Understanding the ground deformation - Darfield and Christchurch - using GPS and InSAR

Abstract

During a large earthquake there is typically significant movement of the earth’s surface; however, in many cases there is additional movement following the earthquake (postseismic deformation). This movement can be due to continued slip (afterslip), relaxation of stresses with the earth following the earthquake (viscoelastic relaxation), movement of fluids with the earth (poroelastic response), or a change in volume related to recovery of strength within damaged material following an earthquake (dilatancy recovery). In this project we attempt to determine the mechanisms contributing to postseismic deformation following several earthquakes in the South Island of New Zealand. The earthquakes considered include the Mw 7.8 Dusky Sound earthquake (15 July, 2009), the Mw 7.1 Darfield earthquake (4 September, 2010), and the Mw 6.2 Christchurch earthquake (22 February, 2011). We are also interested in determining whether time-varying stresses following the Dusky Sound earthquake could have altered the stress state in Canterbury in such a way to trigger the Darfield earthquake.

Surface displacements following an earthquake are observed using both Global Positioning System (GPS) and Interferometric Synthetic Aperture Radar (InSAR) observations. The GPS observations consist of either continuously recording sites (cGPS) or sites that are measured periodically to determine the movement of the ground surface during that period. The InSAR observations are satellite radar images that tell us the movement of the ground surface toward or away from the satellite between successive passes of the satellite. Using these observations, we then evaluated the contributions to surface deformation from each of the mechanisms mentioned previously. For the Dusky Sound earthquake, the observations can be explained as a combination of afterslip and viscoelastic relaxation. For the Darfield earthquake, we find that the surface displacement following the earthquake can be explained as a combination of poroelastic response and dilatancy recovery, with little contribution from afterslip. For the Christchurch earthquake, the surface displacement can be accounted for using a combination of afterslip, dilatancy recovery, and poroelastic rebound.

Once we had determined the mechanisms responsible for surface displacement following the Dusky Sound earthquake, we then computed the stress changes at the time and location of the Darfield earthquake. Although we find a slight stress increase on the Greendale fault (the primary fault involved in the Darfield earthquake), it does not appear large enough to have triggered the Darfield earthquake without some additional contribution from another source.

Another outcome of this project was the development of a new technique that allows us to simultaneously evaluate the contributions due to afterslip and viscoelastic relaxation. Previous methods required assumptions and approximations to evaluate these effects. The new method should thus be quite useful in future studies of postseismic deformation following large earthquakes.

---

Technical Abstract

This project examines postseismic deformation following several earthquakes in the South Island of New Zealand. The earthquakes considered include the Mw 7.8 Dusky Sound earthquake (15 July, 2009), the Mw 7.1 Darfield earthquake (4 September, 2010), and the Mw 6.2 Christchurch earthquake (22 February, 2011). Of particular interest are the possible mechanisms involved in postseismic deformation. The mechanisms considered are afterslip, poroelastic response, viscoelastic response, and dilatancy recovery. We are also interested in determining whether time-varying stresses following the Dusky Sound earthquake could have altered the stress state in Canterbury in such a way to trigger the Darfield earthquake.

We made use of both InSAR and GPS observations to examine the deformation field following these earthquakes.  Using an inversion method based on the Principal Component Analysis-based Inversion Method (PCAIM), we resolved the deformation following the Dusky Sound earthquake into components due to afterslip and the viscoelastic response to coseismic slip and afterslip. Using our preferred afterslip/rheological model, we then projected the total stress change at the time of the Darfield event onto both the Charing Cross fault (thought to be the initial rupture surface) and the Greendale fault (main rupture surface). The predicted stress changes from our model would reduce the Coulomb failure stress on the Charing Cross fault and increase the Coulomb failure stress on the Greendale fault; however, in both cases the predicted stress changes are only on the order of a few kPa, and are therefore unlikely to promote or retard earthquake occurrence on their own.

For the Darfield earthquake, we find that the observed postseismic deformation can be explained as a combination of poroelastic response and dilatancy recovery, with little contribution from postseismic afterslip.  For the Christchurch earthquake, the observed deformation can be accounted for using a combination of afterslip, dilatancy recovery, and poroelastic rebound.

As an outgrowth of our work on the Dusky Sound earthquake, we have also developed a new inversion methodology:  Iterative Decoupling of Afterslip and Viscoelastic relaxation (IDAV). In contrast to the iterative approach used in our Dusky Sound inversion, which used an iterative approach to account for the fact that the Green’s functions were all elastic, the IDAV approach allows the direct use of viscoelastic Green’s functions in an inversion.  Due to the sparsity of data for the Dusky Sound event and the computational expense of generating viscoelastic Green’s functions, this method was not used in our analysis; however, this method should be quite useful for performing coupled inversions in regions with good data coverage.

References for publications accepted as final report:

John Beavan, Mahdi Motagh, 2012 - Fault slip models of the 2010-2011 Canterbury, New Zealand, earthquakes from geodetic data, and observations of post-seismic ground deformation-NZ Journal of Geology and Geophysics, v.55, p 207-221 - doi.org/10.1080/00288306.2012.697472

M Motah, J Beavan, E Fielding, M Haghshenas, 2014 - Postseismic ground deformation following September 2010 2010 Darfield, New Zealand, earthquake from TerraSAR-X, COSMO-SjkyMed and ALOS data - Geoscience and Remote Sensing Letters, GRSL-00607-2012 - doi: 10.1109/lgrs.2013-2251858