Sub-project 4: Landslide disaster scenarios predictions and loss modelling 

Evaluation of the risk due to landslides needs a good understanding of the geological setting, material behaviour, and physical mechanisms, as well as the use of adequate, flexible computational models to make the predictions. Experience shows that getting sufficiently reliable parameters is one of the main challenges in risk analysis. Crucial for a better comprehension and actual management of the risk is a strategy combining the definition of the predisposing and the natural or human triggering factors, suitable conceptual models, and the identification of relevant physical input properties and calibration requirements. Being concerned with landslide risk assessment, this Sub-Project is closely related to Sub-Projects 1, 2 and 3, and its main objectives will be: (i) to perform a probabilistic and deterministic assessment of both non-seismic and seismic landslide risk scenarios at a few representative sites selected and characterised in Sub-Projects 1; (ii) to select and calibrate landslide loss models, and to evaluate losses by implementing new procedures of early warning systems in relation with Sub-Projects 1, 2 and 3 and finally, (iii) to recommend a best practice of risk assessment for landslides. The Sub-project is coordinated by BRGM, and features also the participation of AUTH, NGI, and SGI-MI.

Disseminated material related to this Sub-Project can be found by clicking each of the links below:

Deliverables

Presentations

Reports

Publications

Events

Meetings

A summary of each of the tasks involved in this Sub-Project is provided below: 

Task 1.4.1: Characterisation and hazard assessment

Hazard assessment, i.e. the determination of the probability of occurrence of unfavourable events for a given location and time exposure, is one of the first steps in the landslide risk assessment method. It is largely dependent on the availability and the quality of the (geo-referenced) geological, geo-morphological, and geotechnical data. In this task, and in close relation with task 1.1.4 of Sub-Project 1, we propose to assess landslide hazard at 2-3 landslide sites affected by significant current movement potential, either by rainfall or earthquake triggering, and by presence of population centres and transportation or other infrastructures. The major activities will consist of:

• Improving and validating an existing deterministic modelling tool for predicting hydraulic- or earthquake-triggered slide movements on large areas. For this purpose, a multilayered St Venant formulation will be implemented, replacing the sliding materials by equivalent viscous fluids (e.g. viscoplastic fluids, etc.), considering frictional conditions between layers and at the top and bottom of the slide, and accounting for time evolution of the pore-pressure within the sliding mass.
• Establishing relevant seismic and non seismic risk scenarios for the selected landslide sites; • Applying deterministic and probabilistic approaches within the landslide scenarios identified for selected sites;
• Evaluating future movements and induced damage on these sites;
• Checking the reliability of simplified approaches for predicting future movements at the selected sites, based on geological and gross geomorphological and geotechnical data.

Task 1.4.2: Loss estimation models for urban areas (mainly lives) and displacement thresholds for lifelines

As part of the risk assessment approach, consequences (i.e. losses) of landslides in terms of permanent ground displacements and their effects to environment, urban areas and infrastructures are to be estimated, using the methodologies, methods and tools developed in all previous tasks of the project. This task will essentially aim at producing a critical review of existig models for estimating human losses in urban areas and possible damage to lifelines. The output of this task is also useful for Sub-Projects 10 and 11, dealing with urban areas and infrastructures.

Task 1.4.3: Early warning systems to calibrate loss models and evaluate losses

Data is needed to develop methods for loss model calibration. To reduce risk due to landslides, systmes are to be developed, capable of monitoring long term slide evolution and ground deformations, and to forewarn of an impending hazard. The reliability of loss models can be verified and quantified only through predictions with the models and, where possible, through comparisons with actual landslides, hopefully new ones (prototype measurements). In close relation with Sub-Project 1, this task will aim at:

• Prioritising verifications and calibrations that need to be done;

• Designing the verification programme;
• Calibrating loss models from Task 1.4.2 using early warning systems;
• Evaluating the consequences of landslides using the calibrated loss models within the scenarios established in Task 1.4.1 for the selected landslide sites.

Task 1.4.4: Recommended best practice for landslide risk assessment

Slide hazard and risk mapping constitute an important task and have to be carried out to evaluate risk levels. Procedures for mapping slide hazard and risk need to be implemented, inclusing selection of criteria for setting priorities. Experience shows that getting sufficiently reliable parameters (e.g. on geology, material behavior, physical mechanisms), that will be useful for complex computational models and for mapping is one of the main challenges in risk analysis. Hence, the major research effort in this task will be to provide a synthesis of the outcome of Sub-Projects 1 to 4, addressing the following main questions:

• Which, where and when should the proposed deterministic/probabilistics modelling approaches for ground displacement predictions be applied, depending on geological settings and triggering forces (seismic/non-seismic)?
• What are the useful criteria for landlslide hazard and risk ampping, considering sliding potentail and consequences of sliding for different triggering mechanisms and types of slides?
• What are the current methods for establishing parameter uncertainty (e.g. statistical data, engineering judgement)?
• How can safety and risk levels for specific slopes and various slide types be quantified, notably those related in Sub-Project 1?