Assessment of earthquake locations in 3‐D deterministic velocity models: A case study from the Altotiberina Near Fault Observatory (Italy)

Abstract

The accuracy of earthquake locations and their correspondence with subsurface geology depends strongly on the accuracy of the available seismic velocity model. Most modern methods to construct a velocity model for earthquake location are based on the inversion of passive source seismological data. Another approach is the integration of high‐resolution geological and geophysical data to construct deterministic velocity models in which earthquake locations can be directly correlated to the geological structures. Such models have to be kinematically consistent with independent seismological data in order to provide precise hypocenter solutions. We present the Altotiberina (AT) seismic model, a three‐dimensional velocity model for the Upper Tiber Valley region (Northern Apennines, Italy), constructed by combining 300 km of seismic reflection profiles, six deep boreholes (down to 5 km depth), detailed data from geological surveys and direct measurements of P and S wave velocities performed in situ and in laboratory. We assess the robustness of the AT seismic model by locating 11,713 earthquakes with a nonlinear, global‐search inversion method and comparing the probabilistic hypocenter solutions to those calculated in three previously published velocity models, constructed by inverting passive seismological data only. Our results demonstrate that the AT seismic model is able to provide higher‐quality hypocenter locations than the previous velocity models. Earthquake locations are consistent with the subsurface geological structures and show a high degree of spatial correlation with specific lithostratigraphic units, suggesting a lithological control on the seismic activity evolution.

Mixed‐Mode Slip Behavior of the Altotiberina Low‐Angle Normal Fault System (Northern Apennines, Italy) through High‐Resolution Earthquake Locations and Repeating Events

“…Thus, the PF does not root in the low‐angle ATF and occurs entirely in its hanging wall. This finding agrees with the recent work by Latorre et al (). The authors relocated in a deterministic velocity model a similar earthquake catalogue recorded during the 2010 Pietralunga sequence, showing that these events occur in the high P wave velocity layer composed of dolomites and anhydrites of the Triassic evaporites ( V P within 6.1 and 6.3 km/s), a lithology prone to the generation of both seismic and aseismic slip behaviors (De Paola et al, ; Trippetta et al, ).…”

“…This thickened zone corresponds to fault portions where the ATF surface marks the contact between different lithologies (i.e., phyllitic layer of the acoustic basement and the underlying crystalline basement). At larger depths, the ATF surface is entirely located within the crystalline basement; thus, it does not crosscut any velocity contrast (from Latorre et al, ). Another significant feature is the presence of multiple subparallel fault surfaces between 10 and 15 km depth (insets c and d).…”

Seismic Reflection Profiles and Subsurface Geology of the Area Interested by the 2016–2017 Earthquake Sequence (Central Italy)

“…In fact, the seismic distribution derives from the locations published by Chiaraluce, Di Stefano, et al (), using an approach based on a simple, 1‐D velocity model. Previous experience suggests that the earthquake locations may be significantly improved, as far as more sophisticated velocity models and relocation algorithms are adopted, as observed for the case of 2009 L'Aquila (Chiaraluce, ; Valoroso et al, ) or of the 1997–1998 Colfiorito earthquakes (Latorre et al, ). Similarly, our proposed geological model is the result of a preliminary interpretation of a small part of the available seismic reflection database, which may be significantly improved in the future, by integrating other seismic profiles and possibly other geophysical data.…”

“…The depth conversion of the interpreted seismic profiles was performed using interval velocity data, measured in the deep boreholes of the region, and considering the data sets compiled in previous studies of the Umbria‐Marche Apennines stratigraphy, reported in Table (Bally et al, ; Barchi et al, ; Bigi et al, ; Latorre et al, ). The velocity values adopted for our simplified velocity model are summarized in Figure .…”

Aseismic deformation associated with an earthquake swarm in the northern Apennines (Italy)

“…We furthermore assume a homogeneous density and a constant Poisson ratio of 0.25, as already assumed for the calculation of the Green's functions. We derive the rigidity modulus for the shallowest line of patches (i.e., depths up to 1.7 km) using the seismic P wave velocity profile for the region, i.e., using a V P of 4 km/s for the shallow section and V P of 6 km/s for the deep section [ Latorre et al , , Figure 2]. We obtain a rigidity modulus for the shallow region equal to 13.3 ± 2.2 GPa.…”