Reevaluating the Experimental Foundation for the Rheology of Crust-Forming Minerals

点击翻译按钮获取中文摘要

The Earth's crust is constantly deforming in response to applied forces, and at depths where temperatures and pressures are high (typically about ten miles or more below the surface), rocks slowly flow (or "creep") rather than breaking as they do at shallower depths. The mathematical relationship describing this creep is called a "flow law", and it depends on the flow laws of the rock's component minerals. Many experiments have been done to define mineral flow laws, but they all have one drawback: the experiments have to be completed within months rather than millions of years, so samples must be deformed ten billion times faster than in the Earth. To get around this problem, the equations derived from laboratory experiments have to be as precise as possible. Korenaga will address this issue by applying sophisticated statistical methods to obtain new mineral flow laws from experimental deformation data that has been produced over the years by other scientists. He will recruit a group of undergraduate students to do this analysis for a wide range of minerals, which will give them valuable research and data science experience. Using these new flow laws, geophysicists will be able to more confidently apply their numerical models to understanding plate tectonics and how slow deformation deep in the Earth leads to earthquakes.Incorporating realistic rock mechanics via ductile flow laws has become increasingly common in geophysical modeling, as seen in recent studies on the dynamics of continental lithosphere. However, considerable extrapolation is involved when using experimentally-derived rheology in numerical modeling. Strain rates attained in laboratories are usually on the order of 10-5 s-1, which is ten orders of magnitude faster than geological strain rates (10-15 s-1). When estimating a flow law from rock deformation data, therefore, it becomes essential to conduct a rigorous statistical analysis, by considering all experimental uncertainties, so that a flow law can be trusted for extrapolation over ten orders of magnitude. This project plans to conduct a series of reanalysis of published deformation data for crust-forming minerals, capitalizing on the Markov Chain Monte Carlo (MCMC) inversion method that has been developed over the last decade to investigate the rheology of olivine The project is based on the two-fold potential of the MCMC-based reanalysis of published deformation data. First, a short-term project using MCMC inversion has a definite appeal for physics undergraduates who already have familiarity with the technique, and a reanalysis project can serve as their entry point for earth sciences. The use of MCMC in data analysis is now part of standard data science, so this practical aspect also helps to attract undergraduates with solid quantitative skills. Second, producing a series of reanalysis results, through multiple summer intern projects, will build a collection of case studies for the deformation of different minerals and rocks. The aim of building a collection of case studies is to provide a “critical mass” so that the community of crustal dynamics, including numerical modelers, rock mechanicists, and field geologists and geophysicists, will recognize the importance of the rigorous analysis of rock deformation data.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.