Regional earthquake likelihood models (RELMs) for probabilistic seismic hazard analysis (PSHA) around the southern off-coast of Sumatra-West Java

This study investigates the application of time-independent models for the regional earthquake likelihood models earthquake forecast testing based on the modified probability gain (mG) for seismic hazard study and analysis (SHA) in the southern off-coast of Sumatra-West Java utilizing a 15 years region time length. This study used a declustered catalog of shallow earthquakes, dividing it into two observation periods: 1978–1992 and 1993–June 2006. Two rate models were developed to estimate earthquake occurrence probability for a given magnitude and time window. The normalized probability gain between these two observation periods was then calculated, leading to the mG estimation. The mG, in turn, provides insights into seismicity patterns for the SHA. The SHA model was constructed using a uniform background seismicity rate model based on the declustered earthquake with Mw ≥ 5.0, weighted by the mG. The peak ground acceleration (PGA) versus exceedance probability over 50 years was estimated and evaluated for several sites. They are Bengkulu, Liwa, and Kota Agung. They incorporated an amplification factor derived from the horizontal-vertical Spectral Ratio method, which allows for the estimation of surface PGA values. This research suggested a better understanding of future seismic hazards and informs mitigation efforts before large earthquakes.