Subduction zones‐their influence on traveltimes and amplitudes of P‐waves

SUMMARY The aim of this study is to demonstrate what effects subduction zones, i.e. laterally inhomogeneous, high‐velocity anomalies in the Earth's mantle, can have on the traveltimes and the amplitudes of teleseismic recordings, and to show that the amplitude anomalies are large enough that they can be used in the determination of the structure of subducting slabs. The method used for the computation of the seismograms in these laterally inhomogeneous media is the Gaussian beam method which, in contrast to the ray method usually used, gives correct amplitudes at caustics and in critical regions. The restriction to 2‐D models in the downdip direction of the slab, i.e. the assumption that there is no change along strike of the subduction zone, reduces the number of free parameters. For the short‐period waves considered here it is sufficient that the section of the slab be uniform along strike for a few hundreds of kilometres. This geometry is satisfied by many subduction zones. Furthermore we expect pronounced effects of the subduction zones to be in the downdip direction because the wave will then sample the anomaly to a large extent. The influence of the following parameters of the slab on teleseismically recorded traveltimes and amplitudes is studied in detail: maximum velocity anomaly, depth of penetration, thickness of the slab and dip change. The influence of the source location, i.e. depth and epicentre, is also considered. We compare seismograms from this modelling to those computed from a radially symmetric reference earth and observe the following effects: negative traveltime residuals, occurrence of caustics and focusing h2reasing the magnitude by up to +0.7 magnitude units, defocusing and the creation of shadow zones. Receiver regions with large amplitude anomalies are not necessarily regions where large traveltime anomalies occur. Regions with little change in their traveltime anomalies can show dramatic changes in amplitude (larger than Δ m b =± 1). Small structures of subduction zones are better resolved by amplitude information than by traveltime information. A combined interpretation of traveltimes and amplitudes yields much better resolution of the structure of subduction zones because traveltimes and amplitudes contain independent information about the structure studied. This helps to eliminate ambiguities which may not be resolved when using only one type of information (traveltimes or amplitudes) especially in aseismic parts of slabs where no additional information from seismicity is available. Observations of amplitudes for deep Kurile events recorded at European stations show anomalies that can be explained by a slab with a penetration depth of 670 km, thus indicating the usefulness of this approach in the determination of subduction zone structure.