GNGTS 2015 - Atti del 34° Convegno Nazionale

30 GNGTS 2015 S essione 2.1 Monitoring of deep seated underground fluids within the frame of the DPC-INGV S3 Project (2nd year) G. Martinelli 1 , P. Bonfanti 2 , A. Dadomo 3 , F. Gentile 4 , F. Gherardi 5 , F. Italiano 6 , L. Pierotti 5 , A. Riggio 4 , M. Santulin 4 , A. Tamaro 4 1 ARPA, Emilia Romagna, Reggio Emilia, Italy 2 INGV Catania, Italy 3 Geologist, freelance, Fiorenzuola d’Arda, Italy 4 INOGS, Trieste, Italy 5 CNR-IGG, Pisa, Italy 6 INGV, Palermo, Italy Introduction. Detection of short-term crustal strain variations may represent an important tool for monitoring ongoing seismogenic processes. Beside direct geodetic observations, indirect information may be obtained by monitoring deep seated fluids whose conditions may be significantly affected by volumetric strain fluctuations. Sites particularly sensitive to this kind of observations are usually found along active faults, in thermal springs or in deep wells reaching confined deep reservoirs able to act as natural strain-meters (Bodvarsson, 1970). These sites were identified as suitable for short term and medium term earthquake prediction experiments (Martinelli, 2015; Martinelli et al. , 2015). In these areas, observations relative to deep seated fluids (groundwater piezometries, spring outflow, Co 2 and Rn emissions) may provide useful information about preparation of future earthquakes. In this line, project DPC- INGV-S3 on “Short Term and Medium Term Earthquake Prediction and Preparation” promoted collection and analysis of geofluids data in order to: 1. identify groundwater and gas trends characterized by time series >3 years (or by gas chemical analysis, including Radon) to be compared with other observations relative to strain filed variations and thermal anomalies detected from satellite observations (Tramutoli et al., 2013); 2. map ranges of CO 2 flux in Italy with indication of potential maximum non meteorological fluctuations deriving by crustal deformations or by seismicity rate variations. The following, a possible model linking geofluids fluctuations and strain field variations is presented. Then, main outcomes of the two-year activity of the project will be shortly outlined. Mechanisms for earthquake hydrologic precursors. About all geochemical anomalies observed before earthquakes have been retained attributable to deep fluids pressure variations induced by crustal deformative processes since fluid pressure is proportional to stress and volumetric strain. The stress-strain relationship for an isotropic, linearly elastic porous medium was studied, among others, by Rice and Cleary (1976 and references therein) and by Roeloffs (1996). In particular the stress tensor σ kk , the volumetric strain ε kk and the fluid pressure p under undrained conditions can be described as: p = - Bσ kk / 3 (1) p = -2GB(1 + ν u )ε kk /3(1 + ν u ) (2) where G is the shear modulus, B is the Skempton coefficient, and ν u is the Poisson’s ratio under undrained conditions. As a consequence the fluid pressure is proportional to stress and volumetric strain. Thus groundwaters can be, in principle, utilized as natural strainmeters being water about incompressible and monitorable by large scale networks. To better constrain the candidate area eventually identified by anomalous signals detected from large scale networks attempts have been carried out to identify areas affected by maximum amplitude signals (Popov and Vartaniyan, 1990; Vartanyan et al. , 1992). Piezometric level recorded in the Po Plain. PCA analysis of the PR-RE-MO cluster of piezometric wells identify in PC1 a sort of V trend while PCAanalysis of the FE-RA-BO cluster of piezometric wells identify a long term increasing trend. Not detrended graphs probably better