A rapid response magnitude scale for timely assessment of the high frequency seismic radiation

Abstract

In this work the scaling of seismic moment (M0) and radiated energy (Er) is investigated for almost 800 earthquakes of the 2016–17 Amatrice-Norcia sequences in Italy, ranging in moment magnitude (Mw) from 2.5 to 6.5. The analysis of the M0-to-Er scaling highlights a breaking of the source self-similarity, with higher stress drops for larger events. Our results show the limitation of using M0, and in turn Mw, to capture the variability of the high frequency ground motion. Since the observed seismicity does not agree with the assumptions on stress drop in the definition of Mw, we exploit the availability of both Er and M0 to modify the definition of Mw and introduce a rapid response magnitude (Mr), which accounts for the dynamic properties of rupture. The new Mr scale allows us to improve the prediction of the earthquake shaking potential, as shown by the reduction of the between-event residuals computed for the peak ground velocity. The procedure we propose is therefore a significant step towards a quick assessment of earthquakes damage potential and timely implementation of emergency plans.

Detecting long-lasting transients of earthquake activity on a fault system by monitoring apparent stress, ground motion and clustering

“…The linear scaling between the logarithm of E S and M 0 holds over 5 orders of magnitude for seismic moment (Fig. S2), and the scaled energy (i.e., E S /M 0 ) variations with M 0 well compares to previous results obtained for several tectonic areas 38 , and in particular for Central Italy 35 (Fig. S3).…”

“…It is worth noting that changing the node constrained to zero corresponds to changes of both the A and C j , as well as for D and G j , which compensate each other, so that the coefficient B and F in Eqs (2) and (3) are unaltered. The attenuation models obtained for the Irpinia area agree with those obtained for the Central Italy seismic sequence 35 (Fig. S5).…”

“…S1). We assessed E S and M 0 by extracting from the S-wave time window proxies for these two source parameters and correcting them for attenuation along the path 34,35 (see “Method” section). So far, applications where E S is estimated for small magnitude earthquakes has been rather limited, mainly due to difficulties in its estimation from band-limited recordings 36 .…”

“…We adopted a procedure proposed for the Central Italy seismic sequence 34,35 , where M 0 and E S are derived from two proxies measured on direct S-waves: the peak displacement (PD S ) and the cumulative squared velocity (IV2 S ), respectively. Processing was carried out as follows: P-waves have been hand-picked on three-component ground velocity recordings sampled at 125 Hz, while S-waves arrival has been estimated considering location and a 1D velocity model optimized for the Irpinia area; Butterworth band pass pre-deconvolution filter with high and low pass corner frequencies at 0.5 Hz and 40 Hz, respectively; Instrumental correction and computation of acceleration, velocity and displacement.…”

“…Inspired by previous works 32,33 , we mapped the heterogeneity distribution of τ a over the Irpinia fault system. The source parameters M 0 and E S used to derive τ a are obtained by applying the same approach used for the 2016–2017 Central Italy seismic 34,35 . Then, focusing on the more densely instrumented central sector of the Irpinia fault (i.e., a segment 60 km long), where the Mw 6.9, 1980 earthquake occurred, we studied the temporal evolution of τ a , δBe and η considering small and moderate magnitude earthquakes and looking for trends which are coherent among the parameters.…”