![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
Characterization of landslide kinematics with a long range terrestrial laser scan: a methodological approach |
VerfasserIn |
J. Travelletti, T. Oppikofer, J.-P. Malet, M. Jaboyedoff |
Konferenz |
EGU General Assembly 2009
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 11 (2009) |
Datensatznummer |
250026462
|
|
|
|
Zusammenfassung |
The objective of this work is to present a methodology for analyzing large displacements of
landslide with a terrestrial laser scan (TLS) and to characterize the acquisition and
computation errors of the displacement fields. Several high resolution TLS observations (0.3
to 4 pt.cm-2) were acquired in representative plots of the Super-Sauze mudslide in
2007 and 2008: the main scarp in the upper part, the medium part exhibiting the
highest displacement rates and the toe in the lower part. The TLS equipment is
an Optech ILRIS-3D. All the processing has been performed with the Polyworks
software.
Among the procedures influencing the quality of the derived displacement fields, the
alignment of the scans is the most sensitive. The distribution of statistical noise associated to
equipment errors follows a normal law and does not significantly influence the quality of the
displacement field (μ = 0.1 cm, Ï = 1.0 cm). As well, local changes in surface
soil moisture do not significantly influence the quality of the displacement field;
although the intensity of the signal is drastically decreased (~24% of the maximum
intensity), observations on nearly saturated and unsaturated plots still indicate a tolerable
error band of μ = 0.3 cm and Ï = 0.2 cm at a distance of 30 m from the laser
scan.
To quantify the displacement field from the original point clouds, several approaches
can be used: (1)Â point cloud comparisons (e.g. algorithm looking for the shortest
points along a vector), (2)Â rebuilding of object geometry (TIN model analysis), and
(3)Â difference of DEMs. In order to characterize displacements with an important horizontal
component, it is demonstrated that the object recognition method is more efficient
to characterize the kinematics on relative smooth topography than point clouds
algorithms.
To characterize displacement with a more important vertical component, such as the
collapse of material from the main scarp of the mudslide, a “jackknife” procedure was used to
identify the best interpolation techniques for producing the DEM. A differential DEM
analysis allowed to define the volume of the collapse (~23.000Â m3) as well as a progressive
subsidence of the area downslope.
The quality of the alignment is the most sensitive parameter influencing the
accuracy of the laser scan observations. A good coverage among the scans and
the inclusion of stable parts are necessary to maximize the alignment procedure,
but the number of scans to acquire has also to be minimized in a survey planning. |
|
|
|
|
|