 |
| Titel |
Modelling the Probability of Landslides Impacting Road Networks |
| VerfasserIn |
F. E. Taylor, B. D. Malamud |
| Konferenz |
EGU General Assembly 2012
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 14 (2012) |
| Datensatznummer |
250058994
|
|
|
|
|
|
| Zusammenfassung |
| During a landslide triggering event, the threat of landslides blocking roads poses a risk
to logistics, rescue efforts and communities dependant on those road networks.
Here we present preliminary results of a stochastic model we have developed to
evaluate the probability of landslides intersecting a simple road network during a
landslide triggering event and apply simple network indices to measure the state of
the road network in the affected region. A 4000 x 4000 cell array with a 5 m x 5
m resolution was used, with a pre-defined simple road network laid onto it, and
landslides ‘randomly’ dropped onto it. Landslide areas (AL) were randomly selected
from a three-parameter inverse gamma probability density function, consisting
of a power-law decay of about –2.4 for medium and large values of AL and an
exponential rollover for small values of AL; the rollover (maximum probability)
occurs at about AL = 400 m2 This statistical distribution was chosen based on
three substantially complete triggered landslide inventories recorded in existing
literature. The number of landslide areas (NL) selected for each triggered event
iteration was chosen to have an average density of 1 landslide km-2, i.e. NL = 400
landslide areas chosen randomly for each iteration, and was based on several existing
triggered landslide event inventories. A simple road network was chosen, in a ‘T’
shape configuration, with one road 1 x 4000 cells (5 m x 20 km) in a ‘T’ formation
with another road 1 x 2000 cells (5 m x 10 km). The landslide areas were then
randomly ‘dropped’ over the road array and indices such as the location, size (ABL)
and number of road blockages (NBL) recorded. This process was performed 500
times (iterations) in a Monte-Carlo type simulation. Initial results show that for a
landslide triggering event with 400 landslides over a 400 km2 region, the number
of road blocks per iteration, NBL,ranges from 0 to 7. The average blockage area
for the 500 iterations (A¯ BL) is about 3000 m2, which closely matches the value
of A¯ L for the triggered landslide inventories. We further find that over the 500
iterations, the probability of a given number of road blocks occurring on any given
iteration, p(NBL) as a function of NBL, follows reasonably well a three-parameter
inverse gamma probability density distribution with an exponential rollover (i.e., the
most frequent value) at NBL = 1.3. In this paper we have begun to calculate the
probability of the number of landslides blocking roads during a triggering event, and
have found that this follows an inverse-gamma distribution, which is similar to that
found for the statistics of landslide areas resulting from triggers. As we progress to
model more realistic road networks, this work will aid in both long-term and disaster
management for road networks by allowing probabilistic assessment of road network
potential damage during different magnitude landslide triggering event scenarios. |
|
|
|
|
|
|