Abstract
The MW 8.3 (GCMT) Illapel megathrust earthquake is investigated. The objective is to find out which features of the previously published rupture scenarios can be resolved by using a regional strong-motion network (epicentral distances 130–260 km) and source models with a few parameters only. Low-frequency waveforms (<0.05 Hz) at nine stations (Centro Sismológico Nacional, Chile) are subjected to modeling. Various representations of the source are used: (i) multiple-point source models based either on iterative deconvolution or simultaneous inversion of source pairs, (ii) models of circular and elliptical uniform-slip patches, employing synthetic and empirical Green's functions, respectively. This variety of methods provides consistent results. The earthquake appears to be a segmented rupture progressing from an early (deep) moment release to a later (shallow) one, towards the northwest. The source models of slip-uniform patches synchronously suggest a low rupture speed of 1–2 km/s. Despite the different data sets and methods used in this study, the estimate of rupture speed is consistent with independent publications. As for ambiguity in literature regarding the depth and timing of the rupture, our paper clearly prefers the models including a ∼20–30 s delay of the shallow moment release compared to the initial deep one. The strong-motion data set and low-parametric models proved to be competitive with more sophisticated approaches like multi-parameter slip models using a variety of regional geophysical observables. These results, together with the results from other studies for smaller events, show that strong-motion networks can be useful for studying rupture processes in a wide range of magnitudes, thus promoting the improvement of regional strong-motion networks in poorly instrumented regions.
Highlights
An Mw8.3 (Illapel) earthquake occurred on September 16, 2015 in central Chile, causing significant strong ground motions, most likely due to a combined effect of rupturing a deep (~30 km) and shallow (~15 km) slip patch, respectively, indicating an along-dip segmentation of the megathrust (Melgar et al, 2016)
The strong-motion data set and low-parametric models proved to be competitive with more sophisticated approaches. This result implies a need to improve regional accelerometer networks in South America, which might eventually help in resolving source process of possible future large events, e.g. in Patagonia
The objective of this work is to develop low-parametric models of the Illapel earthquake, helping to answer specific questions like this: (i) Was most of the seismic moment released in a single patch, situated close to surface, northwest of the epicenter, or in a deeper patch situated closer to the epicenter? (ii) If two patches represent a correct source model, what was the timing of the two moment-release episodes? In order to accomplish these goals we choose a low-parametric modeling (ISOLA multiple-point source (MPS) modeling) as a suitable tool for the Illapel earthquake
Summary
An Mw8.3 (Illapel) earthquake occurred on September 16, 2015 in central Chile, causing significant strong ground motions, most likely due to a combined effect of rupturing a deep (~30 km) and shallow (~15 km) slip patch, respectively, indicating an along-dip segmentation of the megathrust (Melgar et al, 2016). The objective of this work is to develop low-parametric models of the Illapel earthquake, helping to answer specific questions like this: (i) Was most of the seismic moment released in a single patch, situated close to surface, northwest of the epicenter, or in a deeper patch situated closer to the epicenter ( closer to the locus of the high-frequency radiation identified by the backprojection techniques)? (ii) If two patches represent a correct source model, what was the timing of the two moment-release episodes? In order to accomplish these goals we choose a low-parametric modeling (ISOLA multiple-point source (MPS) modeling) as a suitable tool for the Illapel earthquake. At the same time, being interested in a robust model of the source complexity, we use the lowest frequencies enabling deterministic modeling. To further validate our MPS results we apply the empirical Green’s function technique for calculating apparent source functions and inverting them into finite source models composed of slip-uniform elliptical patches
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
