Submarine landslide in southern Africa?

The award for the strangest landslide story of the week goes to this one, from southern Africa. Late last week reports started to emerge of strange tidal patterns around the southern Cape of Africa (Fig. 1).

Fig. 1: Google Earth image of the southern Cape of Africa, showing the location of St Helena Bay and Plattenburg Bay.

In particular, on 21st August at about 8:45 am (local time) a series of tidal oscillations were recorded in St Helena Bay (Fig. 2). These tidal oscillations were reportedly large enough to damage some factories, pull a car into the sea and to cause problems for boats navigating into the harbours. A couple of days later, a report emerged from the National Sea Rescue Institute (NRSI) suggesting that the "mini-tsunami" was caused by a seismically-induced submarine landslide. Fig. 2 shows that some anomalous tidal activity certainly did occur.

Fig. 2: Tide gauge data for Port Nolloth showing the anomalous behaviour on the morning of 21st August 2008 (graph from the South African Weather Service).

Now, of course submarine landslides can trigger localised tsunamis (as occurred in Papua New Guinea in 1998), and the most likely cause of such an event is an earthquake. However, aspects of this cause do not seem to ring true. First, southern Africa is not a strongly seismic area. Second, no earthquake event has been recorded at the time that the tidal oscillations were observed. And third, the oscillations appear to have occurred over a long time period, and at markedly different times along the coast. One would expect that a landslide induced tsunami would be quite short-lived and to arrive at the coast over quite a short time period. Thus, the evidence does not really seem to support this mechanism.

In the last couple of days a more considered analysis from NSRI has been reported, recognising that the cause is more likely to be meteorological than the effects of a landslide. The South African Weather Service have proposed that atmospheric gravity waves might have been the cause, although they are sitting on the fence a little.

Whilst these tidal oscillations are clearly strange, a landslide cause does not really seem to be very likely.