Civil Engineering Information Management
Resume:
Geophysical Research Abstracts
EGU General Assembly 2010
Author(s) 2010
Uncertainty Quanti cation on the Determination of Debris Flow Run-Out
for Quantitative Risk Analysis
Byron Quan Luna (1), Zenon Medina-Cetina (2), Jean-Philippe Mallet (3), Cees van Westen (1), and Victor Jetten
(1)
(1) United Nations University-ITC School for Disaster Geo-information Management, Enschede, The Netherlands
(********@***.**), (2) Zachry Department of Civil Engineering, Texas A&M University, College Station TX, USA
(*****@****.***), (3) CNRS - University of Strasbourg, School and Observatory of Earth Sciences, Strasbourg, France
In recent times, a number of dynamic run-out models for debris ows have been developed for risk analysis, for
the creation of zonation plans, and for the design of potential mitigation measures. Dynamic run-out models are
capable of characterizing the material distribution, the ow intensity, and the zone of potential impact. Estimating
the intensity of rapid landslides like debris ows is fundamental for quantifying the hazard on a speci c location.
Dynamic models allow the effect of released volumes as well as rheological behaviours to be modeled for
different scenarios. However, these models are still based on simple assumptions on the physical mechanisms
controlling the ow and are based on resistance parameters that cannot be measured directly during an event. As
a consequence, these models are associated with large uncertainties, which must be addressed in a proper risk
analysis.
This work introduces a systematic identi cation, characterization and propagation of the uncertainties present
in debris ow hazards analysis, with the aim of populating a joint probability density function of the input
parameters of a given debris ow model, when conditioned on eld observations. From this distribution, likely
model responses would allow for generating best estimates and con dence measures of extreme run-out distances.
To demonstrate the implementation of this method, a two-dimensional dynamic run-out model is considered that
solves the conservation equations of mass and momentum. This general methodology facilitates the consistent
combination of physical models with the available observations. Expected outputs like extension, depth and
velocity can be used as input into vulnerability and quantitative risk analysis for risk mapping and regulatory
zoning. The outlined procedure provides a useful way to produce hazard or risk maps for the typical case
where historical records are either poorly documented or even completely lacking, as well as characterizing the
con dence limits on the zoning of interest.
Contact this candidate