3-D attenuation image of fluid storage and tectonic interactions across the Pollino fault network

The Pollino range is a region of slow deformation where earthquakes generally nucleate on low-angle normal faults. Recent studies have mapped fault structures and identified fluid-related dynamics responsible for historical and recent seismicity in the area. Here, we apply the coda-normalization method at multiple frequencies and scales to image the 3-D P-wave attenuation (QP) properties of its slowly deforming fault network. The wide-scale average attenuation properties of the Pollino range are typical for a stable continental block, with a dependence of QP on frequency of $Q_\mathrm{ P}^{-1}=(0.0011\pm 0.0008) f^{(0.36\pm 0.32)}$. Using only waveforms comprised in the area of seismic swarms, the dependence of attenuation on frequency increases [$Q_\mathrm{ P}^{-1}=(0.0373\pm 0.0011) f^{(-0.59\pm 0.01)}$], as expected when targeting seismically active faults. A shallow very-low-attenuation anomaly (max depth of 4-5 km) caps the seismicity recorded within the western cluster 1 of the Pollino seismic sequence (2012, maximum magnitude Mw = 5.1). High-attenuation volumes below this anomaly are likely related to fluid storage and comprise the western and northern portions of cluster 1 and the Mercure basin. These anomalies are constrained to the NW by a sharp low-attenuation interface, corresponding to the transition towards the eastern unit of the Apennine Platform under the Lauria mountains. The low-seismicity volume between cluster 1 and cluster 2 (maximum magnitude Mw = 4.3, east of the primary) shows diffuse low-to-average attenuation features. There is no clear indication of fluid-filled pathways between the two clusters resolvable at our resolution. In this volume, the attenuation values are anyway lower than in recognized low-attenuation blocks, like the Lauria Mountain and Pollino Range. As the volume develops in a region marked at surface by small-scale cross-faulting, it suggests no actual barrier between clusters, more likely a system of small locked fault patches that can break in the future. Our model loses resolution at depth, but it can still resolve a 5-to-15-km-deep high-attenuation anomaly that underlies the Castrovillari basin. This anomaly is an ideal deep source for the SE-to-NW migration of historical seismicity. Our novel deep structural maps support the hypothesis that the Pollino sequence has been caused by a mechanism of deep and lateral fluid-induced migration.