James W. Dewey

Coping with earthquakes induced by fluid injection

Hazard may be reduced by managing injection activities Large areas of the United States long considered geologically stable with little or no detected seismicity have recently become seismically active. The increase in earthquake activity began in the mid-continent starting in 2001 (1) and has continued to rise. In 2014, the rate of occurrence of earthquakes with magnitudes (M) of 3 and greater in Oklahoma exceeded that in California (see the figure). This elevated activity includes larger earthquakes, several with M > 5, that have caused significant damage (2, 3). To a large extent, the increasing rate of earthquakes in the mid-continent is due to fluid-injection activities used in modern energy production (1, 4, 5). We explore potential avenues for mitigating effects of induced seismicity. Although the United States is our focus here, Canada, China, the UK, and others confront similar problems associated with oil and gas production, whereas quakes induced by geothermal activities affect Switzerland, Germany, and others.

Seismic Gaps and Source Zones of Recent Large Earthquakes in Coastal Peru

The earthquakes of central coastal Peru occur principally in two distinct zones of shallow earthquake activity that are inland of and parallel to the axis of the Peru Trench. The interface-thrust (IT) zone includes the great thrust-fault earthquakes of 17 October 1966 and 3 October 1974. The coastal-plate interior (CPI) zone includes the great earthquake of 31 May 1970, and is located about 50 km inland of and 30 km deeper than the interface thrust zone. The occurrence of a large earthquake in one zone may not relieve elastic strain in the adjoining zone, thus complicating the application of the seismic gap concept to central coastal Peru. However, recognition of two seismic zones may facilitate detection of seismicity precursory to a large earthquake in a given zone; removal of probable CPI-zone earthquakes from plots of seismicity prior to the 1974 main shock dramatically emphasizes the high seismic activity near the rupture zone of that earthquake in the five years preceding the main shock. Other conclusions on the seismicity of coastal Peru that affect the application of the seismic gap concept to this region are: (1) Aftershocks of the great earthquakes of 1966, 1970, and 1974 occurred in spatially separated clusters. Some clusters may represent distinct small source regions triggered by the main shock rather than delimiting the total extent of main-shock rupture. The uncertainty in the interpretation of aftershock clusters results in corresponding uncertainties in estimates of stress drop and estimates of the dimensions of the seismic gap that has been filled by a major earthquake. (2) Aftershocks of the great thrust-fault earthquakes of 1966 and 1974 generally did not extend seaward as far as the Peru Trench. (3) None of the three great earthquakes produced significant teleseismic activity in the following month in the source regions of the other two earthquakes. The earthquake hypocenters that form the basis of this study were relocated using station adjustments computed by the method of joint hypocenter determination.

Seismicity associated with the Sumatra-Andaman Islands earthquake of 26 December 2004

The U.S. Geological Survey/National Earthquake Information Center (usgs/neic) had computed origins for 5000 earthquakes in the Sumatra–Andaman Islands region in the first 36 weeks after the Sumatra–Andaman Islands mainshock of 26 December 2004. The cataloging of earthquakes of m b (usgs) 5.1 and larger is essentially complete for the time period except for the first half-day following the 26 December mainshock, a period of about two hours following the Nias earthquake of 28 March 2005, and occasionally during the Andaman Sea swarm of 26–30 January 2005. Moderate and larger ( m b ≥5.5) aftershocks are absent from most of the deep interplate thrust faults of the segments of the Sumatra–Andaman Islands subduction zone on which the 26 December mainshock occurred, which probably reflects nearly complete release of elastic strain on the seismogenic interplate-thrust during the mainshock. An exceptional thrust-fault source offshore of Banda Aceh may represent a segment of the interplate thrust that was bypassed during the mainshock. The 26 December mainshock triggered a high level of aftershock activity near the axis of the Sunda trench and the leading edge of the overthrust Burma plate. Much near-trench activity is intraplate activity within the subducting plate, but some shallow-focus, near-trench, reverse-fault earthquakes may represent an unusual seismogenic release of interplate compressional stress near the tip of the overriding plate. The interplate-thrust Nias earthquake of 28 March 2005, in contrast to the 26 December aftershock sequence, was followed by many interplate-thrust aftershocks along the length of its inferred rupture zone.

Testing an earthquake prediction algorithm

A test to evaluate earthquake prediction algorithms is being applied to a Russian algorithm known asM8 TheM8 algorithm makes intermediate term predictions for earthquakes to occur in a large circle, based on integral counts of transient seismicity in the circle. In a retroactive prediction for the period January 1, 1985 to July 1, 1991 the algorithm as configured for the forward test would have predicted eight of ten strong earthquakes in the test area. A null hypothesis, based on random assignment of predictions, predicts eight earthquakes in 2.87% of the trials. The forward test began July 1, 1991 and will run through December 31, 1997. As of July 1, 1995, the algorithm had forward predicted five out of nine earthquakes in the test area, which success ratio would have been achieved in 53% of random trials with the null hypothesis.

Relocation of teleseismically recorded earthquakes near Tennant Creek, Australia: Implications for midplate seismogenesis

The three MS > 6 Tennant Creek, Australia, earthquakes of January 22, 1988, were preceded by small and moderate earthquakes in 1986 and 1987. We have used the method of joint hypocenter determination to estimate the positions of teleseismically recorded 1986–1987 earthquakes, the 1988 main shocks, and aftershocks with respect to a system of surface fault scarps. The 1986–1987 shocks and the 1988 main shocks nucleated near the center of the zone of surface scarps, where fault-segment boundaries at depth are implied by complexities in the distribution of scarps at the surface. This suggests that the fault segments that ruptured in 1988 were already in existence in 1986–1987, which is consistent with the hypothesis that strong midplate earthquakes occur on preexisting faults. The redetermined 1986–1987 hypocenters are, however, also consistent with the hypothesis that midplate seismicity is localized by stress concentration due to bulk rheological heterogeneity of the crust, because they are situated on a regionally prominent gravity anomaly. The teleseismically recorded seismicity of the Tennant Creek region prior to the 1988 main shocks has the temporal pattern of a swarm followed by a lull. The concentration of swarm earthquakes between two scarps is consistent with models in which precursory-swarm earthquakes correspond to faulting that is spatially distinct from the site of primary main-shock faulting. The time interval between swarm and main shocks is similar to intervals between intermediate-term precursory swarms and main shocks in regions that have much higher rates of tectonic loading; this similarity suggests that the time intervals between precursory swarms and subsequent main shocks are not strongly influenced by the rate of tectonic loading, but are determined primarily by time dependence of the failure process. The spatial distribution of teleseismically recorded aftershocks is in most respects like that of aftershocks recorded by a local network of portable stations in the half year following the main shocks. The set of teleseismically recorded aftershocks, like the set of locally recorded aftershocks, includes some events that occurred well away from the causative faults of the main shocks. At the length scale of the 1988 main-shock rupture, the distribution of aftershocks occurring more than one year after the main shocks is not representative of the distribution of earlier aftershocks.