Reassessment of source parameters of 'major' southern African earthquakes

Moderate to large earthquakes within an earthquake catalogue contribute significantly to the seismic hazard and risk assessment results of any region. Thus it is prudent to ensure these events have reliable source parameters (epicentres and magnitude). The dataset of events compiled in this study contains a total of 117 instrumentally recorded events of magnitude M >= 5.0, whose parameters were obtained from the Council for Geoscience (CGS) and International Seismological Centre (ISC) databases. The events are mostly located in South Africa with a few in neighbouring countries. Parametric data made up of all available phase data and amplitudes associated with each of the earthquakes were compiled. The availability of these data enabled the earthquake epicentres and magnitude values to be recalculated using the velocity model and the local magnitude relation that are currently being used by the CGS in its analysis of national seismic data. The accuracy of the relocations was determined by producing and analysing three parameters, the azimuthal distribution of seismograph stations (GAP), root-mean-square of travel time residuals (RMS) and epicenter location error data. The analysis of these parameters showed that there was an improvement in the accuracy of the relocated events. Using the ISC location algorithm, iLOC, eight preselected events were further analysed. From this analysis, two earthquakes were found to satisfy the conditions for Ground Truth (GT5(95%)) candidacy whilst four events satisfied the criteria for GT20(90%) candidacy.