Compiling a homogeneous earthquake catalogue for Southern Africa

Abstract

An accurate seismic hazard assessment can only be carried out if a homogeneous and sufficiently complete catalogue for the study area is available. Since the catalogue for southern Africa was last updated in the early 1990s for the Global Seismic Hazard Assessment Program (GSHAP), it is necessary that a new updated, homogeneous and complete earthquake catalogue be compiled that includes data acquired during the last two decades. The process of compiling the new earthquake catalogue for southern Africa (South Africa, Lesotho, Swaziland, Botswana, and Namibia) was done as part of the Global Earthquake Modelling (GEM) project. The data from published and unpublished sources, and databases from the South African National Seismograph Network (SANSN), Bulawayo (BUL), the Geological Survey of Botswana, the National Earthquake Information Centre (NEIC), the Advanced National Seismic System (ANSS), and the International Seismic Centre (ISC) were retrieved and evaluated. After the data from the different sources were merged, duplicates and induced earthquakes were removed. The catalogue was unified with all magnitude types converted to moment magnitude (MW). Unifying the southern African catalogue to one magnitude scale had multiple challenges, considering that the catalogue is mostly incomplete, and it was therefore not easy to derive relations between different magnitude scales. The question of which method and relations are most suitable for converting all magnitude scales to MW had to be addressed. To ensure that all the events are independent, several procedures were carried out to decluster the catalogue and most suitable method selected. The final catalogue includes all available events, i.e. historical, and instrumental events from 1690 to December 2011, excluding fore- and aftershocks and induced events. This catalogue has 920 events with MW ≥ 4 whereas GSHAP has only 100 events with MS ≥ 4 in the southern African region. The largest event in the final catalogue occurred in 1952 and is located in the Okavango Delta region in Botswana with magnitude MW = 6.7. The maximum likelihood method was used at each point on a grid covering the study area, to estimate the spatial distribution of the b-value and the activity rate. The maximum curvature method was used for estimating the spatial distribution of magnitude of completeness, which was also substantiated with the Gutenberg-Richter, time-scale and spatial-scale magnitude of completeness graphs.