- © 2014 by the Seismological Society of America
Online Material: Discussion, figures of resolution matrix analysis of alternative cell geometries; alternative station distributions.
When geoscientists study the structure and history of old orogens, basins, and cratons, information from teleseismic data may provide important constraints on the present distribution of seismic velocities. The existing networks of permanent observatories are typically quite sparse, so temporary arrays are installed for a limited period of time with carefully designed measurement geometry. A dense grid of stations should ideally be deployed, but logistics, like road access and coastlines, combined with limited resources, commonly result in a distribution of the available seismometers along profiles perpendicular to the supposed strike direction of the target structures. This study focuses on the tomographic resolution of velocity structure from teleseismic travel‐time residuals. In particular, we identify some potential pitfalls when using a conventional tomographic model with regular cells, and we use a ray‐adapted model type that provides optimized resolution with much fewer artifacts.
The resolution and reliability of such an image depend on the density of stations (receivers), the diversity of azimuths of incoming information, and parameterization (shape and size of the cells) of the study area. Generally, the teleseismic ray incidence angles are rather steep, so vertical smearing in tomographical models is a problem that cannot be avoided but must be reduced as much as possible. Earthquakes (sources) are mostly located at plate boundaries. The period for data acquisition is limited so that, in most cases, the datasets are clearly biased with respect to azimuths and distances. A global‐scale illustration of this problem is shown in Figure 1 in which events with Mw>5.5 occurring in 2006 are plotted. Rose diagrams for every 25° of latitude by 50° of longitude show the azimuthal distribution of teleseismic P arrivals, that is, for source–receiver distances between 30° and …