 |
| Titel |
Using Distant Sources in Local Seismic Tomography |
| VerfasserIn |
Bruce Julian, Gillian Foulgr |
| Konferenz |
EGU General Assembly 2014
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 16 (2014) |
| Datensatznummer |
250090344
|
| Publikation (Nr.) |
 EGU/EGU2014-4574.pdf |
|
|
|
|
|
| Zusammenfassung |
| Seismic tomography methods such as the “ACH” method of Aki, Christoffersson & Husebye
(1976, 1977) are subject to significant bias caused by the unknown wave-speed structure
outside the study volume, whose effects are mathematically of the same order as the
local-structure effects being studied. Computational experiments using whole-mantle
wave-speed models show that the effects are also of comparable numerical magnitude
(Masson & Trampert, 1997). Failure to correct for these effects will significantly corrupt
computed local structures.
This bias can be greatly reduced by solving for additional parameters defining the shapes,
orientations, and arrival times of the incident wavefronts. The procedure is exactly analogous
to solving for hypocentral locations in local-earthquake tomography. For planar incident
wavefronts, each event adds three free parameters and the forward problem is surprisingly
simple: The first-order change in the theoretical arrival time at observation point B
resulting from perturbations in the incident-wave time t0 and slowness vector s
is δtB - δt0 + δs -
rA = δtA, the change in the time of the plane wave at the
point A where the un-perturbed ray enters the study volume (Julian and Foulger,
submitted). This consequence of Fermat’s principle apparently has not previously been
recognized.
In addition to eliminating the biasing effect of structure outside the study volume, this
formalism enables us to combine data from local and distant events in studies of local
structure, significantly improving resolution of deeper structure, particularly in places
such as volcanic and geothermal areas where seismicity is confined to shallow
depths.
Many published models that were derived using ACH and similar methods probably
contain significant artifacts and are in need of re-evaluation. |
|
|
|
|
|
|