 |
| Titel |
Strength of megathrust faults and its control on subduction-zone seismotectonics |
| VerfasserIn |
A. Tassara, R. Hackney |
| Konferenz |
EGU General Assembly 2009
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 11 (2009) |
| Datensatznummer |
250029265
|
|
|
|
|
|
| Zusammenfassung |
| Predicting where devastating subduction-zone earthquakes could occur requires identification
of key fault parameters from geophysical observations. Global and regional-scale
comparisons of subduction-zone seismicity against gravity and bathymetry anomalies show
that a large percentage of the moment released by co-seismic slip during megathrust
earthquakes concentrates over regions characterized by gravity and bathymetry lows. This has
been interpreted after assuming that high interplate friction implies high mechanical coupling
between the subducting slab and the overriding forearc and, therefore, a depressed forearc
topography leading to low gravity anomalies. This interpretation supports the accepted
paradigm as to that seismic asperities (areas of large co-seismic slip) are strong
patches of the plate interface characterized by high friction. Attempting to test this
idea from a physically compelling perspective, we apply a wavelet formulation
of the classical spectral isostatic analysis to invert grids of gravity anomalies and
bathymetry/topography into maps of flexural rigidity for several subduction zones
worldwide. Flexural rigidity explicitly links those surface observables with the integrated
mechanical strength of the lithosphere. For the special case of subduction zones
where two tectonic plates are in contact along a low-angle fault, it can be argued
that local- to regional-scale lateral variations of flexural rigidity are mostly due to
spatial changes of the shear strength along the interplate fault caused by the physical
conditions of the subduction channel, which supersede large-scale variations of
thermo-mechanical properties of both converging plates. In our interpretation, high/low
flexural rigidity along subduction zones means high/low shear stress supported by the
megathrust fault and, therefore, high/low friction and/or low/high pore pressure along the
subduction channel. We use the gravity model EIGEN-GL04C, which combine data from
satellite missions with marine and land gravity measurements, to calculate flexural
rigidity for most subduction zones worldwide. In addition, we compile a seismicity
catalogue for moderate to giant earthquakes (5.5 < Mw < 9.6) occurred during the last
hundred years along those subduction zones. Comparing flexural rigidity at the
epicentral location of subduction earthquakes, with their moment magnitude, released
seismic moment and the slip distribution of giant megathrust events, we confirm that
subduction-zone earthquakes tend to nucleate at regions of the interplate fault characterized
by high strength. However, a large portion of the total seismic moment accounted by
subduction earthquakes during the last hundred years was released by rupturing weak
segments of the plate interface characterized by low flexural rigidity, a tendency which
is dominated by the growing pattern characterizing the three giant earthquakes
(Mw>9) occurred during the twenty century (Kamchatka 1952, Chile 1960, Alaska
1964). The unexpected Mw9.3 Sumatra-Andaman 2004 earthquake is anomalous
with respect to other giants because it propagated throughout several high-strength
segment of the subduction zone. Paleoseismological studies in Chile and Sumatra
suggest that giant earthquakes in the later region have a recurrence interval that is 2-3
times higher than the interval between giants in the former region. This intriguing
behavior fits into a model suggested by our results, as to that the common way for
an earthquake to grow toward giant magnitudes is for the rupture front entering a
conditionally-stable, weak segment of the megathrust (the case of twenty century giants); a
less common scenario (typified by the Sumatra-Andaman 2004 earthquake) supposes the
synchronized rupture of several high strength patches of the interplate fault that are
all loaded near a critical shear stress. Our findings partially contradicts previous
interpretations based on the correlation of gravity and bathymetry anomalies with
seismogenic behaviour and can potentially change paradigms forming the physical basis to
understand processes associated with the initiation and growth of subduction earthquakes. |
|
|
|
|
|
|