S U M M A R YMoment tensor inversion of early XX century mechanical seismograph recordings offers the opportunity to estimate source parameters for a number of important earthquakes. This knowledge is crucial for understanding regional tectonics and estimating seismic hazard; but processing of the historical seismograms is often difficult. Especially, the rotation of horizontal seismograms into radial (P-SV ) and transverse (SH) components tends to fail; for example, if one polarity is wrong, if the alignment of the seismograms is not correct, or if one component is lost. To avoid these kind of problems, we use a moment-tensor-inversion scheme that processes the original single-component horizontal traces without rotation, and rotates the theoretical Green's functions instead. We show how this approach simplifies the analysis, and allows a better use of available data. We present moment tensor solutions for two destructive earthquakes in southwestern Europe in 1909. The April 23 earthquake near Benavente in the lower Tagus valley (Portugal) had a moment magnitude M W = 6.0 and an estimated centroid depth of 10 km. Our preferred moment tensor solution indicates reverse faulting (nodal planes have strike/dip/rake of N51 • E/52 • /83 • and N242 • E/38 • /99 • ). We propose a blind thrust beneath the Tagus valley sediment basin as the responsible fault. The June 11 earthquake near Lambesc, Provence (France) was found to be slightly smaller (M W = 5.5) and related to oblique reverse faulting at 4-km depth. Nodal planes of the preferred solution have strike/dip/rake of N80 • E/53 • /53 • and N311 • E/50 • /128 • , subparallel to the Trévaresse fold.

No abstract

SUMMARY We examine and model analogue recordings from 6 early mechanical seismographs for the 1910 June 16 earthquake at Adra, Southern Spain. Modern standard, time‐domain analysis techniques were applied to the historical data to estimate the source parameters of the event: The regional sparse network data were inverted for the deviatoric seismic moment tensor. The best moment tensor solution corresponds to a M0= 1.50 · 1018 Nm, MW 6.1 oblique strike‐slip event at 16 km depth. Our preferred faulting solution is: strike 122°, dip 80°, rake −137°, in very good agreement with available neo‐ and seismotectonic data. The source time function of this earthquake was estimated by deconvolving recordings of a MW 5.5 aftershock that occurred the same day. The time function indicates a total rupture time of 4.5 s, corresponding to estimates for mainshock rupture length of 12 km, and stress drop of 29 bar.

No abstract

No abstract