A magnitude 6.5 earthquake struck central Botswana on Monday, 3 April 2017 at 7h40pm, the largest to occur on this con-tinent in 11 years. The location was unusual as most large tectonic quakes in southern Africa are closely associated with the East African Rift System, an extension failure zone stretching from Djibouti by the Red Sea all the way to southern Mozambique. The key words here are ‘most tectonic’: the most seismically active regions in Africa are the gold mining dis-tricts along the edge of the Witwatersrand Basin, but these events are associated with mining and not with plate tectonics.
With an estimated depth of 29km, the source of the M6.5 tremor is not related to either mining or even fracking, as was guessed by some fake-news articles on the internet. It is simply too deep to be influenced by human activity. And, with 25 global seismic stations having recorded the quake, the source location is too accurate to allow for much uncertainty (see wave paths in Figure 1 below).
Thanks to the high data quality, seismologists were able to invert the source mechanism: normal slip failure with either shallow dip towards ENE or a steep dip towards WSW (see the beach ball symbol in Figures 1 and 2 below)., We cannot be sure which one occurs, as there are always two possible solutions to the model commonly used to determine the failure mechanism from incident wave patterns.
The tremor location was distant from towns and major civil infrastructure, being approximately 130km from Moijabane, cen-tral Botswana. In Serowe, the closest settlement, residents reported ‘it was shaking heavily, the light bulbs were shaking, the windows almost cracked’; in Gaborone, ‘windows rattled like mad’ and, at least one person has ‘neva been so scared in my life.’ Others reported strong shaking for several seconds, plates and vases falling off shelves, and some people evacuating houses.
In the Johannesburg CBD, where seismic waves arrived around 4 minutes later, high-rise buildings swayed for 10–15 sec-onds, couches were shifting and dogs ‘went mad’. However, no structural failures were reported and no significant injuries were reported to persons (or pets!). Shaking, or at least notable vibrations, were felt as far away as Knysna and Walvis Bay.
These eyewitness reports allow for the compilation of a Modified Mercalli Intensity (MMI) map, which is based on the effect an earthquake has on its environment. Obviously, the greater the distance to the source, the lower is the MMI. Mercalli in-tensity is complementary to the Richter scale, which is a measure of the source strength. Richter Magnitude is fixed and does not change with distance.
Interestingly, the slip duration was around ten seconds and the energy release profile shows changes over time, which indi-cates that varying amounts of strain energy were released until the failure process terminated and the slip surface locked up again and stabilised (see source time function in Figure 3 below). The seismograms also show a number of small slips prior to the main rupture, i.e. rupture was probably initiated on a small scale and then escalated from there (see Figure 4 below).
Amazing what you can learn from seismic data … and unless you are a mine seismologist, you don’t even have to go down there!