| dc.description.abstract | Landslides have an impact on inhabitants, properties and economic activities, depending on a variety
of movement types and velocities. Therefore, a management approach is required to assess the
possible risks of an area. The mountainous region of the Alps is susceptible to landslides for example,
as geophysical properties and meteorological characteristics destabilize the slope. In 1982, a
landslide was triggered in these mountains above the Barcelonnette and Saint-pons municipalities
called the La Valette. This landslide shows a rotational slide in the upper parts and earth and mud
flows in the middle to lower parts. These flows have regularly endangered the residential areas
downhill, with the largest flow having a volume of 50.000 m3.
Although these flows have not reached the downhill valley yet, through morphological developments and climate change the intensity and frequency of the flows might increase. Therefore, the goal of this research was to assess the future
societal impact of the La Valette by defining these morphological and climate changes.
To research the morphological developments, historical photogrammetry was used, as
quantifications can be derived from historical aerial images for timestamps where little to no
research was done. For the La Valette landslide, historical photogrammetry was not yet used within
the literature. In this research, the semi-global matching algorithm was able to create point clouds
for 1982, 1988, 1990, 1997 and 2000 with totals of approximately 70 million and point densities
between 2 and 4 points per square meter. These point clouds were used to create five DEMS of
differences combined with a 2021 DTM from IGN, resulting in changes like the development and
retrogression of the main scarp, the steepening of the slope and an extension towards the North
West. Furthermore, these clouds were also used to research the vertical displacement through
multiscale model to model cloud comparison, deriving accumulation and displacement locations that
were previously generalized.
To research the climate developments, Meteoblue was used to obtain historical meteorological
precipitation events and the predicted precipitations and temperatures between 2020 and 2100.
Within the historical precipitation events, multiple peaks were present surrounding the failure
events. Furthermore, climate predictions indicate that whilst precipitation sums and heavy rainfall
events remain stable or decrease, the number of heat days and the warm spell indexes will increase.
Considering that snow thaw and the variety of extreme weather events might increase, the intensity
and frequency of the failure events could also increase.
To define the future societal impact, a risk zonation was created using the previously generated
information. By creating the theoretical destructive potential of flows to the valley and the
vulnerability of the elements at risk, a risk zonation was constructed for the study area with five risk
classes. This resulted in residential areas downhill and in the valley to be situated in high to very high
risk classifications. Furthermore, 30.6 percent of the 1317 addresses was situated in the high to very
high classifications and 74.9 in the moderate to very high classifications. To conclude this research, it
can therefore be seen that the societal impact might increase in the future. For future research
however, it is recommended to create a more intensive cooperation with IGN, as this organization
could provide camera information and LIDAR datasets to improve the precision and accuracy of the
data. Furthermore, the historical images could also be digitized with a higher resolutions to improve
the point densities, which created constraints in this research. | |
