Frequency-Dependent Nature of Pn in Western China: Gaussian Beam Modeling of Data from the Hi-CLIMB Experiment (original) (raw)

In realistic models of the crust and the upper mantle where vertical gradients in wave speeds and major interfaces are present, propagation of Pn, an important seismic phase at regional distances, involves complex effects of interference. Such effects can result in wave-trains whose frequency-contents and amplitudes vary with distance in counter-intuitive ways. In this new project, we are investigating the propagation of Pn beneath western China using data from Hi-CLIMB (An Integrated Study of the Himalayan-Tibetan Continental Lithosphere during Mountain Building). This experiment is one of the largest broadband seismic experiments to date, with more than 210 deployments at close station-spacing of 3-8 km over a distance of 800 km. The linear array is complemented by a regional array of comparable aperture, producing an unprecedented dataset for Eurasia. For the current project, which just began in May of 2008, we are organizing a dataset for regional seismic events recorded by the Hi-CLIMB arrays and will show examples of long seismic profiles over apertures of 500 km from several different azimuths. We are also investigating methods of modeling and inversion based on Gaussian beams (GB), which offer several distinct advantages. First, GB modeling can be applied both for interference waves, which have caustics, and for pure head waves. Second, GB modeling can handle laterally varying media, an important aspect that cannot be investigated by standard methods such as reflectivity. Third, GB is computationally efficient, suitable for analyzing large datasets. We will illustrate our approaches with data from an earlier experiment using explosions. Based on numerous events recorded by the Hi-CLIMB array, we will investigate frequency-dependent propagation of Pn over a large region in western China using GB and also finite-difference schemes. Attributes of seismic wave-trains, including arrival times, amplitudes of signal-envelopes, and instantaneous pulse frequencies will be used to constrain how structures in the crust and the upper mantle affect the propagation of Pn. The objective is to achieve self-consistent models of Pn propagation in western China which are free from assumptions such as a frequency-independent factor for geometric spreading. To this end, our results should advance efforts in isolating effects of frequency-dependent propagation from those of pure-inelastic attenuation (Q), leading to improved methodologies for discrimination and yield estimates at regional distances.