Group-velocity tomography of South America and the surrounding oceans

This paper presents the results of a study of the dispersion characteristics of broad-band fundamental-mode surface waves propagating across South America and the surrounding oceans. Broad-band waveform data from about 765 events from 1977 to 1996 recorded at 48 individual stations have produced 7000 Rayleigh-wave and 4800 Love-wave dispersion curves. We present group-velocity maps from 20 to 150 s period for Rayleigh waves and from 20 to 100 s for Love waves. Measurement uncertainties estimated from cluster analyses average about 0.025–0.030 km s− 1, but are larger for the short-period Love waves. We estimate surface resolving kernels and show that the average resolution across South America is about 6°–8° for Rayleigh waves and 7°–10° for Love waves below 100 s period but degrades at longer periods and near the periphery of the maps. The estimated maps produce a variance reduction relative to the Preliminary Reference Earth Model (PREM) of more than 90 per cent at short periods, more than 80 per cent for Rayleigh and Love waves below about 100 s period, but nearer to about 70 per cent at longer periods. Synthetic experiments are presented to estimate the bias caused by azimuthal anisotropy using the recent global model of Trampert & Woodhouse (1996). This bias, which is worse for Love than for Rayleigh waves, may be substantial (up to 5 per cent in some places), but it is expected to be largely off the continent and is not correlated with the main features of the observed maps. Many known geological and tectonic structures are observed in the group-velocity maps. Of particular note are the signatures of sedimentary basins (e.g. Maturin-Llanos Basin, Marañon-Ucayali-Madre de Dios complex, Chaco-Tarija Basin, S. Paraná Basin, the basins in the W. Caribbean and W. Gulf of Mexico), variations in crustal thickness (e.g. Andes, Altiplano, Brazilian Highlands), continental roots (e.g. Guyana and Guaporé shields, Sao Francisco Craton), and the Galapagos thermal anomaly. Comparison of the estimated group-velocity maps with those predicted by CRUST5.1/S16B30 is qualitatively good, but there are significant differences in detail that provide new information that should help to calibrate future crustal and upper-mantle models of South America.