Upper mantle Q structure beneath the East Pacific Rise

Abstract

A model of Qβ as a function of depth has been determined for the upper mantle beneath the East Pacific Rise. The model is appropriate for shear waves with periods from roughly 15 to 30 s. The amplitudes of multibounce SH phases which propagate within the upper mantle were used to constrain the Q structure. The multibounce phase amplitudes were measured with respect to the S wave amplitude for each seismogram. The data consist of S, SS, SSS, and SSSS waves for distances from 30° to 80°. The sources are earthquakes which occurred on transform faults associated with the East Pacific Rise and the seismograms were recorded by World‐Wide Standard Seismograph Network and Canadian Seismic Network stations in western North America. Mantle beneath oceanic crust less than 15 Ma and tectonically active continental crust is sampled. Since the S wave data primarily sample the deep mantle, the resulting upper mantle Q model depends on the Q structure assumed for the lower mantle. We assumed a constant Q of 312 for the mantle beneath 650 km depth. With this assumption, the upper mantle phase amplitudes require a Q of about 70 in the upper 150 km of the mantle. From 150 to 400 km depth, Q increases to 180. The Q in the transition zone, from 400 to 650 km depth, was found to be 180, significantly lower than the lower mantle Q. The resulting Q model is consistent with several surface wave studies of mantle attenuation beneath the Pacific. However, our results predict a QScS (210) which is significantly higher than many direct measurements of ScS attenuation indicate. Furthermore, the large range in measured QScS is inconsistent with lateral Q variations being confined to the upper mantle if the region we have studied represents the low Q extreme of upper mantle. We conclude that large lateral variations in Q exist in the lower mantle.

Attenuation Tomography of the Earth's Mantle: A Review of Current Status

“…In general, models obtained to date at the global scale concur with regional scale models on the existence of a correlation of lateral variations in Q with tectonic province, in the first 250 km of the upper mantle, with high Q under shields and low Q under oceans, particularly so under young oceans. Surface wave derived models (ROMANOWICZ, 1990(ROMANOWICZ, , 1995DUREK et al, 1993) tend to indicate high attenuation under young oceans (as confirmed by regional studies, e.g., DING and GRAND, 1993). ROMANOWICZ (1994b) demonstrated the correlation of Q with the age of the sea floor in the first 250 km of the upper mantle, both in the Pacific and in the Atlantic Oceans (Fig.…”

“…The existence of large lateral variations of Q in the crust (see MITCHELL, 1995 for a review) and the upper mantle, is well documented from various regional studies, either using surface waves (e.g., MITCHELL, 1975;MITCHELL, 1978, 1981;NAKANISHI, 1979a;BUSSY et al, 1993) or body waves, such as multiple ScS phases (NAKANISHI, 1979b;SIPKIN and JORDAN, 1980;LAY and WALLACE, 1983;CHAN and DER, 1988;SIPKIN and REVENAUGH, 1994) or multiple S and depth phases (SHEEHAN and SOLOMON, 1992;WIENS, 1990, 1994;DING and GRAND, 1993). These lateral variations can be an order of magnitude larger than observed lateral variations in velocity, commonly exceeding 50-100%.…”

Attenuation of Broadband P and S Waves in Tonga: Observations of Frequency Dependent Q

“…Improved data coverage and measurement techniques have led to numerous models of attenuation on both global scales (DUREK et al, 1993;ROMANOWICZ, 1995;BHATTACHARYYA et al, 1998) and regional scales (REVENAUGH and JORDAN, 1991;SHEEHAN and SOLOMON, 1992;DING and GRAND, 1993;SIPKIN and REVENAUGH, 1994;FLANAGAN and WIENS, 1994;MITCHELL, 1995). A universal feature of these models is the presence of a low-Q asthenosphere and upper mantle with lateral variations in attenuation which are quite strong.…”

Comparison between Time-domain and Frequency-domain Measurement Techniques for Mantle Shear-wave Attenuation

“…In this study, we have chosen a frequency-independent attenuation model which is equivalent to those used in similar analyses of long-period seismic data (BHAT-TACHARYYA et al, 1996, 1998DING and GRAND, 1993;FLANAGAN and WIENS, Figure 11 A comparison between time-domain (a) and frequency-domain (b) measurements for teleseismic S waveforms. We observe considerably more scatter in the time-domain measurements, probably due to biased picks.…”

“…Several body-wave attenuation studies involve the analysis of long-period waveforms to constrain Q i variations in the mantle (SIPKIN and JORDAN, 1980;REVENAUGH and JORDAN, 1991;BHATTACHARYYA et al, 1996BHATTACHARYYA et al, , 1998WALLACE, 1983, 1988;CHAN and DER, 1988;SHEEHAN and SOLOMON, 1992;DING and GRAND, 1993;REID and WOODHOUSE, 1997;SIPKIN and REVENAUGH, 1994;WIENS, 1990, 1994). The attenuation of seismic body waves is typically obtained from their pulse shapes.…”