Byron Quan Luna

On 4 January, 2009, more than 5 million cubic metres of limestone and calcareous breccias detached from the "Los Chorros" hill and travelled along a tributary ravine of the Chixoy river in the municipality of San Cristóbal Verapaz, department of Alta Verapaz, Guatemala. At the time of this landslide, several persons were crossing roads and foot trails downstream of the release area. As of 14 January, authorities had reported 38 casualties, 50 missing and 5 injured persons. Along the landslide path, a 1.2 km segment of the 7W National Highway was destroyed, cutting the sole access route between San Cristóbal Verapaz and the western department of Quiché where numerous inhabitants of Alta Verapaz commute to work, especially during the coffee harvest season from October to March (with its peak on January). In response to this disaster, the Guatemalan government established four priorities: search and rescue activities, relief aid to victims and their families, evacuation of villages at risk and selection and construction of a temporary access route and a permanent road. In an attempt to provide additional elements to decision-makers of the Guatemalan authorities, this report is aimed to characterise the context of this landslide from a geosciences perspective. Preliminary assessments of the 4 January event and of other potentially unstable zones identified in the surrounding areas are also performed. The first accounts of ground instabilities in this area date back to 1590 when a 4.0 MS earthquake was associated with the collapse of a karst cave. In 1881, a Guatemalan newspaper reported that the San Cristóbal (Chichoj) lagoon was created after a ground subsidence was triggered by an earthquake. In 1983, after less than one year of operations, a ~50 m segment of the 26-km long pressure tunnel in the Chixoy hydro electrical project was damaged due to an anhydrite karst produced during tunnel operations. In response to this event, repair and strengthening works were carried out and power plant operation was re-started in 1985. At the end of November 2008, small landslides occurred in the surroundings of the area of the 4 January event and on mid December, a few number of larger slides occurred killing 2 persons and blocking the 7W National Highway. Rumbling noise was often reported by passersby. No heavy rainfall seems to be associated with the triggering of these events and rainfall accumulations during November 2008 (transition from rainy to dry season) were below the monthly normal rainfall. Immediately prior to these landslides, there are no earthquake events located in this area by the Guatemalan seismic network. During November 2008, three cold fronts affected Guatemala, producing freezing temperatures in some locations especially during the third week of November. In the surroundings of the area of the 4 January landslide, frost susceptibility ranges from low to medium. The landslide took place in a catchment that follows a NNE fault which to the south intersects the EW Chixoy-Polochic fault (part of the transcurrent boundary of the North American and the Caribbean plates) 5 km downstream of the release area. Some fumarolic activity is currently observed in the landslide site. Based on the above elements and on observations from field reconnaissance missions, some hypotheses are formulated to explain the conditioning and triggering factors for the events in November and December and particularly for the 4 January landslide. These hypotheses are aimed to help to identify other potential instabilities in the surroundings. Back-calculation of flow parameters for the 4 January landslide has been possible based on estimation of velocities using video footage of the event and simulations with two different models for landslide dynamics across three-dimensional terrain: DAN3D developed at the University of British Columbia and RAMMS developed at the WSL Institute for Snow and Avalanche Research SLF. These back-calculated parameters have enabled the estimation of impact areas due to landslides that can be potentially released in surrounding slopes with similar instability features to the "Los Chorros" hill. Even though these estimations may be improved as more information becomes available for this event, the authors hope that this report contributes with some elements for decision-makers regarding short- and intermediate-term activities in response to this disaster.

In landslide risk research, the majority of past studies have focused on hazard analysis, with only few targeting the concept of vulnerability. When debris flows are considered, there is no consensus or even modest agreement on a generalized methodology to estimate physical vulnerability of the affected buildings. Very few quantitative relationships have been proposed between intensities and vulnerability values. More importantly, in most of the existing relationships, information on process intensity is often missing or only described semi-quantitatively. However, robust assessment of vulnerabilities along with the associated uncertainties is of utmost importance from a quantitative risk analysis point of view. On the morning of 13th July 2008, after more than two days of intense rainfall, several debris and mud flows were released in the central part of Valtellina, an Italian alpine valley in Lombardy Region. One of the largest muddy-debris flows occurred in Selvetta, a fraction of Colorina municipality. The result was the complete destruction of two buildings, and damage at varying severity levels to eight others. The authors had the chance to gather detailed information about the event, by conducting extensive field work and interviews with local inhabitants, civil protection teams, and officials. In addition to the data gathered from the field studies, the main characteristics of the debris flow have been estimated using numerical and empirical approaches. The extensive data obtained from Selvetta event gave an opportunity to develop three separate empirical vulnerability curves, which are functions of deposition height, debris flow velocity, and pressure, respectively. Deposition heights were directly obtained from field surveys, whereas the velocity and pressure values were back-calculated using the finite difference program FLO2D. The vulnerability was defined as the ratio between the monetary loss and the reconstruction value. The monetary losses were obtained from official RASDA documents, which were compiled for claim purposes. For each building, the approximate reconstruction value was calculated according to the building type and size, using the official data given in the Housing Prices Index prepared by the Engineers and Architects of Milan. The resulting vulnerability curves were compared to those in the literature, and among themselves. Specific recommendations were given regarding the most suitable parameter to be used for characterizing the intensity of debris flows within the context of physical vulnerability.

Entrainment of channel path material, and material deposition during run-out are key features of many rapid landslides like debris flows. Such mechanisms are able to change significantly the mobility of the flow, through rapid changes of the flow volume and of its rheology. Models using both a constant rheology and a constant volume cannot yield accurate forecast of debris-flows characteristics (velocity, discharge, flow height, spreading area), especially for debris flows occurring in heterogeneous torrential watersheds characterized by various geological settings and surficial deposits. The objective of this paper is to present and test a simple 1D debris-flow model with a material entrainment concept based on limit equilibrium considerations and the generation of excess pore water pressure through undrained loading of the in situ material. The debris flow model propagation is based on a one dimensional finite difference solution of a depth-averaged form of the Navier-Stokes equations of fluid motions. The flows are treated as a one phase material, which behaviour is controlled by different rheological characteristics depending on the liquid/solid ratio. In this model, users are able to implement a change in rheology at the onset of entrainment The model is tested on a debris flow event that occurred in 2003 in the Faucon torrent, and for which a detailed database on the sediment budget per reaches is available.

Estimating the magnitude and the intensity of rapid landslides like debris flows is fundamental to evaluate quantitatively the hazard in a specific location. Intensity varies through the travelled course of the flow and can be described by physical features such as deposited volume, velocities, height of the flow, impact forces and pressures. Dynamic run-out models are able to characterize the

Abstract Inside the framework of the European research network Mountain Risks, an interdisciplinary research group has been working in the Consortium of Mountain Municipalities of Valtellina di Tirano (northern Italy). This area has been continuously affected by several mountain hazards such as landslides, debris flows and floods that directly affect the population, and in some cases caused several deaths and million euros of losses. An aim of the interdisciplinary work in this study area, is to integrate different ...