DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación
Artículo científico (JournalArticle)
Subduction earthquakes controlled by incoming plate geometry: The 2020 M > 7.5 Shumagin, Alaska, earthquake doublet
In 2020, an earthquake doublet, a M7.8 on July 22nd and a M7.6 on October 19th, struck the Alaska-Aleutian subduction zone beneath the Shumagin Islands. This is the first documented earthquake doublet involving a megathrust event and a strike-slip event. The first event partially ruptured a seismic gap, which has not hosted large earthquakes since 1917, and the second event was unusual as it broke a trench-perpendicular fault within the incoming oceanic slab. We used an improved Bayesian geodetic inversion method to estimate the fault slip distributions of the major earthquakes using Interferometric Synthetic Aperture Radar (InSAR) wrapped phase and Global Navigation Satellite Systems (GNSS) offsets data. The geodetic inversions reveal that the Shumagin seismic gap is multi-segmented, and the M7.8 earthquake ruptured the eastern segment from 14 km down to 44 km depth. The coseismic slip occurred along a more steeply, 26◦ dipping segment, and was bounded up-dip by a bend of the megathrust interface to a shallower 8◦ dip angle connecting to the trench. The model for the M7.6 event tightly constrained the rupture depth extent to 19-39 km, within the depth range of the M7.8 coseismic rupture area. We find that the M7.6 event ruptured the incoming slab across its full seismogenic thickness, potentially reactivating subducted Kula-Resurrection seafloor-spreading ridge structures. Coulomb stress transfer models suggest that coseismic and/or postseismic slip of the M7.8 event could have triggered the M7.6 event. We conclude that the segmented megathrust structure and the location of intraslab fault structures limited the rupture dimensions of the M7.8 event and are responsible for the segmentation of the Shumagin seismic gap. Our study suggests that the western and shallower up-dip segments of the seismic gap did not fail and remain potential seismic and tsunami hazard sources. The unusual earthquake doublet provides a unique opportunity to improve our understanding of the role of the subducting lithosphere structure in the segmentation of subduction zones., This research was supported by the Natural Environmental Research Council (NERC) through the Centre for the Observation and Modelling of Earthquakes, Volcanoes and Tectonics (GA/13/M/031) and the LiCS large grant (NE/K011006/1). This research was also supported by a Chinese Scholarship Council-University of Liverpool joint scholarship awarded to YJ (201706450071). PJG contribution was supported by the Spanish Ministerio de Ciencia e Innovación research project, grant agreement number PID2019-104571RA-I00 (COMPACT) and the Beca Leonardo a Investigadores y Creadores Culturales 2020 of the Fundación BBVA, grant agreement number IN[20]_CMA_CCT_0015. RB acknowledges support by NSF award EAR-1801720. Global Centroid Moment Tensor (GCMT) Solutions are from https://www.globalcmt.org/CMTsearch.html. Earthquake information is based on the catalogs from U.S. Geological Survey (USGS) (https://earthquake.usgs.gov/earthquakes). Sentinel-1 InSAR data are copyrighted by the European Space Agency, and 6 interferograms can be downloaded from Zenodo (https://doi.org/10.5281/zenodo.5722342). The GNSS stations with three-component coseismic offset estimates for Jul 22 2020 M7.8 event and Oct 19 2020 M7.6 event are retrieved from http://geodesy.unr.edu/news_items/20200723/us7000asvb_5min_rapid_20200723.txt and ftp://data-out.unavco.org/pub/products/event/cwu_201019_2055_eq59_coseis_final.evt. The 21 GNSS stations with three-component daily offsets to estimate the postseismic decay time are retrieved from Nevada Geodetic Laboratory (http://geodesy.unr.edu/). The Subduction zone geometry model is retrieved from Slab2 (https://www.sciencebase.gov/catalog/item//). The bathymetry data is retrieved from SRTM30_PLUS (https://topex.ucsd.edu/WWW_html/srtm30_plus.html). The earthquake catalog is retrieved from USGS (https://earthquake.usgs.gov/earthquakes/search/). The coastal data is retrieved from NOAA (https://www.ngdc.noaa.gov/mgg/shorelines/). This is a contribution of the CSIC Thematic Platform PTI-Teledect (https://pti-teledetect.csic.es/). CRUST 1.0 model is retrieved from https://igppweb.ucsd.edu/~gabi/crust1.html. The authors thank the Editor Jean-Philippe Avouac, Jeffrey Freymueller, Tim Wright, and an anonymous reviewer for their insightful comments. The authors thank Lingling Ye for providing the coseismic slip model for the 2020/07/22 M7.8 earthquake as shown in Fig. 5(b). The authors also thank Lucía Pérez-Díaz for her help on the plot of Fig. 7(b)., Peer reviewed