GNGTS 2015 - Atti del 34° Convegno Nazionale

GNGTS 2015 S essione 3.2 101 section or for areas coinciding with the abnormal structures. The sequences are also tested with field data to assess the validity of the theoretical results. Choice of the acquisition patterns. Recently many authors have dealt the optimization of acquisition sequences of data, which can be formed by different arrays, dipolar lengths and distances, in order to seek the optimal sequences that ensure a realistic imaging, without the need for an excessive number of measurements. Stummer et al. (2004) have experienced an accurate approach that uses sensitivity distributions to calculate an estimate of the resolution matrix of the model. The goal is therefore to seek the optimal sequences that ensure a realistic imaging with high resilience of the subsoil, without the need for an excessive number of measurements that would compromise the economic viability of the survey. Several simulations on 2D model have been made, using the software RES2DMOD (Loke, 2014), to study the changes in resolution and reliability when the acquisition pattern changes. The choice of the acquisition sequences is based on the variation of the ratio between the maximum dipole length a max and the maximum dipole order n max , trying to keep a similar amount of number of measures. Data sets of apparent resistivity values have been calculated using the multiple gradient (MG) array (Dahlin and Zhou, 2004; 2006) and the three most common arrays for multichannel measurements: dipole-dipole (DD), pole-dipole (PD) and enner-Schlumberger (WS). For DD, PD and WS arrays sequences retain approximately the same depth of investigation and a not so different total amount of data. For these arrays 8 data sets were generated in which the maximum potential spacing a increases and the maximum n factor decreases (Tab. 1), so increasing the number of current dipoles and therefore the time of acquisition (Tab. 2). Tab. 2 - Number of measures and current dipoles for each data set.    Nr. of measures    Nr. of current dipoles Pattern DD PD WS MG DD PD WS MG 1 1820 1855 1225 533 69 69 1225 36 2 1887 1938 1479 904 135 69 1479 65 3 1848 1914 1518 1227 198 69 1347 93 4 1784 1864 1504 1596 258 69 1384 122 5 1665 1755 1440 1921 315 69 1320 150 6 1635 1740 1425 2292 369 69 1251 179 7 1512 1624 1344 - 420 69 1082 - 8 1296 1404 1188 . 468 69 1043 - For the multiple gradient array (MG) we choose particular sequences, in order to minimize the number of current dipoles (and therefore time), but maintaining a resolution comparable to those of the other arrays (Fiandaca et al. , 2005). The sequences were determined by subdividing the electrode layout AB max into equal parts and by locating the current electrodes at the ends of each part: AB = AB max / e, e varying from 1 to 8 (Martorana et al. , 2009). However, this sequence does not ensure a uniform lateral coverage comparable to the classical sequences in which all the electrodes are in turn used as the current electrodes. To overcome this drawback, the dipoles of the current are increased by dividing the forwarding step of the current dipole by a coverage factor c . Six different sequences were considered, with c ranging from 1 to 6 (Tab. 1). Forward modelling and inversion. The adopted resistivity models are similar to those used by Szalai et al. (2013). Model #1 shows ten resistive prisms (100 Ωm) of the same square section (2m*2m), equally spaced and a background of 10 Ωm. The depth of the center of the first prism to the left is 2m and the following prisms are gradually deeper 25 cm until a maximum