Imaging of small-scale structures and time-dependent changes in the rupture area of the 2010 Maule earthquake
Final Report Abstract
The objective of the project had been to image structures near the plate interface and within the subduction channel, and correlate them with frictional properties deduced from geodetic observations, local seismicity and the mainshock rupture model. In the original work plan, we had proposed two methods to approach this objective. First, we were going to determine the local VP/VS ratio within clusters of earthquakes using a double-difference method and high-precision cross-correlation times. Secondly, we were going to analyse secondary phases from local earthquakes in an attempt to image the plate interface. However, the scientific developments led to a somewhat different course. After the local seismicity distribution was published by Lange et al. [2012] and others, offering a far better insight into the geometry of the plate interface, it became clear that the rupture models or backprojection results published at the time might be biased with respect to the location of local seismicity (or any structures found by application of the proposed techniques), hampering detailed comparison and interpretation. The reason is that the classic rupture backtracking is calibrated with the hypocentre, and therefore might show a severe bias with respect to structures and seismicity derived from local observations. We therefore decided it was necessary to modify the backprojection technique by calibrating the procedure based on local aftershock locations and also allow for spatial variability of the station statics; an additional motivation to pursue this line was prior experience of the project staff (Mauro Palo) in using array techniques. This work has led to a more detailed image of the high frequency emissions during the main earthquake phase. In particular the correct positioning with respect to the local seismicity has demonstrated the correspondence between the emission points for high frequency seismic radiation and the location of aftershocks to the point that the unusual two separate activity peaks seen in the plate interface aftershock seismicity on the downdip end of the plate interface are also reflected in the backprojection results. This correspondence suggests strongly that the frictional heterogeneity leads to both a rough rupture in the main shock (as evidenced by high frequency emissions), and incomplete strain release, resulting in significant afterslip activity. The algorithm developed for the 2010 Maule earthquake was also used subsequently in the analysis of the 2014 Iquique (MW = 8.2) and 2015 Illapel (MW = 8.3) earthquakes in Northern and Central Chile, respectively. We then proceeded to try and use double difference cross-correlation times between P- and S waves to determine directly VP/VS ratios using the method of Lin and Shearer [2007], as planned in the proposal work programme. However, after extensive tests with real and synthetic data, it became clear that the method does not produce robust results in most settings, even apparently favourable ones. Additional theoretical work clarified the fundamental reason, i.e. that bias due to deviations between P and S emergence angles does not cancel even when averaging over well-formed spherical clusters. Although this is a negative result as far as the main objectives of the proposal are concerned, we thought it important to invest the time to write up this methodological result in order to prevent others from investing time in the application of this method without a thorough consideration whether it will be applicable. For the final part of the work package, which was focussed on imaging interfaces using secondary phases, there was only very little time left in the contract for the project staff. Therefore only very preliminary observations could be made. These are promising, though, such that we will pursue that line of investigation on our own time.
Publications
- Gradual unlocking of plate boundary controlled initiation of the 2014 Iquique earthquake, Nature, 512, 299–302, 2014
Schurr, B., G. Asch, S. Hainzl, J. Bedford, A. Hoechner, M. Palo, R. Wang, M. Moreno, M. Bartsch, Y. Zhang, O. Oncken, F. Tilmann, T. Dahm, P. Victor, S. Barrientos, and J.-P. Vilotte
(See online at https://doi.org/10.1038/nature13681) - High-frequency seismic radiation from Maule earthquake (MW 8.8, 2010 February 27) inferred from high-resolution backprojection analysis, Geophys. J. Int., 199, 1058–1077, 2014
Palo, M., F. Tilmann, F. Krüger, L. Ehlert, and D. Lange
(See online at https://doi.org/10.1093/gji/ggu311) - Applicability and bias of VP /VS estimates by P and S differential arrival times of spatially clusterered earthquakes, Bul. Seism. Soc. Am., 106, 1055–1063, 2016
Palo, M., F. Tilmann, and B. Schurr
(See online at https://doi.org/10.1785/0120150300) - The 2015 Illapel earthquake, central Chile: A type case for a characteristic earthquake?, Geophys. Res. Let., 43, 574–583, 2016
Tilmann, F., Y. Zhang, M. Moreno, J. Saul, F. Eckelmann, M. Palo, Z. Deng, A. Babeyko, K. Chen, J. C. Baez, B. Schurr, R. Wang, and T. Dahm
(See online at https://doi.org/10.1002/2015GL066963)
