Developing an earthquake model based on simultaneous peak ground acceleration occurrences using the D-vine copula approach

Indonesia is susceptible to natural disasters including earthquakes, volcanic eruptions, tsunamis, floods, and landslides. This catastrophe wreaked havoc on infrastructure, residences, and businesses, resulting in enormous economic losses. One of the frequent natural catastrophes in Indonesia is an earthquake, particularly in the province of Banten, one of potential areas exposed to megathrust earthquake. Peak ground acceleration (PGA) can be used to measure earthquake risk, but current calculations are univariate, meaning that seismic hazard calculations are performed independently across regions. In reality, seismic conditions in a region are influenced by seismic conditions in neighboring regions, making the independent calculation of PGA less pertinent. In this article, we propose to construct a model for earthquakes based on PGA values by incorporating the dependencies among PGA occurrences via the D-vine copula method. We discovered that the greater the distance between a quake-affected location and the epicenter, the greater the influence of ground motion from nearby locations. These findings can be used as a tool to mitigate earthquake occurrences in Indonesia, a similar strategy can also be implemented in other regions.