THOR Projects for 2018
Coupled Hydro-Mechanical Analysis of Hillslope Failure: From numerical modelling and UAV-based field observations to landslide hazard assessment
We propose to advance our understanding of landslide processes, predictive models and quantitative hazard assessment, by studying inherent, destabilizing and triggering landslide factors; and monitoring failure/post-failure landslide mechanisms using UAV-based photogrammetry. We specifically plan to simulate the hydro-mechanical response of hillslopes using a hybrid finite-discrete numerical technique, capable of capturing continuously the pre-failure progressive deformation, the hillslope failure instability and the landslide post-failure discontinuous deformation (e.g. flow and toppling). We will use UAV-based photogrammetry and image correlation techniques to calibrate and update our numerical model; and we will synthesize our results into a probabilistic framework for quantitative displacement hazard assessment due to landslides.
Assessing Flood Hazards by Mapping Rivers with Autonomous Surface Vehicles (ASVs)
River floods are the single most costly natural hazard, resulting in an annual average of 50 billion dollars of damage worldwide over the past ten years. Rivers are conduits of water and sediment, and their evolving geometry – through erosion and deposition - determines in part where and when flooding will occur. Extreme weather events are becoming more common, and rivers have the potential to self-adjust their geometries to increase flow capacity and reduce the impact of floods. However, we currently lack data on river bed topography and flow rates for the vast majority of rivers globally. This project will leverage the existing approaches to planning, control, and state estimation to ensure feasible experiments and successful generation of results while creating a space for novel algorithm development (e.g. reducing run-time complexity for path planning through complex currents). More importantly, this project sits at the intersection of robotics and Earth sciences, and although in its infancy, has potential to generate mutually beneficial research to make contributions in both fields of study. As such, this project fits perfectly into CAST, in that novel robotics technology will be used to enable geological sampling studies that were previously not possible, and with THOR, as the project directly addresses a technological gap of direct relevance to the most costly natural hazard worldwide.
Understanding Near Fault Seismic Hazard and Tsunami Generation Mechanisms at Thrust Faults
Many large, destructive earthquakes occur along thrust faults (e.g., the 2011 Mw 9.0 Tohoku-Oki earthquake, the 2008 Mw 7.9 Wenchuan earthquake, and the 2004 Mw 9.2 Indian Ocean earthquake). Megathrust faults line the Pacific Rim and intraplate thrust faults lie near or within many major metropolitan centers, including Los Angeles. This highlights the risk that thrust earthquakes pose in terms of near-fault shaking and tsunami generation. In part, the shaking on the hanging wall can be much amplified by the interaction between the waves released by the rupture and the free surface, which can potentially even open the fault and may result in large fast slips and much larger ground shaking than typically envisioned. This is currently a topic of scientific debate and some controversy and has been the subject of a Nature paper recently published by the Rosakis group (Gabuchian et al., 2017)
We propose to study the ground motion associated with thrust faults using laboratory experiments and finite element numerical simulations. One of the main goals is to understand whether thrust faults might actually open as the earthquake rupture approaches the free surface. While our previous laboratory experiments indicated that fault opening is possible (Gabuchian et al., 2017), those measurements had a reduced spatial resolution and were based on a limited number of tests. Our newly developed ultrahigh speed imaging method allows us to capture the full-field behavior of dynamic ruptures (Rubino et al., 2017) and is an ideal technique to study thrust fault motion. Simultaneously, numerical simulations, making use of realistic frictional laws measured in the laboratory, will help us to design suitable experimental configurations as well as to explore a broader parameter regime that is possible in the lab.
Autonomous Drone-based Observation and Sampling of Volcanic Gases
We propose development of drone-based autonomous sampling of mantle-derived gases from high-temperature fumaroles at active volcanoes. The composition of such gases can provide unique information about the deep mantle sources of magmas and the long-term evolution of these sources; although the abundances of a full suite of gaseous species will be measured, we are particularly interested in the isotopic composition of xenon (Xe) for this purpose. Such autonomous collection of volcanic gases will be combined with monitoring capability of volcanic hazards.
Robotic Observations of Boundary Layer Turbulence in Marginal Ice Zones
Ocean observations are typically collected in one of three ways, all of which have significant limitations: (i) sea-going research cruises that are expensive and require extensive, at-sea human commitment; (ii) satellite observations that provide high temporal and broad spatial coverage, but only capture surface properties; (iii) autonomous floats that have no on-board control and often drift out of areas of interest. Increasingly, fleets of mobile heterogeneous platforms are being used to obtain observations over the broad range of temporal and spatial scales needed to understand those aspects of the ocean that strongly impact the climate system, e.g. upper ocean heat and carbon uptake or the dynamics of the rapidly-changing high-latitude marginal ice zones (MIZs). The goals of this project are (i) to improve our understanding of boundary layer turbulent fluxes in the ocean's MIZ and their impact on sea ice extent, and (ii) to explore the feasibility of using a combined in situ robotic and modeling approach to optimizing sampling strategies in the MIZ.