Seismic Tomography Models for Alaska: Validation, Iteration, and Complex Anisotropy

University Of Alaska Fairbanks Campus, Fairbanks AK

Investigators

Abstract

Seismology provides the best opportunity to investigate the interior structure of Earth. Seismometers at the Earth's surface record seismic waves from local and distant earthquakes, and these recordings can be used to derive three-dimensional images of the Earth's interior, from its innermost solid iron core to the surface. Creating these seismic images requires recordings from real earthquakes and from simulated earthquakes that calculate how waves propagate through realistic three-dimensional models of Earth's structure. By comparing real recordings with simulated recordings, it is possible to improve the models of Earth's structure. In this project, these simulations of seismic waves will be used to investigate previously established models of the subsurface structure of Alaska, as well as to develop new models based on recently recorded earthquakes. The results will enable a better understanding of the relationship between earthquakes and faulting in Alaska, while providing a more accurate method for determining the places where certain ground motion is expected to be relatively strong due to the presence of sedimentary layers at the Earth's surface. The project will establish general procedures for other scientists to interrogate Earth structure models made by other methods and data sets. The project promotes free and open software development and training opportunities, and it advances two topics relevant to society—computational science and elasticity—with potential benefits to seismic hazard assessments, oil and gas seismic imaging, materials science, and structural engineering. Seismic images of Earth's interior structure, also known as tomographic models, are commonly produced by approximate methods, and they are typically qualitatively compared to one another. These images and comparisons raise fundamental questions regarding the accuracy of the images as well as how to interpret them in the context of the compositional and thermal structure and dynamics of the Earth. This project addresses three facets of this problem. First, it will use seismic wavefield simulations to generate the simulated seismograms needed for comparison with the recorded seismograms. This will enable the most accurate physics to be deployed within the imaging problem. Second, it will apply wavefield simulations to a reference data set to directly and fairly compare previously derived tomographic models. Third, it will formalize these procedures and offer them to others by hosting a virtual workshop featuring how to access tomographic models from the Earth Model Collaboration and how to perform seismic wavefield simulations in the open-source software package Specfem3D. The focus of these efforts in on Alaska, which has had exceptional coverage of EarthScope seismic stations since 2017 and exhibits extreme subsurface tectonic complexity caused by collision at the eastern margin of the Aleutian-Alaska subduction zone. The availability of earthquakes, the quality of station coverage, the complexity of subsurface structure, and the complexity of subsurface geodynamics all provide motivation for using Alaska as a focus for the broader effort of validating and improving tomographic models. The efforts will consider a more complex representation of Earth structure, in the form of tilted transverse isotropy, which is present in minerals and rocks (for example, shale) but which is challenging to determine at larger scales, such as the crust and uppermost mantle. 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.