Coseismic Slip Model Evaluation and 3-month Stress Driven Afterslip Simulation of the July 29, 2021 Mw 8.2 Chignik Earthquake.

On 29 July 2021, an Mw 8.2 megathrust earthquake struck the Alaska Peninsula, which was the largest earthquake in the U.S. since 1964. Quantifying the exact coseismic slip and the following afterslip distribution of this earthquake provides us the opportunity to evaluate the slip budget and the earthquake hazard potential in the Alaska-Aleutian subduction zone. In this study, we first use several published coseismic finite rupture models as input to simulate the postseismic deformation over the first three months after the earthquake, in which time period afterslip is likely the dominant effect. We use a rate-and-state frictional afterslip model to do the forward modeling and find that both the updip and downdip afterslip are necessary to fit the observations. However, we find that most coseismic models can't explain the large afterslip signal observed on the Chirikof island, and also produce large angular misfits at several sites on the Alaska Peninsula. The spatial pattern of the afterslip is mainly controlled by the coseismic slip distribution, while the slip time history is controlled by the frictional parameters. The afterslip spatial pattern thus provides new information about the coseismic slip distribution.

The coseismic slip inversion result hugely depends on the smoothing factor or the chosen roughness of the rupture plane, because there is limited information to constrain offshore slip. Therefore, we use the afterslip spatial patterns as a new data constraint to modify the coseismic slip, to get a coseismic rupture model that can not only fit the coseismic data constrains but also the observed postseismic signal well. We also add to the inversion displacements from GPS campaign measurements on some islands near the rupture zone, which contain the coseismic displacements and 3 weeks of afterslip. We remove the predicted afterslip at those sites and use the remaining coseismic signal to do the coseismic slip inversion, then we do multiple iterations to finally get a stable coseismic rupture model.