Electronic Supplement to
Reassessing the Rupture Process of the 2009 L’Aquila Earthquake (Mw 6.3) on the Paganica Fault and Investigating the Possibility of Coseismic Motion on Secondary Faults

by Julien Balestra and Bertrand Delouis

This electronic supplement presents waveform fitting of the L’Aquila mainshock using different velocity models (Fig. S1) and the time versus distance distribution of slip for the single-fault (1-SF) model (Fig. S2). Figure S3 and Table S1 describe fault models corresponding to a variation in the geometry of the Paganica fault. The fault model incorporating secondary antithetic faults (model 6-SF; Table S2) and the related modeling of the Global Positioning System (GPS) and Interferometric Synthetic Aperture Radar (InSAR) data (Figs. S4 and S5) are included, as is the fault model incorporating secondary synthetic faults (model 5-SF; Fig. S6 and Table S3). Figure S7 shows the uncertainty (one standard deviation) on the slip value and rupture time, and Figure S8 shows inversions of synthetic data incorporating 15% of noise. Figure S9 presents the a posteriori modeling of the leveling data published by Cheloni et al. (2014). The quantification of the improvement of the GPS data fitting with the preferred model 2-SFPSG with respect to model 1-SF is shown in Tables S4–S6.


Figures

Figure S1. Examples of waveform fitting from separate strong-motion inversions for the mainshock carried out with different velocity models. The gray lines represent the observed waveforms. The black continuous lines represent the waveforms computed with the preferred velocity models used in this study (Bi-VM2). The black-dashed lines represent the waveforms computed with the velocity models published in Cirella et al. (2009).

Figure S2. Rupture onset time of the subfaults versus distance to the hypocenter, from the joint inversion carried out with model 1-SF. Dashed lines represent the trend for rupture velocities (Vr) of 1, 1.8, and 2.5 km/s. In this graph, rupture velocity means the average rupture velocity between the hypocenter and a given point in the fault model. White squares represent triggered subfaults with a size proportional to slip amplitude (scale on the graph).

Figure S3. Fault models (a) 1-SFc1 and (b) 1-SFc2 corresponding to variations around model 1-SF (Paganica fault), with a cross section perpendicular to fault strike on the left and a map view on the right. In the cross sections, the heavy dashed gray line with its open circle shows the fault plane of model 1-SF, with its hypocenter for reference. The two-segment black lines with the open stars (hypocenter) correspond to the new fault models tested, with their respective hypocenters. The position of GPS station PAGA is indicated at the surface. The sense of fault movement is indicated by double black arrows. On the map views, the black rectangular frames correspond to the surface projection of the two-segment fault models, with dots inside representing the individual point sources. Black squares indicate GPS stations (station PAGA is labeled).

Figure S4. Map of model 6-SF. Black rectangular frames denote the surface projections of the different fault segments, with dots corresponding to individual subfaults. Black squares indicate the locations of the GPS stations, with station PAGA labeled. Gray arrows show the dipping direction of each segment, starting from the top of the fault segments, and the white star locates the epicenter. Faults: B.F., Bazzano fault; C.C.F., Colle Caticchio fault; F.F., Fossa fault; M.F., Monte fault; and S.F., Stiffe fault.

Figure S5. Modeling of the geodetic data from the joint inversion with model 6-SF. (a) GPS vectors: observed (white arrows) and computed (black arrows). Black rectangular frames and dots inside are the same as in Figure S4. (b) Computed fringes for the Envisat descending interferogram. White rectangular frames indicate the surface projections of the different fault segments, and the white star locates the epicenter. (c) Observed Envisat descending interferogram from Atzori et al. (2009). White arrows point to observed discontinuities in the InSAR fringes. Note that they are not reproduced in the computed interferograms in (b). The white rectangular frame is the surface projection of the Paganica fault. Modeling of the additional SAR interferograms, and of the seismological data, is not shown.

Figure S6. Map of model 5-SF. Black rectangular frames outline the surface projections of the different fault segments, with dots corresponding to individual subfaults. Black squares are the GPS stations, with station PAGA labeled. Gray arrows show the dipping direction of each segment, starting from the top of the fault segments, and the white star locates the epicenter. Faults: A.F., Asini fault; C-S, the Monte Castellano–Monte Stabiata zone (Boncio et al., 2010); P.F., Proticciolo fault; and S.D.F., San Demetrio fault.

Figure S7. Uncertainty on the slip and rupture time parameters. In (a) we reproduce our preferred average model (2-SFPSG) for reference, and (b) shows the standard deviation on the slip values. In the computation of the standard deviation, slip values are weighted by 1/fcost, in which fcost is the cost function (see main article for the exact definition of the cost function). Only models having cost functions within 5% of the best solutions are retained, in the same way as when we compute the average model (a). In (c), the standard deviation on rupture time is shown. Time values are weighted by slip amplitude and by 1/fcost, using only the 5% best models as before. Weighting by slip amplitude is justified by the fact that rupture timing is unconstrained when slip is small. The hatched area corresponds to rupture times between 6 and 7 s after nucleation time (i.e., close to the maximum allowed rupture time, which is 7 s). Due to this constraint, the standard deviation on rupture time is artificially small in this area. The open star locates the hypocenter.

Figure S8. Synthetic tests with fault model 2-SFPSG, with added noise. Same as Figure 22 in the main article, but 15% of random noise was added to all the datasets. (a) Synthetic (input) slip model. (b) Slip distribution from the joint inversion carried out with the synthetic data. (c)–(f) Slip distributions from separate inversion (strong motion, teleseismic, InSAR, and GPS data, respectively), carried out with the synthetic data. The open star locates the hypocenter.

Figure S9. A posteriori modeling of the leveling data published in Cheloni et al. (2014). The slip model used is our preferred one resulting from the inversion with model 2-SFPSG (Fig. 14 in the main article, and Figure S7a). (a) Map of the leveling data of line 9 (light gray squares), lines 124–126 (dark gray squares), and line 197 (black squares) published in Cheloni et al. (2014). Black rectangles indicate the surface projections of Paganica (P.F.), San Gregorio (S.G.F.), and Campotosto (C.F.) faults. (b) Comparison between observed (white circles) and computed (black circles) leveling data of line 9. Dark gray arrows indicate the intersection with the surface trace of the Paganica (P.F.) and San Gregorio (S.G.F.) faults. (c) Comparison between observed (white circles) and computed (black circles) leveling data of lines 124–126. Dark gray arrow indicates the intersection with the surface trace of the Campotosto fault (C.F.). (d) Comparison between observed (white circles) and computed (black circles) leveling data of the line 197. Dark gray arrows indicate the intersection with the surface trace of Paganica (P.F.) and Campotosto (C.F.) faults.


Tables

Table S1. Fixed source parameters for models 1-SFc1 and 1-SFc2 (Fig. S3).

Table S2. Fixed source parameters for model 6-SF (Fig. S4).

Table S3. Fixed source parameters for model 5-SF (Fig. S6).

Table S4. Observed surface displacements at stations PAGA, INFN, and CADO.

Table S5. Computed surface displacement from model 1-SF at stations PAGA, INFN, and CADO.

Table S6. Computed surface displacements from model 2-SFPSG at stations PAGA, INFN, and CADO.


References

Atzori, S., I. Hunstad, M. Chini, S. Salvi, C. Tolomeo, C. Bignami, S. Stramondo, E. Transatti, A. Antonioli, and E. Boschi (2009). Finite fault inversion of DInSAR coseismic displacement of the 2009 L’Aquila earthquake (central Italy), Geophys. Res. Lett. 36, L15305, doi: 10.1029/2009GL039293.

Boncio, P., A. Pizzi, F. Brozzetti, G. Pomposo, G. Lavecchia, D. Di Naccio, and F. Ferrarini (2010). Coseismic ground deformation of the 6 April 2009 L’Aquila earthquake (central Italy, Mw 6.3), Geophys. Res. Lett. 37, L06308, doi: 10.1029/2010GL042807.

Cheloni, D., R. Giuliani, E. D’Anastasio, S. Atzori, R. J. Walters, L. Bonci, N. D’Agostino, M. Mattone, S. Calcaterra, P. Gambino, et al. (2014). Coseismic and post-seismic slip of the 2009 L’Aquila (central Italy) Mw 6.3 earthquake and implications for seismic potential along the Campotosto fault from joint inversion of high-precision levelling, InSAR and GPS data, Tectonophysics 622, 168–185, doi: 10.1016/j.tecto.2014.03.009.

Cirella, A., A. Piatanesi, M. Cocco, E. Tinti, L. Scognamiglio, A. Michelini, A. Lomax, and E. Boschi (2009). Rupture history of the 2009 L’Aquila (Italy) earthquake from non-linear joint inversion of strong motion and GPS data, Geophys. Res. Lett. 36, L19304, doi: 10.1029/2009GL039795.

[ Back ]