P-Wave Scattering in the Earth's Crust and Upper Mantle

Abstract

The codas which follow P-phase onsets on seismic records can be interpreted in terms of waves scattered from random inhomogeneities in the Earth's crust and upper mantle. In particular, the so-called precursors to PP have travel times and slownesses consistent with the interpretation of these signals as waves scattered from direct P-waves incident at the Earth's surface at distances greater than about 20" from either the source or receiver. Significant differences between recordings at the Warramunga array of PP precursors from two Novaya Zemlya explosions are extremely difficult to account for on any interpretation previously suggested, but are fully consistent with the random variations to be expected on the proposed scattering mechanism.The present interpretation accounts for the major observed features of P codas, including their durations and the variations in slowness along the trains. Precursors to P'P' can also be interpreted as scattering phenomena. Thus the interpretation removes the need to postulate sharp reflecting discontinuities in the uppermost few hundred kilometres of the upper mantle.

P-signal complexity and upper mantle structure

“…We believe, on the contrary, that the absorption model of Douglas, Marshall & Corbishley (1971) was initially based on a misidentification of later arrivals at GBA, and that the model may not be generally applicable to the problem of P-signal complexity. It is our opinion that a more consistent interpretation of the other examples given by Douglas et al (1973) is provided by the 'scattering model' recently proposed in explanation of the body-wave coda by Cleary, King & Haddon (1975 and Douglas et al (1973) show P-signals recorded at the UKAEA arrays GBA, India and EKA, Scotland from an underground explosion near Bukhara, USSR. Whereas the P arrival at EKA (A = 47.4") is the simple, 'coda-less' signal typical of most explosion records at teleseismic distances, the GBA record (A = 27.4') has a first arrival (P) with a slowness of about 9.2 s/deg, followed about 4 s later by a larger-amplitude arrival (PHI), with a slowness of about 10.3 s/deg.…”

“…4, because as epicentral distance increases (a) the scattering angle is increased, and (b) the energy scattered from any specific volume arrives over a longer time interval, so that there will be proportionately less energy in the first 30 s of the coda (cf. Cleary et al 1975 , Figs 1 and 2). In terms of this model, complexity is a function, not of Q variation, but of the scattering properties of the upper mantle between source and receiver.…”

“…We may note, however, that in the three examples of complex records in their Novaya Zemlya series (as illustrated in their Figs 3 and 4), the proportion of high-frequency to low-frequency energy does not change in any marked way from the direct P signal to the coda, which is contrary to the prediction of the absorption model. 5 Coda complexity and the scattering model Cleary, King & Haddon (1975) have proposed a model for the generation of the coda train in which energy is scattered from ordinary ray paths by relatively small random inhomogeneities in the crust and uppermost mantle, each element of the inhomogeneous medium becoming effectively a source of scattered waves of the same frequency as the primary wave (Chernov 1960). The amplitudes of the scattered waves depend in general on the characteristics of the random medium, the primary wave amplitude at the point of scattering, the wave frequency, the angular deflection from the ordinary ray path, the volume of secondary sources contributing to the signal, and the usual geometrical spreading effects.…”

P-Wave Velocities in the Upper Mantle Beneath Fennoscandia and Western Russia

“…2(j), (k), (1) and (m)). Scattering by small-scale random irregularities in the crust and upper mantle has been suggested previously as the dominant mechanism in the generation of coda at both short distances (Aki & Chouet 1975) and teleseismic distances (Cleary, King & Haddon 1975;King et al 1975), so it appzars likely that the coda observed at intermediate distances is cognate. If so, the differences in signal duration between the ef first arrivals and the ab arrivals suggest that scattering irregularities may be particularly pronounced near 420-km depth which seems plausible and is consistent with the pronounced distance extension of the ab branch as discussed above.…”

The use of the Born approximation in seismic scattering problems

“…Similarly, the coda of P and precursors to PP have been attributed to strong single scattering (see Husebye 1975 andCleary, King &). a% axq cjipq -ni = q(x), x E g2, where + A?2 = A#, the internal boundary of 22, and a radiation condition: ui(x) -U: (x) represents outgoing waves.…”