This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).
This project involves an investigation of different technical approaches used to image the Earth's interior. Much of what is known about the structure, composition and temperature of the deep Earth comes from mapping of elastic properties using seismic waves from earthquakes recorded on hundreds of seismic stations distributed worldwide. All of the computational tools and techniques used to image the deep Earth are approximate, and significant uncertainty exists concerning the accuracy and precision of models that are the result of so-called seismic tomography. This project uses synthetic data sets, generated using a highly accurate computer code and a prescribed Earth structure, and standard analysis techniques to image the interior structure using the synthetic data. Comparison of the prescribed and recovered Earth structures allows the true resolution of common analysis techniques to be investigated and quantified. The project aims to develop guidelines for the applicability and limitations of different techniques.
Large data sets of long-period seismograms for a suite of Earth models with different elastic and anelastic characteristics are generated using the spectral-element method. Realistic distributions of earthquakes and seismic stations are used to mimic real conditions, and earthquake mechanisms are determined directly from the data. Surface-wave propagation anomalies are measured from the data with standard techniques, and the anomalies are interpreted in terms of Earth structure employing methods of different levels of theoretical and computational complexity. In particular, the performance of ray theory and two-dimensional and three-dimensional Born scattering approaches are compared for different three-dimensional Earth models.
