Sensitivity of Tsunami Data to the Up-Dip Extent of the July 2021 Mw 8.2 Alaska Earthquake

Abstract

A large tsunamigenic earthquake of magnitude Mw 8.2 occurred on the Alaska-Aleutian subduction zone in July 2021. To reveal the characteristics of the event, we first applied spectral and wavelet analyses to the induced tsunami recorded both at the local and Pacific-wide sea level observation networks. Because the earthquake was relatively deep (∼30 km), the resultant maximum tsunami amplitudes were only ∼5 and ∼50 cm in the open ocean and coastal area respectively. However, owing to the unique geological feature of the region, the tsunami had dominant periods of 57–73 min, which are longer than that typically generated by similar-size megathrust earthquakes. Furthermore, we compared multiple source models inferred from various data sets and evaluated their performances in reconstructing the observed tsunami waveforms. The comparison results suggest that the up-dip limit of the rupture area must be restricted at depth of ∼20 km to accurately reproduce the observed tsunami waveforms. Shallower slips beyond the prescribed limit led to an overestimation of the tsunami amplitude. This implies that the earthquake was unlikely to rupture the plate interface on the near trench section.

Rupture Model for the 29 July 2021 MW 8.2 Chignik, Alaska Earthquake Constrained by Seismic, Geodetic, and Tsunami Observations

“…Due to having further slip that extends up‐dip to near the continental shelf break, the computed tsunamis from the Liu et al. (2022) model are too early for all DART stations and overpredict amplitudes (Figure S3 in Supporting Information S1; Mulia et al., 2022).…”

Rapid shallow megathrust afterslip from the 2021 M8.2 Chignik, Alaska earthquake revealed by seafloor geodesy

“…The similar distributions of down-and up-dip moment are an important constraint that follows, in part, from our conservative use of subfaults that reflect average slip over relatively larger areas. Even if there were strongly variable co-seismic slip at depth, producing greater peak slip values over localized patches (35,36,40,41), our results show that on average deeper and shallower moments are equilibrated.…”

“…Plate convergence rate at the Semidi section is ~6.3 cm/year, and the slip deficit rate (the amount of plate boundary slip stored elastically in the upper plate) varies west to east from 33 to 100% (39). The Chignik event nucleated at ~26 km depth and ruptured up-dip toward the northeast (35,36,40,41) and the inferred rupture zone of the 1938 tsunamigenic M w 8.2 to 8.3 earthquake (36,42). Previous analyses suggested that Chignik co-seismic slip was confined below the intermediate depth portion of the megathrust below the continental shelf (~20 to 25 km), consistent with the lack of a large associated tsunami (36,(40)(41)(42).…”

“…The Chignik event nucleated at ~26 km depth and ruptured up-dip toward the northeast (35,36,40,41) and the inferred rupture zone of the 1938 tsunamigenic M w 8.2 to 8.3 earthquake (36,42). Previous analyses suggested that Chignik co-seismic slip was confined below the intermediate depth portion of the megathrust below the continental shelf (~20 to 25 km), consistent with the lack of a large associated tsunami (36,(40)(41)(42). Liu et al (36) assert that the megathrust immediately updip of the principal Chignik rupture patch retains a substantial slip deficit (their peak slip was ~8.2 m) that could be recuperated in a future tsunamigenic earthquake or via shallow afterslip.…”

“…See fig.S1for GNSS station codes. Orange vectors, Chignik co-seismic displacement; red vectors, Chignik cumulative co-and post-seismic displacements; pink vector, predicted displacement at GNSS-A station SEM1 from the co-seismic model(40); white arrows, along-strike rupture extents of substantial recent trench earthquakes; black arrows, boundaries of trench sections.…”