Seismic imaging of continents and their margins: New research at the confluence of active and passive seismology

Now that the world has well and truly crossed the threshold to the 21st century, it seems timely to review the progress of lithospheric seismic imaging and comment on recent developments in the field. In the context of this special issue, lithospheric seismic imaging refers to any kind of seismic study that gives rise to multidimensional (2-D or 3-D) views of the Earth's interior, from the shallow crust to the base of tectonic plates. For active sources, the primary tools for delineating Earth structure include reflection seismic imaging, reflection tomography, wide-angle reflection and refraction tomography and cross-hole tomography. In cases involving modest numbers of unknowns (e.g. in 2-D), forward modelling approaches are often also used with reflection and wide-angle data. In the case of passive sources, local earthquake tomography, teleseismic tomography, ambient noise tomography, and receiver function migration are all capable of retrieving lithospheric structure. With the notable exception of Peaceful Nuclear Explosions (Nielsen et al., 1999; Priestley et al., 1994; Ryberg et al., 1996), active source imaging is largely confined to the crust and mantle lithosphere. Its use is widespread in the exploration industry, particularly in delineating sedimentary structures favourable to hydrocarbon entrapment (e.g. Sheriff and Geldart, 1995). On a broader scale, active source imaging has been widely used to vastly improve our understanding of crustal and uppermost mantle structure and processes, and enjoys a history over many decades involving hundreds of experiments around the globe (Prodehl and Mooney, 2012; Thybo et al., 2011). Passive source imaging, which historically has exploited seismic energy from earthquakes, is capable of elucidating structure at a variety of scales, from the shallow crust to the whole Earth (Rawlinson et al., 2010a). Compared to active source imaging, the cost of data acquisition is generally much lower, but this advantage is offset by the spatial and temporal irregularities of natural seismicity, which have a profound effect on data coverage. Although active and passive source studies are often carried out independently, it has long been recognised that potential overlaps exist, and a joint treatment of both classes of data may yield improved results compared to separate analyses. For example, both Ankeny et al. (1986) and Thurber (1983) include refraction traveltimes from explosive sources in a local