Po/So Synthetics For A Variety of Oceanic Models and Their Implications For the Structure of the Oceanic Lithosphere

Abstract

S U M M A R Y Polso synthetic seismograms were computed for a variety of oceanic structures in order to model data from an t?zb = 5.7 earthquake, recorded during the OSS IV seismic experiment. Satisfactory modelling of Polso waveshapes and frequency content was achieved with no lateral heterogeneity, but with a small random vertical heterogeneity in the mantle, superimposed on a mean velocity structure which is consistent with seismic refraction data. The random heterogeneity in the P and S velocities appears to be about 2 per cent with a scale length of 5 km. This kind of heterogeneity is easily achieved by varying slightly the major mineral components in either the peridotite or eclogite mantle mineralogy and can be another cause of observed upper mantle anisotropy.Using a step-function source time behaviour, our numerical modelling also indicates that in the upper mantle Q p is roughly proportional to rising from about 450 at f = 1 Hz to about 800 at f = 15 Hz, and that Q, is roughly proportional to rising from about 900 at f = 1 H z to about 1800 at f = 15Hz. A slight decrease in Q, from these values in any small zone of the upper mantle reduces So amplitudes drastically and we believe this is why So is sometimes missing from observed seismograms.The computed synthetic record sections were also used to examine the relation between intrinsic Q and the apparent Q inferred from data in earlier studies. Apparent Q was found to be lower than intrinsic Q below about 5 H z and higher than intrinsic Q above 5 Hz. When these differences are accounted for, the mantle Q's of this study agree with the mantle Q's of Butler et al. (1987) to within 10 per cent. There can no longer be any doubt that oceanic upper mantle Q's are very large, above 5 Hz, and that Q, is about twice as large as Qp.

Upper mantle Q structure beneath old seafloor in the western Pacific

“…Estimates of Q p and its frequency dependence for P n in the western Pacific vary widely and are based on a variety of different techniques including falloff rates of the coda, reduction in amplitude with distance along a linear array, maximum distance the phase is observed, and spectral ratios [Oliver and Isacks, 1967;Walker et al, 1983;Butler et al, 1987;Sereno and Orcutt, 1987;Brandsdóttir and Menke, 1988;Mallick and Frazer, 1990;Roth et al, 1999;Kennett and Furumura, 2013;Shito et al, 2013], although the general consensus is that Q p for P n in the lithosphere at 10 Hz is on the order of 1000 or greater. The thickness of the low-attenuation lithosphere is even less clear.…”

Toward the reconciliation of seismological and petrological perspectives on oceanic lithosphere heterogeneity

“…But, such a purely stratified structure produces rather distinct pulse‐like arrivals rather than the continuous coda observed (as illustrated in Kennett and Furumura [, Figures 7c and 7d]). Further using just such 1‐D variation requires a much stronger level of variability to produce the same strength of coda than is required for 2‐D and 3‐D heterogeneity [ Mallick and Frazer , ; Kennett and Furumura , ].…”

T waves from the great 1994 Bolivian deep earthquake in relation to channeling of S wave energy up the slab

“…The timing of the T phase at RAR can be explained only through propagation as an S wave between Mangaia and Rarotonga. Efficient propagation of high frequency oceanic upper mantle phases (Po and So) has long been documented [Walker, 1977;Talandier and Bouchon, 1979] and modeled [Mallick and Frazer, 1990], but it remains unclear why the bulk of the T wave hitting Mangaia would convert to So rather than Po. Mangaia is an unusual island, featuring exceptionally old exposed volcanics [Turner and Jarrard, 1982], and an uplifted limestone ting, with a small recent fringing reef [Yonekura et al, 1988].…”