Seismic Hazard and Earthquake Source Parameters in the North Sea

Abstract

Contemporary seismicity delineated in the North Sea by modern instrumental seismograph networks, deployed and analysed most intensively since 1980, allows study of coseismic crustal deformation by integrating information on statistical seismic hazard and earthquake source parameters. The seismicity observed during 1980–86 in the North Sea shows four earthquakes exceeding local magnitude 4.0 ML, three of which are in the northern North Sea; none exceed 5.0 ML. The longer view of historical seismicity since 1900, needed to assess seismic hazard, shows 30 earthquakes of magnitude 5 or larger and extreme value analysis estimates an annual expection of 10^-2 for an earthquake exceeding magnitude 6 anywhere in the North Sea. An upper bound to earthquake magnitude around 6.75 is compatible with evidence from other studies in and adjoining the North Sea: palaeoseismic data in Scotland has been interpreted as suggesting earthquakes of magnitude 6.5–7.0 in the last 10 years BP and the second Storegga slide in the northern North Sea has been associated tentatively with a magnitude 7 earthquake around 7 × 10³ years BP. However, cumulative strain energy release diagrams applied to the historical seismicity indicate a current upper bound to magnitude of 6.44.

Seismotectonic source parameters of recent earthquakes in the northern North Sea have also been obtained. Local magnitudes of the 13 earthquakes analysed in detail span the magnitude range 1.9–4.8 ML. Spectral analysis of shear waves from the largest of these earthquakes, 4.8 ML, provides a seismic moment estimate around 4.0 (±.7) × 10¹⁴ N m. The Madariaga model of fault rupture allows estimation of seismotectonic parameters for the same earthquake: preliminary estimates are a fault of radius 270(±70) m on which the average slip was 06 (±.05) m accompanied by a stress drop of about 90 bars. The local drop in strain during this earthquake is of order 1.3 × 10^-4. Regression between these parameters allows estimation of the total coseismic crustal deformation during 1980–86, which is small. Extrapolation to historical earthquakes of larger magnitude is not accomplished satisfactorily, this may be because of inadequacies in the historical catalogue or the moment-magnitude relationship may not be uniformly linear.