SEISMIC IMAGING OF SUBDUCTED SLABS - TRADE-OFFS WITH DEEP PATH AND NEAR-RECEIVER EFFECTS

The most promising procedures for determining detailed velocity structure of subducting lithosphere (seismic tomography and residual sphere analysis) require isolation of near-source contributions to total travel-time anomalies of seismic waves from earthquakes in the slab. Effects of heterogeneous structure along the raypaths through the deep mantle, lithosphere and crust beneath the recording stations must be removed from the data, using either empirical procedures or aspherical models. Comparisons of observed S wave travel-time anomalies for earthquakes in the Kurile slab with both empirical path corrections and corrections computed for a high resolution tomographic shear velocity model beneath North America indicate that near-source contributions are significantly smaller than suggested by recent slab modeling efforts that fail to fully remove the distant path effects. Imaging deep slab velocity structure and depth of penetration is likely to prove more difficult than originally expected, given that slab velocity heterogeneity may be relatively weak at depths greater than 300 km.