Application of the EEPAS earthquake forecasting model to Italy

Abstract

Summary The EEPAS (Every Earthquake a Precursor According to Scale) forecasting model is a space–time point-process model based on the precursory scale increase ($\psi $) phenomenon and associated predictive scaling relations. It has been previously applied to New Zealand, California and Japan earthquakes with target magnitude thresholds varying from about 5 to 7. In all previous application, computations were done using the computer code implemented in Fortran language by the model authors. In this work we applied it to Italy using a suite of computing codes completely rewritten in Matlab. We first compared the two software codes to ensure the convergence and adequate coincidence between the estimated model parameters for a simple region capable of being analysed by both software codes. Then, using the rewritten codes, we optimised the parameters for a different and more complex polygon of analysis using the Homogenized instRUmental Seismic catalogue (HORUS) data from 1990 to 2011. We then perform a pseudo-prospective forecasting experiment of Italian earthquakes from 2012 to 2021 with Mw≥5.0 and compare the forecasting skill of EEPAS with those obtained by other time independent (SUP, SVP, and PPE) and time dependent (ETAS) forecasting models using the information gain per active cell (IGPA). The preference goes to the ETAS model for short time intervals (3 months) and to the EEPAS model for longer time intervals (6 months to 10 years).

Comparison between alarm-based and probability-based earthquake forecasting methods

“…The probabilistic models EEPAS and ETAS are fitted by Biondini et al ( 2023 ) to forecast earthquakes with magnitude M ≥ 5 . 0 in the Italian re gion, ov er a spatial grid consisting of 177 non-ov erlapped square cells of side L = 30 √ 2 km showed in Fig.…”

“…Ev en the FORE method, based on the occurrence of potential foreshocks, was adapted to this region R of analysis. According to Biondini et al ( 2023 ) such region only considers cells where a M w ≥ 4 earthquake occurred inland in the last four centuries. This is to avoid an overestimation of the performance of all forecasting methods caused by the inclusion of almost completely aseismic areas (like for example the Sardinia Island).…”

“…To implement the alarm-based approach, the analysis and optimization of the rate thresholds for the probabilistic models and of the foreshock magnitude range for the FORE method was conducted for the period 1990-2011, while a pseudo-prospective alarm-based forecasting experiment is conducted for the period 2012-2021. Such periods are the same as those used for the learning and the pseudo-prospective application, respecti vel y of the EEPAS and ETAS models by Biondini et al ( 2023 ).…”

“…In this work, the probabilistic forecasting models described in Biondini et al ( 2023 ) are compared with a deterministic forecasting method based on the occurrence of strong (fore) shocks (hereafter FORE method) described by Gasperini et al ( 2021 ). The latter sets an alarm of duration t every time an earthquake of magnitude 4 .…”

“…The probabilistic models described by Biondini et al ( 2023 ) are the every earthquake a precursor according to scale (EEPAS) model (Rhoades & Evison 2004 ) and the epidemic type aftershock sequence (ETAS) model (Ogata 1988 ). The EEPAS model is based on the increase in rate and in magnitude of the minor seismicity observed before the occurrence of major earthquakes ( ψ-phenomenon) and each earthquake is assumed to contribute to the transient increment of the rate density of future seismicity (Rhoades & Evison 2004 ).…”