Multiple ScS Research Articles

Multiple ScS attenuation and travel times beneath western North America

Variations in attenuation and travel time of multiple ScS whole-mantle reverberations indicate very strong lateral gradients in upper-mantle properties beneath the tectonically active province of western North America. Whole-mantle attenuation estimates (Qscs) obtained by spectral stacking of phases sampling discrete subregions span the entire range previously determined using a similar procedure for a global set of continental and oceanic Qscs measurements. The relationship between travel time and attenuation anomalies, which has been used to argue for predominantly thermal control on the variations in oceanic environments, is poorly defined within the continental regime. However, a correlation between surface heat flow and Qscs variations in the United States indicates that thermal variations do control attenuation heterogeneity. For paths beneath Mexico, stacking 13 pairs of ScS, phases from the 24 October 1980 Huajuapan, Mexico and the 19 June 1982 El Salvador events yields Qscs - 142 _+ 10 in the frequency band 7.8 to 46.9 mHz. Time-domain measurements for the same phases yield a similar estimate of Qs~s = 148 __. 9. The average SCSn+l -- ScS n JB (Jeffreys-Bullen) travel-time residual (htscs) is +0.9 sec (slower than JB). All of the differential travel-time estimates are corrected for dispersion effects in the passband of the observations, and are appropriate for a reference period of about 30 sec. Multiple ScS arrivals from the 29 April 1965 Puget Sound earthquake recorded by WWSSN, CSN, LRSM and Pasadena stations provide the first determination of whole-mantle attenuation beneath the westernmost United States. Pasadena recordings from three instruments provide ScS, paths traversing the Cascade and Sierra Nevada ranges. Stacking 9 ScS, pairs yields Qscs = 208 __. 39 (7.8 to 54.7 mHz) and the same phases give htscs = +2.5 sec. Six pairs of ScS, reverberations under the Pacific Northwest give Qs¢s = 344 - 88 (15.6 to 54.7 mHz) and Atscs = +5.5 sec. A combined estimate for the relatively low heat flow region of the Pacific borderlands gives Qscs = 257 __. 20 (7.8 to 46.9 mHz), comparable to the high Qscs estimates previously determined for South America. On the other hand, 12 SCSn pairs sampling the northern Basin and Range, which has high heat flow, yield a very low Qs~s = 95 -_. 4 (7.8 to 31.3 mHz) with htscs = +5.1 sec. The latter Qscs value is even lower than estimates for young oceanic regions, suggesting extensive partial melting in the low-velocity zone under the northern Basin and Range.

Attenuation of multiple ScS in various parts of the world

A waveform and amplitude matching scheme is used to study regional variations of QScS at low-frequency band between 0.02 and 0.1 Hz for paths across various parts of the world. The scheme estimates attenuation by convolving the first well-recorded, clear ScSH phase with multiples of trial t* operators until the phases so derived match the amplitude decay-rates and waveform characteristics of the observed later multiples. Thirty-five intermediate-to-deep focus earthquakes recorded at seismographic stations throughout the world were used in this study. The regions sampled are the tectonic and shield areas of Eurasia, tectonic regions of North America, shield regions of South America, and tectonic portions of NW, SW and E Pacific Ocean. In general, the regional differentials in t* between successive multiples of ScS are similar to those reported in earlier studies (Nakanishi 1979; Spikin & Jordan 1980). For the tectonic regions of Eurasia, t*ScS are 5.0 ± 1.1 s for a double passage through the mantle, whereas for the shield areas, they are around 2.8 ± 0.4 s. For the tectonic regions of North America, our estimates of t*ScS are 4.8 ± 1.6 s. A mean t*ScS of 6.3 ± 0.9 s has been obtained for paths across the E Pacific while the NW and SE Pacific are characterized by moderate attenuation with average t*ScS values of 4.4 ± 0.7 and 4.2 ± 0.8 s respectively. For central South America, the estimated t*ScS is 2.7 ± 0.7 s. Assuming that the lateral variations in Q observed using the ScS data are confined to a 200–300 km thick layer in the upper mantle, where the lateral velocity variations are also the most pronounced, these results imply lateral variations in shear wave Q of about an order of magnitude. The lateral variations of mantle Q in the upper mantle derived from these analyses of long-period ScS phases correlate well with mantle Q estimates from short period data. It is therefore unlikely that Q(f) appropriate to various types of upper-mantle structures converge at low frequencies of 0.02 to 0.1 Hz as required by the ‘absorption-band shift’ models.

Observations of first-order mantle reverberations

AbstractWe have observed first-order mantle reverberations, specifically SH-polarized ScSn and sScSn phases reflected at near-normal incidence from upper mantle discontinuities, as discrete phases on long-period digital seismograms of the HGLP and SRO networks. Such arrivals correspond to an n or n + 1 member dynamic ray family denoted by {ScSn, Sd±S}, where n is the multiple number of the parent phase, “d+” signifies a reflection from the top of an internal discontinuity at depth d, and “d−” signifies a reflection from the bottom. The travel times and attenuation of these phases place important constraints on the nature of the transition zone. We have employed the phase equalization and stacking algorithm of Jordan and Sipkin (1977) to obtain the differential attenuation operators of {sScSn, S650+S} − sScSn and {ScSn, S650−S} − ScSn phase pairs from deep-focus Tonga events recorded by the HGLP station KIP on the island of Oahu. The apparent Q's of the zeroth-order and first-order reverberations over the frequency band of 10 to 30 mHz have been inverted for the average quality factors of the upper mantle (QUM) and lower mantle (QLM) and the reflection coefficient of the 650 km discontinuity (R650). At 20 mHz, the results are: QUM = 82 ± 18, QLM = 231 ± 60, and |R650| = 0.080 ± 0.004 (at normal incidence). The QLM estimate is significantly less than recent normal mode solutions, suggesting that lower mantle attenuation structure is frequency-dependent at very low frequencies. We have measured the differential travel times between the first-order reverberations {sScSn, S400+S} and {sScSn, S650+S}, which yields a direct estimate of the vertical shear-wave travel time between the discontinuities along the Tonga-to-KIP path. The observed two-way time, 92.2 ± 1.5 sec, is less than predicted by some recent oceanic upper mantle models, indicating either a smaller separation of the discontinuities and/or greater shear velocity in the transition zone. Measurements of differential travel times between first-order reverberations and multiple ScS phases suggest that the former is more likely and are consistent with depths of 400 and 650 km for the transition zone discontinuities. Observations of first-order mantle reverberations can potentially provide unique constraints on lateral heterogeneity of the earth's mantle, especially within the transition zone.

Frequency dependence of Q in the mantle underlying the shield areas of Eurasia, Part 11: analyses of long period data

Summary. Long-period multiple S and ScS phases observed in northern Europe were analysed to determine mantle attenuation in the 0.02 to 0.2 Liz range under the Eurasian shield. Two groups of' events were used: deep FarEastern earthquakes and large earthquakes near the edges of the shield areas of Eurasia. The Q of the upper mantle under the Eurasian shield region was estimated in the time domain by taking amplitude ratios and in the frequency domain by taking spectral amplitude ratios among various arrivals. Under shield regions, SS/S amplitude ratios and multiple ScS amplitude ratios give ti - 2.5-3.0 s in the 0.02 to 0.20 Hz band. The results show that the upper mantle Q under this shield area is higher than the global average, but less than the Q values inferred from our studies of short-period data in the same area. Preliminary results suggest significantly higher /: under tectonic regions in the same frequency band.

Estimation of the frequency dependence of Q from ScP and ScS phases

Summary. A measurement of the relative frequency content of ScP and ScS phases from deep earthquakes provides a simple indication of the average value for Q in the mantle. When the ScP converts to the compressional mode at the core-mantle boundary it begins to attenuate at a slower rate than the ScS. Assuming that the core reflections are frequency-independent, the spectral ratio of the two phases provides a measure of the spectral ratio of the P and S attenuation operators. The average value of mantle Q can be inferred from this information if it is assumed that negligible losses occur in pure compression. Both rays travel downward through the anomalous source region as S. Q is measured on the upward leg of the raypath well away from the subduction zone. The two raypaths diverge somewhat in the lower mantle, so they sample different portions of the core-mantle boundary. They converge closely again in the upper mantle, however, where most attenuation takes place. Clear ScP-ScS observations have been collected from the short period records at 25 WWSSN stations in a variety of tectonic regimes. The relative waveshapes have been analysed in terms of a standard linear solid Q operator. The low-frequency level of Q was set to be consistent with Q measurements from multiple ScS phases. The parameter controlling the increase of Q at high frequency, r,, has been determined by minimizing a quantitative measure of fit between ScP waveforms filtered with standard linear solid Q operators and observed ScS waveforms. The frequency band of the observations is from about 0.5 to 1.5 Hz. The observations constrain T, to be between 0.0 and 0.2s. This implies that the rapid increase of Q with frequency does not occur until frequency exceeds 1 Hz. Good quality ScP-ScS pairs appear to be much more common in the WWSSN record library than good quality ScS-ScS, pairs. As more ScP-ScS data are analysed, we should, therefore, be able to constrain the average mantle Q better in the ScP-ScS frequency band than in the multiple ScS band. No clear correlation exists between the ScP-ScS Q measurements and the tectonic regime of the recording sites. ScS phases are not more strongly attenuated with respect to ScP at oceanic island sites than at stable continent

Multiple SCS travel times and attenuation beneath Mexico and Central America

Multiple ScS phases recorded for the June 19, 1982 El Salvador earthquake (Ms = 7.0) are analyzed to determine whole mantle attenuation and upper mantle travel time anomalies beneath Mexico, Central America and northern South America. Recordings from the ultra long‐period seismographs at Pasadena and Berkeley and the SRO station ANMO provide eight ScSn arrivals with 15 bouncepoints concentrated within Mexico. The whole mantle QScS for these paths is estimated as 145 ± 27 in the frequency range 8 to 40 mHz, and the average ScSn+1−ScSn JB residual (ΔTScS) at a period of 50 sec is +1.5 sec. Eight ScSn phases of similar periods recorded at GDSN stations in South America yield a QScS estimate of 232 ± 89 and ΔTScS = −3.9 sec for paths beneath Central America, and QScS = 93 ± 17 and ΔTScS = +0.3 sec for paths beneath northern South America. These results indicate that the variations in travel times and attenuation beneath continental regions are as pronounced as differences reported between old oceanic and stable shield regimes.

Reassessment of a reported S‐delay under Trindade

We present a correction to a paper by Okal and Anderson (1975) about multiple ScS travel‐time anomalies. We have reanalyzed data for ScS2 surface bounces in the South Atlantic Ocean. From these data an ScS2‐S residual of 23.6 seconds was found by Okal and Anderson (1975). This corresponded to an ScS2 surface bounce point under Trindade island and was inferred to be due to very slow upper mantle associated with the Trindade hot spot. The analysis we present here invalidates this conclusion. The nature of the upper mantle under Trindade is an open issue.

Multiple ScS travel times in the western Pacific: Implications for mantle heterogeneity

Multiple ScS travel times have been obtained by wave form cross correlation from seismograms digitally recorded by the High Gain Long Period (HGLP) and Seismic Research Observatory (SRO) networks. The surface projections of the paths corresponding to these data cross the western Pacific on oceanic crust greater than 100 m.y. old or traverse continental regions. The difference between the median ScSn‐ScSn−1 residuals for all western Pacific paths and all continental paths is +5.2 s, in agreement with our World Wide Standardized Seismograph Network (WWSSN) data (Sipkin and Jordan, 1976). These results support the hypothesis that the average mantle shear velocity of old ocean basins is significantly less than that of old continental nuclei. The medians of both the oceanic and continental residuals for the HGLP and SRO data are more positive than those for the higher‐frequency WWSSN data by amounts consistent with attenuative dispersion, which we take to be direct evidence for such dispersion. The residuals for paths crossing China have a median 2 s greater than the median for all continental paths, supporting the inference from dispersion studies that the upper mantle beneath China is characterized by anomalously low shear velocities. The residuals for western Pacific paths show lateral variations of 5 s or more not correlated in any systematic way with crustal ages along the paths. An analysis of these variations suggests that for horizontal scale lengths of the order of 103 km the amplitude of lateral variability is greater along a SW‐NE axis than along a SE‐NW axis. Mesoscale heterogeneity in the western Pacific may thus consist of predominantly NW trending structures.

Attenuation of multiple ScS waves beneath the Japanese Arc

36 pairs of multiply-reflected ScS waves from deep earthquakes around Japan are analyzed to investigate the anelastic properties of the mantle on the continental and oceanic sides of the dipping slab. The average Q-value for shear waves passing through the mantle on the oceanic side is found to be 226 in the frequency range 10–40 mHz. This Q-value is in good agreement with the Q models SL8 (Anderson and Hart, 1978) and QBS (Sailor and Dziewonski, 1978) which have been derived from free oscillation data. Assuming that the Q-value for the mantle deeper than 400 km on the continental side of the Japanese Arc is the same as that for the model SL8, we obtain a value of Q = 53 in the upper mantle above the dipping slab beneath the Sea of Japan. Higher Q-values are obtained for the mantle behind the northern Izu-Bonin arc.

Frequency dependence of QScS

abstract Data from HGLP instruments at KIP and MAT and WWSS LP and SP instruments at KIP and GUA have been used to study the amplitude characteristics of ScS and multiple ScS waves from deep-focus earthquakes. The data at low frequencies (0.006 to 0.06 Hz) are consistent with our previously published estimate, Q ScS = 156 ± 13 (Jordan and Sipkin, 1977). However, at high frequencies (>0.1 Hz), Q ScS appears to increase rapidly with frequency. Lower bounds on Q ScS are obtained by assuming a flat source spectrum and ignoring any energy losses due to scattering; we find that Q ScS must be greater than 400 at frequencies between 1 and 2.5 Hz. Correcting for a source spectrum with a corner at 0.16 Hz and an asymptotic roll-off of ω −2 , considered appropriate for these events, raises this estimate to about 750. The increase in Q ScS at frequencies above 0.1 Hz is consistent with a spectrum of strain retardation times which has a high-frequency cutoff in the range 0.2 to 1.0 sec. At very low frequencies Q ScS can be estimated from normal mode data; the best available models yield values of about 230. Comparison of these estimates with our data suggests that Q ScS decreases with frequency in the vicinity of 0.01 Hz. Because the scattering coefficient increases rapidly with frequency, the fact that significant ScS amplitudes are observed at high frequencies implies that any bias in Q ScS measurements due to scattering at low frequencies is probably small. We show that, although our data provide only integral constraints on the variation of Q μ with depth, the regions in which Q μ is frequency dependent occupy a substantial portion of the mantle, probably including at least part of the mantle below 600 km depth.

Lateral variations in shear velocity and attenuation in the upper mantle

We have used digitally recorded long period seismic data from the HGLP and SRO networks to obtain estimates of multiple ScS travel times and spectra and have used these to place constraints on the lateral variations in upper-mantle shear velocity and attenuation structure. The travel time results are consistent with our previous findings: to a first approximation, upper-mantle heterogeneity as expressed in vertical travel time differences can be correlated with surface tectonic features. A spectral stacking procedure for retrieving the attenuation operator for SH polarized multiple ScS waves has been applied to seismograms representing a variety of tectonic provinces. The spectral analysis has yielded stable estimates of the phase and amplitude response of the attenuation operator in the frequency interval 6–60 mHz. These estimates indicate that the apparent Q can also be correlated with surface tectonic regimes.

A source of bias in multiple ScS differential travel times determined by waveform correlation

Differential travel times determined by waveform correlation techniques between seismic phase pairs suffer an inherent systematic bias with respect to conventional visual timing methods if the phase pairs have different attenuation histories. Synthetic shear waves are generated with attenuation histories appropriate for multiple ScS waves. The bias between the ScS2‐ScS differential time determined by visual onset and a differential time picked by waveform cross‐correlation ranges between 2.2 and 3.8 seconds late, for an assumed average Qβ of the mantle of 312 and 156, respectively. Corrections are calculated for the various ScSn−ScSn−1 phase pairs for a range of ScS t* values.

Estimation of the attenuation operator for multiple ScS waves

A procedure for retrieving the attenuation operator for SH‐polarized multiple ScS waves by spectral stacking is outlined. Application of the procedure to 17 multiple ScS phase pairs from 5 western Pacific deep‐focus events digitally recorded at Kipapa, Hawaii, and Matsushiro, Japan, yields the following results: (1) Stable estimates of the phase and amplitude response of the attenuation operator are obtained in the frequency interval 6 mHz ‐ 60 mHz; (2) Within this band the apparent Q for ScS waves multiply reflected beneath the western Pacific is 156 ± 13. No significant variation of Q as a function of frequency is observed; (3) Within this band no significant dispersion of multiple ScS waves is observed. However, given the uncertainties in the estimates of the phase response, the existence of the dispersion predicted by linear, causal, constant‐Q models is compatible with the observations.

Lateral heterogeneity of the upper mantle determined from the travel times of multipleScS

We have extended our investigation of lateral variations in upper mantle shear velocity structure by employing multiple ScS phases. An analysis of 108 multiple ScS differential travel times indicates that to a first approximation, upper mantle heterogeneity as expressed in vertical, travel time differences can be correlated with surface tectonic features. Travel times for continental regions with Phanerozoic orogenic histories are significantly greater than those from Precambrian continental regions and correlate with heat flow. Oceanic regions have large delays relative to stable continental regions, and the travel time residuals show a consistent decay as a function of crustal age. The average one-way vertical shear wave travel time difference between oceans and stable continental regions is estimated to be +4.0±0.5 s, not significantly different from the value obtained in our previous study. These results support the hypothesis that shield ocean heterogeneity persists to great depths, probably exceeding 400 km. For shear waves with characteristic frequencies of 0.1 Hz, the two-way vertical travel time through the mantle of the spherically averaged earth is 937.3±1.0 s, about 3 s less than the value derived from eigenfrequency data and 1.6 s greater than the Jeffreys-Bullen value. This result is inconsistent with the hypothesis that the base line discrepancy between free oscillation models and classical body wave travel time models is because of continental bias in the travel times, but it is consistent with the hypothesis that the base line discrepancy is an expression of anelastic dispersion.

A study of lateral inhomogeneities in the upper mantle by multiple Scs travel‐time residuals

An analysis of records of multiply reflected ScS phases from ten deep‐focus earthquakes yields near‐vertical one‐way travel‐time residuals varying from +5.0 to −3.5 s. Continental and oceanic residuals overlap and both indicate large lateral variations. Similar values are found for the older oceanic basins and for continental shields. Most, if not all, of the variations can be attributed to differences in the lithosphere and asthenosphere. The mantle under Hawaii, Iceland and Trindade (South Atlantic) Islands is anomalously slow, all being postulated hotspots. In the case of Trindade, petrological data is consistent with a pronounced low‐velocity zone and large S delays. Some recent studies assumed that stations on oceanic islands were representative of the ocean as a whole and it has been suggested that continent‐ocean differences extend deeper than 400 km. The present results indicate that oceanic islands are anomalous, and that differences between oceans and continents need not extend below 200 km. This does not rule out the existence of deep lateral inhomogeneities, but only the proposal that continents, in general, are faster than oceans to depths greater than 400 km.

Attenuation of shear waves in the upper and lower mantle

abstract The attenuation of seismic waves is a direct measure of the absorption due to nonelastic processes in the earth. The well known difficulties in obtaining body wave amplitude decrement data have been avoided by studying the spectral ratios of multiple ScS and sScS phases from two deep focus earthquakes recorded at near normal incidence. The average Q, for shear, in the mantle is about 600 for the frequency range 0.015 to 0.07 cps. Assuming that equal radiation occurs upwards and downwards from the source the average Q for the upper 600 km of the mantle is determined to be about 200 and about 2200 for the rest of the mantle. The value for Q at the base of the mantle is at least 5000 for shear waves.