| dc.description.abstract | Landslide is one of the greatest natural catastrophes and poses a major threat to lives and property worldwide. As development expands into unstable hill areas under the pressure of increasing population and urbanisation, the risk of landslide has been growing all over the world during the last decades. The present work aims at risk analysis of landslides in an effort to reduce the socio-economic impact of landslides and to mitigate the landslide risk effectively. The risk analysis of landslides is neither simple nor straightforward as the effects of different triggering factors have to be analyzed individually and managed separately. In this view, landslide risk analysis requires a detailed understanding of the physical process of landslides and the development of different methodologies to relate landslide occurrences with the characteristics of triggering factors.
Rainfall and earthquakes are considered as the most common triggering factors for landslides. Due to the contrast in the characteristics of rainfall and earthquake loads, and their subsequent effects on the landslide movements, the evaluation of landslide risk by a single methodology becomes difficult, if not impossible. The present study introduces methods for risk analysis of landslides due to rainfall as well as earthquake events. These methods recognize the conditions that caused the slope to become unstable and the processes that triggered the landslide movement. Often landslides occur due to the combination of rainfall and earthquake events as the probability of concurrent rainfall and earthquake event in any area is not rare. Consequently, it is essential to evaluate the landslide risk by considering the unfortunate combination of independent events such as rainfalls and earthquakes, and the effects of their mutual interactions on landslide movements. Therefore, for regions prone to both earthquake and heavy rainfall events, risk analysis should take into consideration multiple processes rather than one single event to avoid underestimation of the threats caused by their potential interaction. Understanding the interaction of these events can form the basis and provide a rational approach for multi-hazard risk assessment. The methodologies developed in the present work aims to provide systematic and rigorous processes to formalize slope engineering practice and enhance slope management.
Uncertainty is implicit in almost every field of engineering. In geotechnical engineering, the uncertainty is mainly attributed to inherent or spatial variability of soil parameters, limited number of samples, testing and measurement errors, and the modelling techniques which relate the laboratory or in-situ properties with response characteristics in terms of stability and deformation behaviour of soil. Unless all the sources of uncertainty are clearly brought out and included in the risk analysis appropriately, it is not possible to explicitly assess the risk associated with landslides. To this end, probabilistic methods which have the potential to include the spatial and temporal variability of different components should be incorporated in the risk analysis of landslides. The need for sophisticated probabilistic methods for landslide risk analysis arises from the complex response of slopes to external loads and the uncertainties in material properties. Random field theory is employed to model the inherent spatial variability prevalent in naturally occurring soils. While the spatial variation of earthquake ground motions is modelled using the random process concepts. The probabilistic methods have the ability to identify mechanisms responsible for the occurrence of landslides while incorporating uncertainties in the analysis.
The present study also addresses the challenging issue of numerical simulation of large deformation problems in geomechanics. Material Point Method (MPM), which is a mesh-based particle method, is employed to simulate large deformations occurring in landslides. MPM simulates large displacement with Lagrangian material points moving through a fixed Eulerian mesh. MPM has been employed in the thesis to study the behaviour of saturated slopes under seismic excitation.
Documentation of the past landslide events provides the opportunity to advance the research and practice of landslide risk analysis and also in the development of techniques for slope remediation. In general, the timely collection of data from landslide sites is a challenging problem but very important for landslide risk analysis. The present study introduces a systematic framework which aims at timely and systematic collection of data from landslide affected areas. It also updates the information on strength parameters and other conditions existing in the slope after the landslide event based on an inverse analysis.
Overall, it is expected that the work reported in the thesis will furnish useful guidelines for 1) risk analysis of landslides due to rainfall, earthquake and the coalescence of rainfall and earthquake events 2) modelling the various sources of uncertainty in landslide problems 3) modelling large deformations which occur during landslide events, particularly landslides of flow types 4) collection of timely and systematic data of landslide events. | en_US |
