Hazard Modeling of Sulawesi-Maluku Earthquakes Using Space-Time ETAS: Comparing Long Decay and Short Decay Densities

The space-time ETAS (Epidemic-Type Aftershock Sequence) model is a point process-based framework for earthquake modeling that leverages complete seismic catalogs to predict earthquake occurrences by capturing both spatial and temporal dependencies. It incorporates a long decay spatial density function, which models the gradual decrease in aftershock probability with increasing distance from the mainshocks. However, the Sulawesi-Maluku region, characterized by complex tectonic interactions and active subduction zones such as the Halmahera and Sangihe arcs, experiences frequent shallow earthquake swarms, exhibit a higher degree of clustering centered around the mainshocks, which may suggest an alternative of spatial density function within the space-time ETAS model. This study compares the commonly used long decay and short decay spatial density in the space-time ETAS model to fit the earthquake catalog in the Sulawesi-Maluku based on model complexity, computational time, and intensity prediction. Findings reveal that the short-decay function outperforms the long-decay counterpart, efficiently identifying high-seismicity areas in the Halmahera and Sangihe arcs, the Palu Koro-Matano fault zone, and the Seram Sea. This study advances seismic risk assessment and supports disaster mitigation efforts in Sulawesi-Maluku.