11:45 AM - 12:00 PM
[S05-2-06] ANALYSIS OF THE ANALOG SEISMOGRAMS RECORDED DURING THE NOVEMBER 19, 1912 (M~7.0) ACAMBAY, CENTRAL MEXICO EARTHQUAKE: TOWARDS A FINITE SOURCE INVERSION
The Acambay earthquake of November 19, 1912, was an M~7.0 shallow intraplate rupture that broke a large portion of the Acambay-Tixmadeje Fault System in central Mexico. The earthquake produced extensive damage in several towns close to the fault and was strongly felt in Mexico City. This event generated a 40.0-km-long surface rupture trace, with a peak observed dislocation of ~0.5m. To date this is one of the most important continental earthquakes occurred in Mexico. During 1912 six seismographic stations were operating in Mexico with different Wiechert and Bosch-Omori mechanical seismographs. An extraordinarily detailed report shows it was recorded by most of horizontal and vertical instruments at the six stations. Up to 2016, only 11 from the original 22 smoked-paper seismograms, from five of the stations that recorded the earthquake were found to be recovered.
In this work using standard and ad-hoc methodologies we obtained the digital, evenly-sampled time series for the ground motion velocities for each of the available seismograms. To do this, we first carried out a high resolution scanning of the original traces. Then, due to the condition of the seismograms we manually digitized them, following the seismic trace in ascending time. Therefore we corrected the distortions of seismic traces for their pen curvature, uneven paper speed, skews, and non-zero baselines following the Grabrovec and Allegretti (1994) procedure, and additional ad-hoc corrections were made using linear approximations when abrupt changes in traces occurred. After this, we resampled the unevenly digitized traces to a constant-time interval to 100 samples per second using a spline interpolation method. Finally we performed an instrument correction procedure to obtain the actual ground velocities for each instrument and component. Once corrected, the unclipped portions of seismic data were used to perform a linear finite-source inversion for the slip distribution on the fault surface.
In this work using standard and ad-hoc methodologies we obtained the digital, evenly-sampled time series for the ground motion velocities for each of the available seismograms. To do this, we first carried out a high resolution scanning of the original traces. Then, due to the condition of the seismograms we manually digitized them, following the seismic trace in ascending time. Therefore we corrected the distortions of seismic traces for their pen curvature, uneven paper speed, skews, and non-zero baselines following the Grabrovec and Allegretti (1994) procedure, and additional ad-hoc corrections were made using linear approximations when abrupt changes in traces occurred. After this, we resampled the unevenly digitized traces to a constant-time interval to 100 samples per second using a spline interpolation method. Finally we performed an instrument correction procedure to obtain the actual ground velocities for each instrument and component. Once corrected, the unclipped portions of seismic data were used to perform a linear finite-source inversion for the slip distribution on the fault surface.