Source characterization of the December 2018 Anak Krakatau volcano sector collapse

The inversion of long-period seismic waveform data has been an effective tool for characterizing landslide sources. The extension of this method to the characterization of tsunamigenic subaerial and submarine mass failures (SMFs), which can be a significant source of hazard for coastal communities, requires consideration of the additional complexity introduced by the interaction between the sliding material and the water. Using the 22 December Anak Krakatau volcano flank collapse as an example, we show that certain a priori constraints commonly adopted for the characterization of landslide sources may not apply to SMF sources. We introduce a new source model for SMFs, which explicitly accounts for the interactions between the sliding material and the water. Model-predicted (synthetic) seismograms can be computed for any given set of model parameters. A global optimization algorithm can be developed to search for the optimal model parameters by minimizing the waveform misfits between synthetic and observed seismograms. The optimal model parameters for the Anak Krakatau volcano flank collapse are generally consistent with results obtained in previous studies. The obtained optimal model parameters can be used as inputs to tsunami simulation code, potentially leading to more accurate and timely tsunami early warning.