Seismograms recorded at antipodal distances (Δ~180°) are used to measure the attenuation and dispersion of surface waves. The antipode is a focus for surf ace wave energy radiated at all directions from the source. Antipodal records thus give direct estimates of global average properties. Laterally heterogeneous Earth structure degrades the focusing and complicates the data analysis.
Group velocity and Q of 120-300 s Rayleigh waves are obtained from a record (station PTO) of the May 23, 1968 New Zealand earthquake. The effects of aspherical structure on this record are simulated by generating synthetic seismograms for an ellipsoidal Earth. The results show that the bias in antipodal Q measurements varies with time. This property is used to constrain the bias in the PTO measurements.
A seismogram of the August 19, 1977 Indonesian earthquake from station TRN is used to study 250-500 s Rayleigh waves. This record is synthesized by combining first order perturbation theory with a realistic representation of the Earth's heterogeneities. Analysis of the synthetic shows that Q measured from the TRN record is biased as much as 20% and that group velocities are virtually unbiased. The PTO and TRN data yield global average values of Rayleigh wave Q which are comparable to or slightly below those for the model PREM [Dziewonski and Anderson, 1981].
Antipodal Love wave and spheroidal overtone data are also presented. Synthetic seismograms match the observed arrivals well, demonstrating that these wave modes can be profitably studied at the antipode.
