Organisation

PhD Defence Mr Byron Quan Luna

Department of Earth Systems Analysis

Dynamic numerical run-out modelling for quantitative landslide risk assessment

Landslides and debris flows are geomorphologic events that may pose danger to different components of mountainous societies. This danger is not only the result of the process as such, but of the interaction with human systems and their associated vulnerabilities. Understanding, forecasting and controlling the hazard associated to this type of slope movements is still an empirical task which requires a mix of qualitative and quantitative analyses. The development of numerical dynamic run-out models has made a dramatic change in the study of hazardous processes, as these allow the simulation of possible future scenarios, including ones that have no historical evidences.  Dynamic computer models have the potential to simulate geomorphologic processes with an acceptable degree of accuracy. Once this is achieved, a range of potential hazard scenarios can be analyzed and the results can be used to inform local authorities and the population in order to respond to these hazards and plan to reduce the associated risks. For this reason, it is important to evaluate the reliability and consistency of these dynamic models that integrate the physical descriptions of the landslide process in a numerical scheme, embedded in a Geographical Information System (GIS).

A variety of models exists for simulating and for identifying the hazard that the different mass-flow phenomena present. Dynamic run-out models are able to forecast the propagation of material after the initial failure and to delineate the zone where the elements-at-risk will suffer an impact with a certain level of intensity. The results of these models are an appropriate input for vulnerability and risk assessments. An important feature of using run-out models is the possibility to perform forward analyses and forecast changes in hazards. However, still most of the work using these models is based on the calibration of parameters doing a back calculation of past events. Given the number of unknown parameters and the fact that most of the rheological parameters cannot be measured in the laboratory or field, it is very difficult to parameterize the run-out models. For this reason the application of run-out models is mostly used for back analysis of past events and very few studies attempts to achieve a forward modelling with the available run-out models. A reason for this is the substantial degree of uncertainty that still characterizes the definition of the run-out model parameters.

The main objective of this research was to apply, improve and optimize the use of dynamic run-out models in quantitative risk assessment, focusing on the parameterization of the models, and the analysis of uncertainty. Since a variety of models exists for simulating mass-flows and for identifying the intensity of the hazardous phenomena, it is important to assess these models, perform a parameterization and reduce their uncertainties. This will enable to improve the understanding to assess the hazard and will provide the link with vulnerability curves that will lead eventually to generate risk curves and quantify the risk.

This research describes the state of the art in dynamic run-out modelling focusing on continuum depth-average models. Three different dynamic run-out models (MassMov2D, DAN3D and RAMMS) were selected for a sensitivity analysis of their resistance parameters using the Voellmy rheology. Three test sites were used: Barcelonnette in France, Valtellina di Tirano in Italy, and a site in Kerala, India.

A special consideration was given to the entrainment mechanism. The increase of volume once a failed mass is in movement due to entrainment enhances the mobility of the flow and can significantly influence the size of the potential impact area.  In view of this, a 1-D run-out model is presented with an entrainment concept based on limit equilibrium considerations and the generation of excess pore water pressure through undrained loading of the bed material.

An extensive database was made which includes the rheological parameters (Voellmy and Bingham rheologies), release volumes, the type of movement, the environmental setting and other physical characteristics of previously back-calibrated events that have been described by other authors. Using the database, the variability for the rheological parameters was represented as probability density functions. The PDFs were used in a probabilistic framework based on a Monte Carlo simulation to analyze the effect of the uncertainty of input parameters. Combined probability density functions of the Voellmy and Bingham rheology were sampled and a large number (5000) of run-out scenarios were generated. The result was a Gamma probability distribution of possible intensities in selected points of the deposition area. The result obtained from the application of this methodology was the probability of a selected location being affected by a landslide in terms of intensity factors (height or velocity). The generated probability density functions were also applied to a newly developed medium scale model called “AschFlow”, which a 2-D one-phase continuum model that simulates the spreading, entrainment and deposition process of landslides or debris flows at a medium scale in the French test site.

Complexity arises with the interaction of the modelling intensity outputs with the affected elements at risk. For this reason, three physical vulnerability curves that relate the intensity of debris flows and the economic losses were derived from a well documented debris flow event. The event was back analyzed with a dynamic-run out model and the outputs were related to the damage data of elements at risk in order to generate the vulnerability functions. A quantitative risk assessment was carried out using run-out modelling for the Italian study site. Based on the historical events and susceptibility maps, three potential debris flows initiation zones were delimited. These selected areas were modelled with the dynamic run-out model FLO-2D to assess the run-out intensity.  The methodology used in this analysis consisted of several components, such as a detailed analysis of rainfall return periods (10, 50, 100 years return period), the modelling of rainfall-runoff, the analysis of soil samples in the laboratory, the analysis of terrain characteristics, the modelling of the run-out of the debris flows, the application of debris flow height and impact pressure vulnerability curves and the generation of risk curves based on the economic losses. 

This research has contributed to a better understanding of the use of run-out modelling of debris flows, and provides a number of new avenues for the incorporation of uncertainty in this type of analysis, in order to be better make an estimation of potential losses. The results can be applied in cost benefit analysis for the design of risk reduction measures.

The main part of this research was carried out as an Early Stage Researcher inside the European Commission Marie Curie Actions Research Training Network: “Mountain Risks: from prediction to management and governance” within the 6th Framework Programme (http://mountain-risks.eu/). The last part and completion of this research was executed inside the “SafeLand” project within the 7th Framework Programme for research and technological development of the European Commission. (http://www.safeland-fp7.eu/).

Biography

Byron Quan Luna was born on 30th of March of 1975. He obtained his Bachelor degree in Environmental Engineering at the Rafael Landivar University in Guatemala. During the period of 1999 to 2005, he worked as a technical advisor in a private sector company in Guatemala (SEPINSA - Incorporated Professional Services) where his main activities were to plan, develop and evaluate agricultural, forestry and environmental projects. In 2007, Byron Quan Luna received a Master of Science in Environmental Geology and Geohazards from the Department of Geosciences, University of Oslo, Norway with the thesis: “Assessment and modelling of two lahars caused by Hurricane Stan at Atitlan, Guatemala, October 2005”. From 2007 to 2008, he did an internship at the International Centre of Geohazards (ICG) in the Norwegian Geotechnical Institute (NGI) on numerical analysis and computer modelling of mass movements (landslides, debris flows and snow avalanches).

In 2008, Byron Quan Luna started his PhD research in The Netherlands at ITC-University of Twente (UT) as a part of the Mountain Risk project as an Early Stage Researcher inside the Marie Curie Research Training Network. The project main focus is research and training in all aspects of mountains hazards and risks assessment and management. The network intended to develop an advanced understanding of how mountain hydro-geomorphological processes behave and to apply this understanding to living with the hazards in the long-term. In 2012, he was involved in the last part of the SafeLand project “Living with landslide risk in Europe: Assessment, effects of global change, and risk management strategies”

Since his start at ITC-UT in 2008, he has been involved as participant or lecturer in different topical workshops and intensive courses that has taken place in different countries such as Italy, Spain, Austria, Nicaragua, France, The Netherlands, Switzerland and Germany. He was a participant in the LARAM 2008 summer school (International School on "Landslide Risk Assessment and Mitigation") in Ravello, Italy. He has been advisor in several master thesis carried out by ITC-UT students.

Quan Luna, B., Jetten, V.G. (Promotor) , van Westen, C.J. (assistant promotor) and van Asch, Th.W.J. (assistant promotor)  (2012) Dynamic numerical run - out modelling for quantitative landslide risk assessment. PhD thesis University of Twente, Summaries in English and Dutch. ITC Dissertation 206, ISBN: 978-90-6164-330-2

   Full text

previous page
more events