Common-Reflection-Point-Based Prestack Depth Migration for Imaging Lithosphere in Python: Application to the Dense Warramunga Array in Northern Australia

Abstract

Abstract We exploit estimates of P-wave reflectivity from autocorrelation of transmitted teleseismic P arrivals and their coda in a common reflection point (CRP) migration technique. The approach employs the same portion of the vertical-component seismogram, as in standard Ps receiver function analysis. This CRP prestack depth migration approach has the potential to image lithospheric structures on scales as fine as 4 km or less. The P-wave autocorrelation process and migration are implemented in open-source software—the autocorrelogram calculation (ACC) package, which builds on the widely used the seismological Obspy toolbox. The ACC package is written in the open-source and free Python programming language (3.0 or newer) and has been extensively tested in an Anaconda Python environment. The package is simple and friendly to use and runs on all major operating systems (e.g., Windows, macOS, and Linux). We utilize Python multiprocessing parallelism to speed up the ACC on a personal computer system, or servers, with multiple cores and threads. The application of the ACC package is illustrated with application to the closely spaced Warramunga array in northern Australia. The results show how fine-scale structures in the lithospheric can be effectively imaged at relatively high frequencies. The Moho ties well with conventional H−κ receiver analysis and deeper structure inferred from stacked autocorrelograms for continuous data. CRP prestack depth migration provides an important complement to common conversion point receiver function stacks, since it is less affected by surface multiples at lithospheric depths.