Ground-motion simulations of the L’Aquila earthquake Mw6.3 (April 6th, 2009) using aftershocks as empirical Green’s functions

On April 6th, 2009 the L’Aquila earthquake Mw6.3 struck the Abruzzo region (Central Italy) and was at the origin of heavy damages and human losses. This destructive mainshock was followed by numerous aftershocks recorded by a large number of accelerometric stations of the Italian Strong Motion Network (RAN). The aim of this study is to simulate the ground-motions generated by the L’Aquila mainshock Mw6.3 from accelerometric records of aftershocks. We use a simulation method based on empirical Green’s functions (EGF) principle. It consists in simulating ground-motions produced by a large earthquake by using stochastic summation of small earthquake recordings regarded as EGF. This method has the advantage of incorporating wave-propagation and site effects. To take advantage of this interesting dataset, we propose to use successively different aftershocks as empirical Green’s functions in order to reproduce the L’Aquila mainshock Mw6.3. As preliminary investigation, three aftershocks were selected: the April 6th, 2009 (02h37) Mw5.1 earthquake, the April 7th, 2009 (09h26) Mw5.0 earthquake and the April 7th, 2009 (21h34) Mw4.6 earthquake. In a first step, the input parameters have to be determined. The method used in this study presents the great advantage of necessitating only four input parameters: the corner frequency of the small event used as EGF, the seismic moments of the target event and the small event and the stress drop ratio between the large and the small events. We used the seismic moments of the mainshock and the aftershocks computed by INGV (Istituto Nazionale di Geofisica e Vulcanologia). We analyzed the displacement spectra obtained from accelerometric records to determine the corner frequencies. Afterward, the three aftershocks were successively used as empirical Green’s functions to produce 500 different synthetic accelerograms of an Mw6.3 earthquake for each station and each component. In order to test the influence of the stress drop ratio parameter on ground-motion amplitudes, we run different simulations for which the stress drop ratio between the mainshock and each of the aftershocks chosen as empirical Green’s functions is set at different values. We limited our range of stress drop ratio values by keeping only the rupture durations that are reasonable for a magnitude 6.3 event. The simulation results obtained from each aftershock taken as empirical Green’s functions are compared to determine in what extent we are able to have a good reproduction of the actual L’Aquila mainshock recordings. We also investigated how a directivity effect of the rupture process can be taken into account in our simulation method. Lastly, we compared the ground-motion simulations with the estimations given by different empirical ground-motion prediction equations.