P AND PCP TRAVEL-TIME TOMOGRAPHY FOR THE CORE-MANTLE BOUNDARY

Citation
M. Obayashi et Y. Fukao, P AND PCP TRAVEL-TIME TOMOGRAPHY FOR THE CORE-MANTLE BOUNDARY, J GEO R-SOL, 102(B8), 1997, pp. 17825-17841
Citations number
79
Categorie Soggetti
Geochemitry & Geophysics
Journal title
JOURNAL OF GEOPHYSICAL RESEARCH-SOLID EARTH
ISSN journal
21699313 → ACNP
Volume
102
Issue
B8
Year of publication
1997
Pages
17825 - 17841
Database
ISI
SICI code
2169-9313(1997)102:B8<17825:PAPTTF>2.0.ZU;2-Y
Abstract
PcP arrival times reported to the International Seismological Centre ( ISC) are inverted for vertical travel time anomaly at the core-mantle boundary (CMB). We first invert about 2 x 10(6) ISC first P arrival da ta for the three-dimensional (3-D) structure of the mantle and hypocen ter locations iteratively. The PcP data are corrected for the relocate d hypocenters and the 3-D structure, This correction has reduced the o bserved PcP residuals. The corrected PcP data are then inverted for th e vertical travel time anomalies at the CMB, which are due either to v elocity anomaly or topographic anomaly. The spherical average of the i nverted structure requires either a velocity reduction in the boundary layer just above the CMB (by similar to 3% if the layer is 20 km thic k) or a lowering of the CMB by 3 km. The heterogeneity pattern of the CMB is drastically different from that of the overlying D '' layer. Ne gative vertical travel time anomalies appear in the northern high-lati tude region, and positive anomalies appear under Southeast Asia and Mi ddle to South America. The lateral dimensions of these anomalies are o f the order of 40 degrees. The anomalies can be interpreted as either a velocity perturbation in the boundary layer (+/- similar to 7% if th e layer is 20 km thick) or an undulation of the CMB (+/- similar to 7 km). Various tests including the error and resolution estimations indi cate that these three anomalies are the significant, resolvable featur es of the CMB. These anomalies are also corroborated by the result of the joint inversion of P and PcP for both the 3-D mantle and the 2-D C MB. Assuming that the layer is laterally heterogeneous in iron content and deforms at the base in isostatic equilibrium with the outer core, the PcP data are also inverted for the isostatic figure of the bounda ry layer. The velocity perturbation in this case is 5 similar to 6% if the layer is 20 km thick, and the CMB undulation is only +/- 1 simila r to 2 km. Regardless whether the CMB heterogeneity is due to velocity and/or topographic perturbations, it is dominant in the zonal compone nt of degree 2, while the degree 2 pattern in the lower mantle, includ ing the D '', is dominant in sectorial component.