Integrating Time-Dependent and Time-Independent Earthquake Recurrence Models for Fault-Based Seismic Hazard Mapping in Indonesia

Abstract

Indonesia's location at the convergence of three major tectonic plates causes extreme seismic activity, necessitating precise hazard assessments. This research integrates time-dependent (Brownian) and time-independent (Poisson) recurrence models across faults in Indonesia to evaluate rupture probabilities and their implications for probabilistic seismic hazard assessment (PSHA). The Poisson model assumes constant-rate events, whereas the BPT approach estimates rupture probability from tectonic stress accumulation, recurrence intervals, and elapsed time since the last major rupture. Earthquake catalogs from ISC, USGS and BMKG were compiled for shallow events (≤50 km) between 1900 and 2025, with fault parameters derived from national maps. Rupture probabilities and peak ground acceleration (PGA) were estimated using OpenQuake for 10% and 2% exceedance probabilities in 50 years. Results show the BPT model is highly sensitive in identifying late-cycle faults, revealing the highest probabilities (>90%) for the Aceh-South and Batee A segments. Early-cycle faults like the post-2018 Palukoro exhibit significantly reduced probabilities (3.75%). The Poisson model generated a broader hazard distribution, yielding PGA exceeding 1.35 g on Northern Sumatera faults. Integrating these complementary models substantially improves seismic hazard characterization for national scale PSHA in Indonesia.