Prøveforelesning
Se prøveforelesningBedømmelseskomité
Associate professor C.J. van Westen, International institution for Geo-Information Science and Earth Observation, The Netherlands
Professor Steinar Nordal, Dept. of civil and transport engineering, NTNU
Professor Henning Dypvik, Department of Geosciences, University of Oslo
Leder av disputas: Professor Nils Roar Sælthun
Veileder: Farrokh Nadim, Kaare Høeg og Anders Elverhøi
Sammendrag
Landslides are among the most important natural hazards affecting people and property located in the vicinities of steep terrain. Worldwide, 3.7 million km2 of area is exposed to landslides. The exposed population amounts to 300 million or 5% of the world population. High risk zones for landslides comprise 820000 km2 with a population of 66 million people. Recent statistics indicate that up to 80% of landslides are triggered by rainfall and are responsible for up to 90% of the total casualties due to this hazard. Rainfall-induced landslides are highly destructive because they can reach high velocities (>100 km/h) and long distances (several kilometres). From the perspective of Geosciences, this thesis has developed two methods for risk management of rainfall-induced landslides: a method for early-warning systems and one for mapping of hazards. The first method consists of a new mathematical model for rainfall thresholds which has proved to reduce false alarms in 90-100% compared to traditional methods. This new mathematical model has also been applied to the assessment of impacts of climate change on the occurrence of landslides. Also, a new method for mapping of landslide hazards has been developed using a novel statistical technique for selecting the optimum set of parameters for simulating potential landslides in the future. The research was carried out at the Department of Geosciences of the University of Oslo and at the Norwegian Geotechnical Institute, within the framework of the International Centre for Geohazards, one of the Norwegian Centres of Excellence. The studies were motivated by the need to improve the efficiency of early-warning systems for landslides and to develop more objective methods for mapping of landslide hazards. The data for calibration and validation of the methods were obtained from landslide-prone areas in El Salvador, Hong Kong, Nicaragua, the Philippines and Puerto Rico. The methods were implemented using GIS (Geographical Information System) tools and computer programmes specifically prepared for this research.
Kontaktperson
For mer informasjon, kontakt Anne Innes.