GROUND MOTION IN MEXICO CITY DURING THE INTRASLAB EARTHQUAKE OF 19 SEPTEMBER 2017 (MW7.1) REVISITED
The intraslab earthquake of 2017 (Mw7.1) was one of the most destructive earthquakes in the history of Mexico City, second only to the Michoacán interplate earthquake of 1985 (Mw8.0). Several measures of the ground motion reveal that the 2017 event was unusually energetic at CU (the reference hill-zone site) in the critical frequency range 0.4 - 1 Hz. Fourier Acceleration Spectrum at CU was anomalously high in this frequency range. With respect to site specific GMPE for intraslab Mexican earthquakes at CU, the observed PGA was normal but PGV was much greater. Observed Sa, 5% damping, at CU was anomalously high as compared to the predicted Sa at structural periods of 1 ≤ T ≤ 1.8 s (0.55 ≤ f ≤ 1 Hz). The cause of the large ground motion at CU, remains obscure: Were the short distance and large magnitude of the earthquake the main causes of destruction in the city? Was the source unusually energetic in the critical frequency range for Mexico City? Did the rupture involve a directivity towards Mexico City, thus enhancing the ground motion? Was the damage a consequence of an exceptionally unfavorable location of the source with respect to the 3-D structure of the Valley of Mexico? These issues merit a careful analysis of all available data.
On 7 December 2023 an intraslab Mw5.8 earthquake occurred in the proximity of the 2017 event. We analyze the recordings of the 2017 earthquake separately, and of the 2017 and 2023 events together to isolate the cause of the anomalous high-frequency radiation. We take recourse of the 2023 recordings as empirical Green´s functions (EGFs). Synthesized Sa for a Mw7.1 earthquake, using 2023 (Mw5.8) recordings as EGFs and assuming the same stress drop, sigma=3 MPa, for both events, are significantly lower than those observed, irrespective of the azimuth, which supports anomalous nature of the 2017 source. It follows that had the source of the 2017 earthquake had been a scaled-up version of the 2023 event then, quite likely, Mexico City would have been spared the damage and deaths that it suffered. The possibility that the enhanced ground motions in the city in 2017 were due a particular direction of the wavefield incident on the 3-D structure of the Valley of Mexico may be ruled out since the recordings of the 2023 event, which was collocated with the 2017 earthquake, show nothing anomalous.
We conclude that the damage to Mexico City during the 2017 earthquake was due to an anomalously energetic ground motion in the Valley of Mexico in the frequency band of 0.4 to 1 Hz. The role of rupture directivity in enhancing the ground motion appears small; the source was unusually energetic at all azimuths. The simulated response spectra using recordings of the 2023 event as EGFs show that a postulated intraslab Mw 7.1 earthquake, at a distance of ~ 130 km from CU, should cause little or no damage in the city.