The mechanisms of progressive landslides: Kashmir study (URGENCY APPLICATION)

Landslides are a highly destructive geological phenomena that are responsible for over 10,000 fatalities and $20 billion of economic loss per year worldwide. As with many other hazards, the impact of landslides in terms both of loss of lives and financial cost is increasing with time, and in general the largest impacts are felt by the poorest sectors of the global society. Unfortunately, at present our understanding of the processes that occur within a slope to cause failure is rather weak. Until we develop this knowledge it will be difficult to forecast where and when landslides will occur in the future, and to develop effective and reliable short- and medium-term warning systems. The Kashmir Earthquake, which struck Pakistan on 8th October 2005, triggered several thousand landslides. Some of these were very large, with the biggest, at Hattian, burying a village in 245 m of debris. It is thought that the landslides were responsible for about 20,000 deaths out of a total of about 86,000. These landslides continue to cause major disruption to life in the affected area. However, surprisingly the number of landslides that were actually triggered is much lower than would normally be expected for an earthquake of this size in a steep, mountainous area. It is likely that this is because the weather conditions before and indeed after the earthquake were exceptionally dry, which has meant that the water table has been low. However, our reconnaissance visit in January 2006 showed that many slopes across the affected area have extensively sheared and deformed by the earthquake, and are thus likely to fail in the future. As Kashmir has a monsoonal climate, this is most likely to occur in July 2006 when the rainfall as it its most intense. Thus, the Kashmir earthquake has created a unique set of circumstances in which the location and likely timing of large-scale landslides are predictable. We seek to use this opportunity by setting up a monitoring system on four of these slopes. This system will measure the movement of the slope with time at four points for each landslide. These data will be complemented by rain gauge information at each site, so that we know how much rain is associated with the movement events. We will also use a terrestrial laser scanning system to measure the geometry and morphology (shape) of the slopes, and conventional mapping to understand the geology. The aim is to look at the patterns of acceleration and deceleration as the slopes move towards final failure, and to determine how much movement occurs before final collapse. These results will be used to assess our existing models of landslide movement, and to refine and enhance them. In addition, we hope that our instruments can be used in the short- and medium-term by the Pakistan authorities to provide a basic warning system of impending collapses.