Open AccessRegional upper mantle S ‐velocity models from phase velocities of great‐circle Rayleigh waves
Author(s) -
Lévéque J. J.
Publication year - 1980
Publication title -
geophysical journal of the royal astronomical society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.302
H-Index - 168
eISSN - 1365-246X
pISSN - 0016-8009
DOI - 10.1111/j.1365-246x.1980.tb02608.x
Subject(s) - geology , great circle , rayleigh wave , geodesy , earth structure , geophysics , oceanic basin , mantle (geology) , phase velocity , tectonics , seismology , surface wave , geometry , physics , optics , mathematics
Summary A linear inversion approach is used to interpret a wide set of great‐circle Rayleigh wave phase velocities in terms of regionalized Earth models. In a first step ‘pure‐path’ fundamental Rayleigh wave phase velocities are estimated by linear regression from great‐circle measurements of phase velocity in the period range 150‐300 s, using both published and new observations. For these data a method based on the deviations of observed eigenfrequencies from those for a spherically averaged model of the Earth, is used after subtraction of the effect of ellipticity in the range of geometrical optics. A regionalization of the Earth is fixed a priori , models of variation of structure with depth being sought for each region; this regionalization is based on a variation with age, both for oceans and continents. Four regions are distinguished: (1) ‘young ocean’ regions (age less than 30 Myr), (2) ‘old ocean’ regions (age greater than 30 Myr), (3) ‘shield and platform’ regions and (4) ‘tectonic’ regions. In a second step, the ‘pure‐path’ phase velocity curves are then interpreted in terms of S‐wave velocity models by a linear inversion scheme. The resolution of the data with regard to surface structure is discussed, and care has been taken to constrain continental and oceanic crustal structures in the starting models. In the upper 250 km, the well‐known strong difference between oceanic and continental structures clearly appears in the resulting models. In the depth range 300‐450 km, no resolvable differences appear between the ‘shield and platform’ and ‘old ocean’ models; yet, slight differences between the ‘young ocean’ and the ‘old ocean’ models are indicated. Also for the ‘tectonic’ region, which includes both subduction zones and mountainous areas, the model contains a 2 per cent higher velocity zone between 300 and 450 km depth. The results are in agreement with independent regional studies and lead to the conclusion that deep lateral S ‐velocity variations are related to recent tectonic processes.