Verifica della qualità dei dati di tomografia elettrica e relative influenze sui risultati d'inversione

It is well established that applied geophysics provides effective support to more traditional direct investigation methods, which typically supply only punctual and often incomplete information. Among the many geophysical techniques available, electrical resistivity tomography (ERT) is one of the most used for both geological and hydrogeological characterization of the subsurface. Nevertheless, a careful data analysis and a correct interpretation of the ERT data are mandatory, in order to obtain reliable information regarding the investigated domain. Traditionally, the ERT data are acquired by investigating each electrode quadrupole only one time (so-called “direct measurement”) and the data thus obtained are directly processed by means of geophysical inversion software. This procedure, however, may provide wrong inversion results, which in turn could lead to a misinterpretation of the actual resistivity distribution. This is mainly due to the lack of a proper data error estimation. To overcome this issue, it is possible to add to the acquisition scheme the so-called “reciprocal measurements”, where the current electrodes and the potential electrodes of the same quadrupole are switched. As a consequence, for each quadrupole two transfer resistance values are available. Moreover, these values should be identical, according to reciprocity effect. Any inconsistency can be considered an estimation of the measurement error, therefore this acquisition type allows filtering the available data, rejecting those measurement couples with an error greater than a fixed threshold. In this work, based on data analysis and inversion, we compare the results obtained with traditional ERT acquisitions, i.e. using only direct measurements, with the acquisition of both direct and reciprocal values. The main aim is to check the quality of the ERT data acquired following the two approaches described above. This is achieved by means of appropriate processing and data inversion software, in combination with graphical representation tools. The analyzed data come from two different field sites: the construction site of the new high-speed trains. Railway station in Bologna and the reconstructed embankment of the Frassine river in Megliadino San Fidenzio (PD). The main results show that acquiring both direct and reciprocal measurements gives more effective and reliable results with respect to the dataset made up of only direct measurements.