Perhaps the most infamously unexpected earthquake was that of 17 December 1811 that shook the historically quiescent middle Mississippi valley with an estimated magnitude of 7 on the Richter scale. The area centred on New Madrid has been resonating with seismic events of lesser magnitude ever since. So too has the area around Charleston, South Carolina on the passive Atlantic margin of the USA, which experienced a magnitude 7 earthquake in 1886. Geophysicists now know to expect major earthquakes at some time in some place along active plate margins, especially subduction zones and boundaries dominated by strike slip motion, although prediction is an art to be learned if indeed it will ever be possible. Yet even small tremors far from plate boundaries within continental parts of plates are a continual worry. The shock of totally unexpected devastation in New Madrid and Charleston makes seismic-risk assessors mark the card of any such events, especially if repeated. Ideally, plate interiors should be rigid and safe. The magnitude 7.9 Sichuan event in May 2008, which caused more than 80 thousand deaths along a fault with no history of activity, reinforced worry. All three examples were situated in areas with old faults, of which most areas of continental crust have plenty, though some are hidden. Somehow tectonic forces had built up and eventually they failed.

Protracted activity might seem to foretell more big ‘quakes. However, it now appears that faults in continental interiors behave very differently from those at plate boundaries: aftershocks, even some with magnitude 6, continue for centuries in the first case, but only for a few years or decades at tectonically active margins (Stein, S. & Liu, M. 2009. Long aftershock sequences within continents and implications for earthquake hazard assessment. Nature, v. 462, p. 87-89). The duration of aftershocks in inversely related to the tectonic load sustained by faults. A lesson suggested is that assigning high risk to continental areas with repeated seismicity overestimates the dangers. But does this mean those seismically stable areas in continental interiors pose underestimated risks? The answer is probably ‘Yes’, if they are near to old faults. That is not to say that the Caledonian and Variscan structures that divide Britain into many small blocks are about to ‘go off’ at any time. Some do generate small, noticeable tremors such as that beneath Market Weighton in east Yorkshire at 1 am on 27 February 2008 that woke people up to several hundred kilometres away (including me). Market Weighton was an area of reduced subsidence during Jurassic sedimentation, as a result of flanking Variscan faults in the crust beneath. However, if large structures – high-rise buildings, bridges, dams and power stations – are planned, it would be wise to look in detail at local faults. One approach is to map disturbance of superficial sediments that in Britain would show activity over the last 18 to 11 thousand years since ice sheets melted. Another is to check bedrock geology for the last major movements on faults. It may become possible to develop models of seismic cyclicity for all large structures to give realistic assessments of risk in the future.

See also: Parsons, T. 2009. Lasting earthquake legacy. Nature, v. 462, p. 42-43.