No abstract

We analyzed 21 rockfalls that occurred in limestone cliffs of the Chartreuse Massif (French Alps). These rockfalls were detected both by Terrestrial Laser Scanning or photogrammetry and by a local seismological network. The combination of these methods allowed us to study relations between rockfall properties (location of detachment and impacts areas, volume, geometry, and propagation) and the induced seismic signal. We observed events with different propagation modes (sliding, mass flow, and free fall) that could be determined from digital elevation models. We focused on events that experienced a free fall after their detachment. We analyzed the first parts of the seismic signals corresponding to the detachment phase and to the first impact. The detachment phase has a smaller amplitude than the impact phase, and its amplitude and duration increases with rockfall volume. By measuring the time delay between the detachment phase and the first impact, we can estimate the free‐fall height. We found a relation Es = aEpb between the potential energy of a rockfall Ep and the seismic energy Es generated during an impact, with a = 10−8 and b = 1.55 and with a correlation coefficient R2 = 0.98. We can thus estimate both the potential energy of a block and its free‐fall height from the seismic signals. By combining these results, we obtain an accurate estimate of the rockfall volume. The relation between the Ep and Es was tested on different geological settings and for larger range of volumes using Yosemite, Mount Granier rockfalls, and with a data set of controlled releases of blocks (Hibert et al., 2017, https://doi.org/10.5194/esurf-5-283-2017, https://www.earth-surf-dynam.net/5/283/2017/).

From 1982 to 1989, more than 520 local seismic events were recorded by a permanent network of 9 telemetered stations over an active gas field (Lacq, France). The geological data of 46 boreholes coupled to several tens of kilometers of seismic profiles, concentrated in a small volume of 15 × 15 × 10 km3, allow the development of a very precise velocity structure, exhibiting a dome geometry limited by faulted strips. The discretized velocity model includes almost 20000 blocks with 500 m × 500 m × 250 m minimum dimensions. A relocation technique using this a priori detailed 3D velocity model allows precise locations of induced earthquakes. 351 relocated hypocenters show that of the seismic activity concentrated on geometrical discontinuities of the local dome structure. The locations of seismic events are correlated with geomechanical layers of the structure. Spatial and temporal migration of seismic ruptures mimics the different permeability rings of the reservoir which are progressively depleted by gas extraction. Most of these events are at present confined close to major thrust planes which limit the dome structure laterally, and in depth onto preexisting faults in the basement below the gas field. The occurence of these deep clusters demonstrates that fluid manipulation in the sedimentary covers can activate preexisting basement faults. This result changes previous assessments of local seismic risk and provides new insights into fault mechanisms in areas with complex geological structure.

No abstract

No abstract