No abstract

No abstract

A high‐resolution analysis of a homogeneous and complete data set of intermediate depth microearthquakes (h > 60 km), extending over an interval of 9.3 years emphasizes a significant deviation at low magnitudes (ML < 4) in a linear frequency‐magnitude distribution. It appears as a distinct change in the slope of the cumulative distribution and as a seismicity deficit followed by a relative enhancement in the corresponding noncumulative curve. This is in agreement with the presence of two characteristic mechanisms and allows the identification of both the magnitude threshold of asperity‐like earthquakes and the transition zone (ML = 3.3 ‐ 3.9) from crack‐like earthquakes (background seismicity) to asperity‐like events. These features are better pointed out on well individualized active zones. The background seismicity shows a tremendeous decrease in its b slope, from 1.03 to 0.57, during a 6‐year interval before the occurrence of a major earthquake in 1986 ( Mw = 7.3), followed by a fast recovery to 1.03 in 2 years after this event. Such a behavior could be correlated to the continuous growth of the shear stress‐free surface on the fault, as in a percolation process, followed by its sudden diminution due to the locking of the fault. The above results provide relevance for the discrete character of faulting and enable a coherent modeling of the earthquake generation process from microearthquake to major event scale lengths.

No abstract

Several decades of faulty exploitation of salt through solution mining led to the creation of an underground cavern containing several million cubic meters of brine. To eliminate the huge hazard near a densely inhabited area, a technical solution was implemented to resolve this instability concern through the controlled collapse of the roof while pumping the brine out and filling the cavern with sterile. To supervise this, an area of over 1 km 2 was monitored with a staggered array of 36 one-component, 15 Hz geophones installed in 12 boreholes about 160-360 m deep. A total of 2,392 seismic events with M w -2.6 to 0.2 occurred from July 2005 to March 2006, located within an average accuracy of 18 m. The b-value of the frequency-magnitude distribution exhibited a time variation from 0.5 to 1 and from there to 1.5, suggesting that the collapse initiated as a linear fracture pattern, followed by shear planar fragmentations and finally a 3-D failure process. The brunching ratio of seismicity is indicative of a super-critical process, except for a short period in mid-February when temporary stability existed. Event relocation through the use of a collapsing technique outlines that major clusters of seismicity were associated with the main cavern collapse, whereas smaller clusters were generated by the fracturing of smaller size nearby caverns. It is shown that one-component recordings allow for stable and reliable point source event mechanism solutions through automatic moment tensor inversion using time domain estimates of low frequency amplitudes with first polarities attached. Detailed analysis of failure mechanism components uses 912 solutions with conditional number CN \ 100 and a correlation coefficient r 2 [ 0.5. The largest pure shear (DC) components characterize the events surrounding the cavern ceiling, which exhibit normal and strike-slip failures. The majority of mechanism solutions include up to 30% explosional failure components, which correspond to roof caving under gravitational collapsing. The largest vertical deformation rate relates closely to the cavern roof and floor, as well as the rest of the salt formation, whereas the horizontal deformation rate is most prominent in areas of detected collapses.