If the
of errors
. weight,
ide false
natching
ias been
udy of à
ADVM.
International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol XXXV, Part B7. Istanbul 2004
3. CASE STUDY
The Hinteres Langtalkar rock glacier (12?45'56" E, 46958'53"
N) is located in the Schober group of the Hohe Tauern range
within the boundaries of the Austrian Hohe Tauern National
Park (see Fig. 3). The rock glacier is situated in the Hinteres
Langtal cirque (see Fig. 4). The foot of the front slope is located
at 2460 m a.s.l. The northern lobe of the rock glacier is 850 m
long and 200 m wide. This highly active part of the rock glacier
is in contrast to the rather inactive southern, smaller part of the
rock glacier. The latter is characterized by a sequence of
adjacent ridges and furrows, which typically relates to
compressive flow. At the rooting zone of both parts of the rock
glacier marked depressions are present. Further information
about the geographical setting and results of geological and
hydrological investigations can be found in Lieb (1987) and
Krainer & Mostler (2001), respectively.
F49° 10° 12° 14°
0 100 km
Vienna
Austria
Hinteres
,— Langtalkar
rock glacier
Munich
e Innsbruck Graz
Figure 4. Orthophoto (9.9.1997) of the study area showing the
Hinteres Langtalkar rock glacier. The box delineates the area
shown in Figures 5-7.
The Hinteres Langtalkar rock glacier is very interesting from a
geomorphological point of view, since the lower end of the rock
glacier tongue has moved into steeper terrain. A sliding process
has started as early as 1992, as the strength of the ice/rock
mixture could no longer resist gravitational forces. This
phenomenon was identified for the first time in aerial
photographs of 1997 (cp. Fig. 4). A similar geomorphological
situation is known from the Ausseres Hochebenkar rock glacier,
Austria (see Kaufmann & Ladstädter, 2003). A monitoring
program has been initiated in order to study the
morphodynamics of the Hinteres Langtalkar rock glacier in
more detail. The program includes (1) reconstruction of the past
kinematic state of the rock glacier using aerial photographs, (2)
determination of the present movement by means of geodetic
Survey, and (3) investigation of alternative monitoring
897
techniques, such as terrestrial laser scanning and satellite-based
radar interferometry. Some results of the photogrammetric work
are presented in the following Section. Geodetic work is briefly
outlined in Kaufmann & Ladstädter (2003). Concerning the
alternative monitoring techniques the reader is referred to Bauer
et al. (2003) and Kenyi & Kaufmann (2003).
3.1 Results of digital photogrammetric work
Aerial photographs at various scales of 11 different overflights
between 1954 and 1999 were made available for detailed
photogrammetric analysis in order to study the spatio-temporal
evolution and Kinematic behavior of the rock glacier. The main
task was to derive 3D displacement/flow vectors using ADVM
software. In a first step all photographs, except those of 1954
which show a thin snow cover, were photogrammetrically
orientated using a Kern DSR-1 analytical plotter. A digital
photogrammetric workstation of Z/l Imaging was at our
disposal later. Subsequently, all photographs were digitized
with a resolution of 10 um using an UltraScan 5000 of Vexcel
Imaging Austria. Elements of exterior orientation were taken
from the analytical plotter. The stereopair of 1998 was selected
for detailed photogrammetric mapping of the rock glacier and
its surroundings. This work was done manually at the analytical
plotter in order to achieve high accuracy and geomorphic
acuity. A similar procedure was performed for the stereopair of
1974. However, the mapping was limited to the area covered by
the rock glacier. High resolution digital terrain models with a
grid spacing of 2 m were derived from both data sets. Pseudo-
orthophotos with a pixel size of 0.25 m were computed from all
photographs selected. Photographs of 1992 and older were
(pseudo-)rectified using the DTM of 1974, the younger ones
using the DTM of 1998. All pseudo-orthophotos were
processed with ADVM software. Finally, 3D displacement
vectors were obtained. In this paper we present a graphical
representation of the spatial distribution of the horizontal creep
velocity for the time period 1969-1974 (cp. Figs. 5-7).
We can summarize the main findings of the case study as
follows: The creeping process of the rock glacier has remained
steady over the years with flow/creep velocities increasing from
the root towards the lower end (frontal slope). A mean annual
horizontal flow velocity for the time period 1969-1991 was
measured at 1.35 m/year at the upper rim of the frontal slope. A
landslide occurred at the steep frontal slope of the rock glacier
between 1992 and 1997. The photogrammetric measurements of
three observation periods after 1997 reveal that the flow
velocity has significantly increased at the lower part of the rock
glacier (max. horizontal movement of up to 2.8 m/year)
compared to previous years. Distinct crevasses with high
longitudinal strain rates have developed as a result of the
extending flow/creep process. Photogrammetric results were
checked by means of precise annual geodetic measurements
(1999-2003).
4. CONCLUSIONS
With the present case study we could demonstrate that the
proposed digital photogrammetric method of deformation
measurement from multi-temporal aerial photographs is feasible
and that the movement of the rock glacier could be measured
with high accuracy. Further developments of ADVM software
will address robust matching in difficult areas, i.e., with low or
missing texture, snow patches and cast shadows. Additional
applied studies must be carried out in order to better tune the
various parameters needed in the semi-automatic workflow.
