Radial variation of compressional and shear velocities in the Earth's lower mantle

Abstract

This paper extends an earlier study (Sengupta & Julian) of travel times of P waves of deep-focus earthquakes to include shear waves. Primary advantage of deep-focus earthquakes is the reduction of anomalies caused by complex structures near the source. The standard deviations of travel times and station anomalies of this study are about half as large as those determined from the data of shallow-focus earthquakes (e.g. Herrin et ul.;Hales & Roberts). Spherically-symmetric velocity models derived from the travel times by a linearized inverse technique have resolving lengths of about 70 km for standard errors in velocity of about 0.02 km/s. No pronounced reversal of either compressional or shear velocity was required at the base of the mantle to satisfy the data, though a small velocity decrease could not be entirely ruled out. Some anomalous rapid changes in compressional velocity gradient were, however, found centred around the depths of 2400 and 2600 km. The models derived in this study agree most closely with that of Herrin er ul. for compressional velocity and the model 1066B of Gilbert & Dziewonski for shear velocity.

Deep Earth Structure: Lower Mantle and D″

“…Global observations of body-wave travel times and measurements of slopes of the travel time curves by seismic array analyses proliferated in the 1960-1990s, with many radially symmetrical Earth models for the lower mantle being produced (e.g., Chinnery and Toksö z, 1967;Dziewonski and Anderson, 1981;Hales and Roberts, 1970;Hales et al, 1968;Herrin, 1968;Johnson, 1969;Kennett and Engdahl, 1991;Morelli and Dziewonski, 1993;Randall, 1971;Sengupta and Julian, 1978;Uhrhammer, 1978). While the variations in lower mantle velocities at a given depth among these models are < 1%, there is still great importance in having an accurate reference model both for earthquake location procedures and for use as a background model in tomographic analyses.…”

Petrology, elasticity, and composition of the mantle transition zone

“…, , I , , , • , , , I , , , I , , , I , , , I , , , • , , , • , 6 and 7. Also shown are the mean and standard error of V• in the upper mantle based on 34 P wave velocity (Vp) profiles and 32 $ wave velocity (Vs) profiles generated from high-resolution travel time studies or waveform modeling studies of body waves or surface waves [Chinnery and Toksoz, 1967;Herrin, 1968;Fairborn, 1969;Randall, 1971;Buchbinder, 1971;Wright and Cleary, 1972;Gilbert and Dziewonski, 1975;Hart, 1975;Kind and Muller, 1975;Dey-Sarkar and Wiggins, 1976;Hart et al, 1976;King and Calcagnile, 1976;Fukao, 1977;Burdick and Helmberger, 1978;Sengupta and Julian, 1978;McMechan, 1979McMechan, , 1981Uhrhammer, 1979;Given and Helmberger, 1980;Hales et al, 1980;Burdick, 1981;Dziewonski and Anderson, 1981;Vinnik and Ryaboy, 1981;Fukao et al, 1982;Nakada and Hashizume, 1983;Grand and Helmberger, 1984a, b;Walck, 1984Walck, , 1985Lyon-Caen, 1986;Lerner-Lam and Jordan, 1987;Paulssen, 1987;Grad, 1988;Graves and Helmberger, 1988;Montagner and Anderson, 1989;Lefevre and Helmberger, 1989;Bowman and Kennett, 1990;…”

Extremal bounds on the seismic velocities in the Earth's mantle

“…The linearized techniques provide uncertainties in velocity and depth at a particular point in a model (e.g. Backus & Gilbert 1970;Johnson & Gilbert 1972a, b ;Sengupta & Julian 1978), rather than establishing a confidence region for the entire model. It will be worthwhile to briefly discuss some of the relationships between the observational r(p) and velocity-depth domains.…”