Long-term GNSS measurements from the northern Adria microplate reveal fault-induced fluid mobilization

- We detected a transient GPS signal in N-Adria, after removing hydrological effects
- The transient is compatible with a porosity wave propagating through the region
- Through tomography, we locate its source on the NW-continuation of Ravne fault
- It originated about 3.5 months before the Bovec (Slovenia) 2004 Mw = 5.2 event
- The causal ~ 0.23 bar effective stress is due to a valve behavior of Ravne fault

Due to the development of denser permanent Global Navigation Satellite System (GNSS) networks over the last decade, the observation of transient deformations has significantly increased, mainly in high strain-rate zones. We analyzed the data from a group of permanent GNSS sites on the N-Adria microplate, where anomalous southward tilting and low-frequency tremors preceded the 1976 Mw = 6.4 earthquake. We present records from different stations of a transient signal with an approximately 2-year period that propagated through the northern edge of Adria, in a region 150 km wide. This represents the first time a transient deformation event has been observed in a continental collision area. We exclude surface and groundwater hydrological load effects because we corrected the data for such effects at seasonal, annual, and multiyear scales. The movement is initially upward, except in one location, with slight tilting parallel to the direction of the main tectonic structures. Later, the opposite behavior is observed.The novel methods used include earthquake location techniques and tomographic inversion of the arrival times. These methods demonstrated that the transient source was located spatially and temporally close to the 2004 Mw = 5.2 event in Bovec (Slovenia), attributed to the activity of the Ravne fault. We interpret the transient rises as the expression of a porosity wave, possibly produced by fault valve behavior of the NW tip of the Ravne fault. The propagation velocity is consistent with this hypothesis. As a further test, we invert the arrival times of the transient through hydraulic tomography to obtain hydraulic diffusivity: the values are compatible with the lithotypes present in the region and the literature. By substituting the tomographic velocity and diffusivity in the solitary/porosity wave equation, we infer an initial effective stress of approximately 0.23 bar, sufficient to alter the equilibrium of some fault segments and influence the subsequent seismicity.