精确测量的地幔体波走时及地幔非均匀性的强度

用波形相关法精确地测定了在世界各地发生的87个6级以上地震的P波, PP波和Pdiff波的503个走时数据。记录这些地震波形的是新建于西太平洋地区的海洋半球地震观测网。我们利用这些高精度的走时数据研究了地幔体波的走时残差的范围及地幔非均匀性的强度。结果表明,P波、PP波和Pdiff波的走时残差最大分别为9 s ,11 s和15 s ,这为地幔层析成像反演中应该使用的体波走时残差数据的范围提供了重要信息。超出这一范围的走时残差数据不应该用于反演中,以免歪曲成像结果。我们发现,当震中距小于40°时,P波走时残差的范围为-6到+9 s。而对于40°到99°之间的震中距,P波走时残差的范围为-3到+5s。由于震中距越大,P波穿透地幔越深,我们这一结果提供了直接和确凿的证据,表明上地幔和地幔转换带中的横向非均匀性的强度要远胜于下地幔。我们精确测量的Pdiff波的走时数据表明,在地幔底部存在显著的低速异常,可能与地幔热柱或者超级地幔柱有关。我们使用了一个最新的三维全球层析成像模型来解释这些体波走时数据的空间变化。

PRECISELY MEASURED TRAVEL TIMES OF MANTLE BODY WAVES: IMPLICATIONS FOR MANTLE HETEROGENEITY AND TOMOGRAPHY

We have measured 503 arrival times of P, PP, and Pdiff waves from seismograms of 87 global earthquakes with magnitudes≥Mw 6.0,as recorded by the newly installed Ocean Hemisphere Project Seismic Network located in the western Pacific region.These mantle body wave data are used to investigate the range of variations in the travel time residuals which indicate the degree of mantle heterogeneity.Our results show that variations of the residuals amount to 9 s for P data, 11 s for PP data, and 15 s for Pdiff data.These findings provide useful information on the maximum travel time residuals for the mantle body waves and on the criteria of using travel times for tomographic inversions.Thus outliers in a data set which exceed the plausible range of travel time residuals can be excluded to avoid biasing the inversion result.Our P-wave data show a large range of travel time residuals (-6 to +9 s) for epicentral distances shorter than 40 degrees and a smaller range of residuals (-3 to +5 s) for longer epicentral distances.Because P rays with a larger epicentral distance penetrate deeper in the mantle,our results provide sound and direct evidence for the inference that the upper mantle and transition zone have much stronger heterogeneities than the bulk of the lower mantle.Our Pdiff data show the existence of significant slow velocity anomalies in the D″ layer, which may be associated with the generation of plumes and superplumes.A new global tomographic model is employed to interpret the spatial variations of travel time residuals of the observed P, PP and Pdiff data.