In Search of Hidden Subsurface Sources of Deformation and Earthquakes: A Catalogue of High Resolution Stress Rates of Surface Expressions
Abstract
Understanding subsurface hazard sources - buried faults, slow slipping faults, locked faults, and deep magma reservoirs - as fully as possible is of paramount importance to hazard mitigation; however, such sources are difficult to detect due to a lack of clear surface expressions. Analysis of GPS velocities has potential to uncover previously unknown sources of deformation that are non-dormant. Currently, such analysis is done in three ways: (1) by fitting prescribed sources, modeled with few free parameters, directly to GPS velocities; (2) by fitting rigid-block models with elastically-behaving zones at prescribed major bounding faults directly to GPS velocities; and (3) by converting horizontal GPS velocities into horizontal strain rates. All of these methods have severe shortcomings. The first two methods are limited by both the data they fit and by the sources they model since surface horizontal velocities are broad-wavelength responses to the subsurface sources, and hence provide low-resolution information, while tight prescription of the nature of the sources or faults allows for little to be learnt about any other processes and sources in the same region. The third method addresses the resolution issue by calculating the strain rates from the horizontal GPS velocities; however, a fundamental issue with this methodology is that it is open to misinterpretation because the strain rates are calculated solely using kinematic and not dynamic principles, as well as using purely mathematical, potentially non-physical, optimization. We propose that the vertical derivatives of horizontal stress (VDoHS) rates, being inherently much more localized, are substantially higher resolution surface expressions of the subsurface sources than are the GPS velocities or the GPS-derived strain rates. Our first area of study calculating the VDoHS rates will be New Zealand, where there is a comprehensive GPS dataset as well as diverse and interesting tectonics. In order to interpret VDoHS rates in this and future studies, we present an extensive catalogue of surface distributions of the VDoHS rates for a complete range of generic subsurface sources: point subsurface sources and shear and tensile cracks. We will show that the lateral extent of lobes in the VDoHS rates provides quite precise estimates of the depths of the sources. For the simplest point sources the primary, large-amplitude lobes extend to surface radii of order twice the source depth, and the patterns are source-depth invariant when all distances are normalised with respect to source depth. These calculations are performed using the routines for uniform half-spaces [Okada 1992] that are used in conventional interpretation of GPS observations. Given that the Earth is not a uniform half-space, further straightforward calculations demonstrate the largely-second-order sensitivity of VDoHS rates patterns at the surface to layered crustal structures. A weak surface layer, for instance, makes the subsurface source appear deeper than it really is. The VDoHS rates are physically the most spatially compact surface manifestations of subsurface sources, and the catalogue we will present is an essential tool for the interpretation of VDoHS rates into type, size and location of subsurface deformation and hazard sources. Okada, Y., 1992. Bull. Seismol. Soc. Am., 82.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2012
- Bibcode:
- 2012AGUFM.G21B..08D
- Keywords:
-
- 1209 GEODESY AND GRAVITY / Tectonic deformation;
- 8164 TECTONOPHYSICS / Stresses: crust and lithosphere;
- 4314 NATURAL HAZARDS / Mathematical and computer modeling;
- 4319 NATURAL HAZARDS / Spatial modeling