Filling in the gaps of the tsunamigenic sources in 2018 Palu Bay tsunami.

The causes of 2018 Palu bay (Indonesia) tsunami are still not entirely clear. There is still an ongoing debate on whether the main cause of the tsunami waves observed was a significant co-seismic tectonic event which occurred underwater or whether it were the multiple landslides detected along the coast and triggered by the earthquake. Data from the paper by Liu et al. (2020) in which the bathymetry of the bay was analysed suggests that landslide-induced waves may have contributed significantly to the tsunami. However, the data presented was incomplete and the information regarding the starting time, magnitude of these waves and the coastal landslide progression has significant uncertainties. In this paper we model each landslide-generated wave with the COMCOT model and track the propagation of the waves to understand their individual contribution at several relevant locations inside the bay where free surface elevation data is available. We then explore the feasible scenarios (i.e. landslide-generated wave configurations and timings) that produce tsunami waves as close as those observed using an optimization technique based on genetic algorithms. Numerical simulations of the chosen scenario point out that landslide-generated waves are likely the main contributors to the tsunami, as they can arrive very fast, at the precise timing, to the locations of interest and can trigger the natural resonant modes of the bay, producing long-period waves that were also observed.