GNGTS 2021 - Atti del 39° Convegno Nazionale

435 GNGTS 2021 S essione 3.2 We set the amplitude of the input voltage to 400 V (resulting current ranging from 0.5 to 2.5A), with a current injection time of 2 s (2 stacks), a time delay of 40 ms and a logarithmic sampling of the IP decay curve using 20 gates (first gate centered at 40 ms, last gate at 1.7 s). The resistivity and normalized chargeability (MN = ecome pivotal. stivity mapping of coastal aquifers has been often conducted employing low- hods, such as electromagnetics (e.g. Viezzoli et al., 2011) or, where resolution is using the electrical resistivity tomography (ERT) (e.g. De Franco et al., 2009). g ERT standalone may not be the best choice in complex geological scenarios where es related to different lithotypes may overlap. e additional contribution given by the induced polarization (IP) technique is be pivotal for resolving the ambiguity arising in such contexts, given the high sponse of clay minerals and anthropogenic contaminants (Slater and Lesmes, 2002). tomographic inversion of direct-current (DC) resistivity and time-domain (TD) IP urely the most applied method for many environmental applications (e.g. Kemna et g recent years, although still rarely applied for mapping saline intrusion (e.g. Slater 02). The TD DC/IP data processing is often conducted by a rapid inversion using the and the integral chargeability as model parameters, discarding the spectral ntained within the IP decay curves without considering the actual transmitter the receiver transfer function (Fiandaca et al., 2012). The spectral analysis can ise parameters, which can be used for predicting key parameters for coastal aquifers, bility and porosity (Kemna et al., 2012). s to explore the potential of the TDIP method for mapping saline intrusion on coastal eling and inversion algorithm ompasses a two-step procedure: firstly, we determine resistivity  and integral  models through a fast ERT/IP inversion using the linear approximation of Li (1994), as implemented in the VEMI algorithm by De Donno and Cardarelli the spectral inversion is performed using a Cole-Cole model parametrization ( ] ) to retrieve the spectral behavior of the selected dataset (Pelton et al., 1978), where esistivity, m 0 the chargeability, τ the relaxation time and c the so-called frequency ting ρ 0 and m 0 models are chos n to b the  and  models achieved at the last the fast ERT/IP inversion expressed in terms of the electric potential measured before and after the current [ u DC , u IPi ] ) at different time gates i = 1 , 2 ,…, N G and the TD step-off response V ( t ) is time t > 0 after the current switch-off through an inverse Fourier transform of the ain response Z q ( ω ) for each quadrupole q (Fiandaca et al., 2012): ) − 2 π ∫ 0 ∞ ℑ ( − Z q ( ω ) iω ) sin ( ωt ) d ω , (1) / activities let the aquifers be recharged by s increased rates of extraction of groundwater i have a huge impact on surface water systems resources has become pivotal. Electrical resistivity mapping of coastal aq resolution methods, such as electromagnetics increased i.e. using the electrical resistivity t However, using ERT standalone may not be th resistivity ranges related to different lithotypes Nowadays, the additional contribution give recognized to be pivotal for resolving the a polarization response of clay minerals and anth Therefore, the tomographic inversion of direct data has been surely the most applied method f al., 2012) during recent years, although still rar and Lesmes, 2002). The TD DC/IP data process DC resistivity and the integral chargeabilit information contained within the IP decay waveform and the receiver transfer function provide piecewise parameters, which can be us such as permeability and porosity (Kemna et al. This work aims to explore the potential of the aquifers. Forward modeling and inversion algorithm Our code encompasses a two-step procedur chargeability  models through a fast ERT Oldenburg and Li (1994), as implemented in (2017). Then, the spectral i version is perfo m =[ ρ 0 ,m 0 , τ , ] ) to retrieve the spectral behavi ρ 0 is the DC resistivity, m 0 the chargeability, τ exponent. Starting ρ 0 and m 0 models are cho iteration using the fast ERT/IP inversion The dataset is expressed in terms of the elect switch-off ( d =[ u DC , u IPi ] ) at different time gate derived at any time t > 0 after the current swi frequency-domain response Z q ( ω ) for each qua V q ( t ) = Z q ( ω = 0 ) − 2 π ∫ 0 ∞ ℑ ( − Z q ( ω ) iω ) sin ( ωt ) d ω ) odels achi v by a fast inversion of DC resistivity and integral chargeability for the L2, L4 and L5 lines are shown in Fig. 2. Fig. 1 - Study area at the Fogliano Lake (Pontina Plain, Central Italy), with indicatio of the TD DC/IP electrical lines (L1-L5). Fig. 2 - Resistivity (a-c) and normalized chargeability (d-f) models for L2 (a, d), L4 (b, e) and L5 (c, f) lines. The selected area for spectral TDIP inversion of the L5 dataset is within the black rectangle.