THOR Projects for 2017

Seismic hazard assessment of the Kathmandu basin via nonlinear ground motion, earthquake cycle and rupture simulations (continuation)

( Asimaki and Ampuero)

On April 25, 2015, Nepal was hit by the magnitude 7.8 Gorkha earthquake. Contrary to the expected devastation for such a large magnitude event in a region with such poor construction practice, the observed structural damage, ground failure effects (e.g. landslides and liquefaction) and recorded ground motions showed only moderate shaking intensities throughout most of the near-field region. The death toll, while very high, was orders of magnitude lower than the number of casualties caused by similar magnitude earthquakes in regions with comparable infrastructure. The few isolated incidences of severe damage were generally associated with topographic amplification along ridge tops or with topography and basin edge effects at the outskirts of Kathmandu.  The fact that the Gorkha earthquake was not nearly as devastating as we originally feared, raises nonetheless numerous urgent questions at the interface between earthquake engineering and basic earthquake research that this project addresses.

Dispersion of a natural hazard: ash clouds


April 14th 2010, the Eyjafjallajökull volcano in Iceland erupts; an enormous ash cloud forms 8km into the atmosphere; the European airspace is closed for six days causing the largest air travel disruption since WWII; about 10 million travelers are affected. The Icelandic eruption was not even the most “explosive”; it was only a 4 on the Volcanic Explosivity Index (VEI). For comparison, Mount St Helens in 1980 was a 5, and Mount Pinatubo in 1991 was a 6. August 17th 2013, the Rim Fire starts in the Stanislaus National Forest in California; a week later, the flames get within miles of the Hetch Hetchy water reservoir and ash falls on the water surface; the reservoir provides 85% of the water consumed by the 2.6 million of people in San Francisco. These are just two of many examples of natural hazards involving ash formation and transport in the atmosphere. The USGS calls ash fall “the most widespread and frequent volcanic hazard” [1]. Unfortunately, predicting the evolution of clouds of ash (and other small particulates) is particularly difficult because these particulates behave differently from the surrounding gas. The important questions to answer are how fast and how far will an ash cloud disperse? In fact, it is not sufficient to know where the ash cloud is; we need to know the ash density distribution within the cloud. To answer these questions, we need a mechanistic understanding of the evolution of ash concentration in turbulent flows characteristic of those found in the atmosphere.

Exploring the potential for glacier monitoring with seismic noise interferometry


Glaciers pose hazards at both short and long time scales. Glacial lake outburst floods (GLOFs)  and glacial avalanches/surges can cause severe damage to human and infrastructures in large  areas within minutes to hours. The most deadly glacier disaster, the 1970 glacier avalanche from Mount Huascarán in Peru, killed more than 6000 people in the town Yungay. At long time scales (years to decades), glaciers respond to climate change, cause global sea level change, and influence availability of water resources. For example, nearly a billion people in Bhutan, Nepal, China, and Indian rely heavily on meltwater from Himalayan glaciers for hydropower generation, agriculture, and ecosystems. Furthermore, the two time scales are linked: recent evidences show that global warming appears to enhance the short time scale hazards, such as GLOFs and surges in high-mountain areas (Bolch et al., 2012).

In this proposal, we will focus on glacial movements  at short time scales, such as surges and avalanches. The physical causes behind these sudden glacier movements are still unclear, which makes them difficult to predict/forecast. Numerous evidences  show that they probably involve changes in glacier basal conditions. In the seminal 1987 paper, Kamb proposed that glacial surge is due to switch of basal water system from concentrated large tunnels to a distributed “layer” as “connected cavities”. The higher  water pressure in the distributed system reduces friction and causes accelerations in ice flow. Unfortunately, this hypothesis has not been fully tested due to spatially and temporally limited data about glacial basal conditions (Tsai et al., 2016).

Therefore, to understand the physical mechanism behind rapid glacial sliding, and mitigate related glacial hazards, we need to develop new methods to complement existing methods (e.g., remote sensing, GPS, borehole measurements) so as to provide continuous monitoring of the glacial sliding interface, especially the distribution of water on the interface.