Location and magnitude from seismic intensity data of historical earthquakes in north-eastern Algeria

Abstract

Accurately determining the location and magnitude of historical earthquakes is crucial for compiling a homogenized seismic catalogue, a key component of regional seismic hazard assessment. This study aims to test and validate the adapted Bakun and Wentworth (Bull Seismol Soc Am 87: 1502–1521, 1997) method on a test-set of 10 historical earthquakes in Algeria (1856–1946) with the aim of revising and updating the Earthquake Catalogue for Algeria (ECA). We calibrated the method using 1083 intensity observations from 14 training-set earthquakes (1949–2003). This approach estimates intensity magnitudes based on an attenuation model and determines the epicentral region by minimizing residuals over a predefined grid. To validate our results, we also tested the Boxer 4.2.1 software (Gasperini et al., Bull Seismol Soc Am 100: 2035–2066, 2010), based on Pasolini et al. (Bull Seismol Soc Am 98: 692–708, 2008) methods, which determine the intensity epicentre by maximizing the likelihood function of an attenuation equation and derives magnitude from an average of multiple estimates. While both methods generally produce comparable results in most cases, discrepancies arise for earthquakes with sparse historical intensity data or non-uniform intensity data distributions. The difference in results highlights the influence of data availability and methodological choices on the accuracy of computing earthquake parameters. Addressing these challenges requires refining intensity attenuation models, expanding macroseismic data collection and integrating historical and instrumental records. These improvements will enhance the reliability of seismic hazard assessments and contribute to a more robust understanding of the seismic activity in Algeria.