GNGTS 2013 - Atti del 32° Convegno Nazionale

do not show the profiles of their georadar investigation and therefore don’t allow a critical evaluation of their measures and data. The divergences between the models of Delle Rose and Leucci (2010) and Margiotta et al. (2012) might be explained from the different results of the respective stratigraphy and geophysical surveys. It must be observed that Margiotta et al. (2012) ignored both the aforementioned borehole survey performed on 2008 and the paper of Calò et al. (2011). However they misunderstand the model of Delle Rose and Leucci (2010) especially when assert that such a model “is based on a stratigraphy consisting exclusively of carbonate deposits, without taking into the due account the crucial presence of cover deposits above the calcarenite bedrock, and primarily that of the organic-rich clays [i.e. marsh deposits]” (Margiotta et al. , 2012, p. 667). Margiotta et al. (2012) described the marsh deposits with a thickness up to of 11 m and overlying (without interruptions) the calcarenites along the coast for an outcropping area wide from 100 to 250 meters. This stratigraphy setting does not fit the boreholes data of the aforementioned project funded by the Ministry of Environment and Protection of Land: in fact only 6 perforations out of 23 crossed palustrine deposits of thicknesses ranging from 0.3 to 1.4 m. Moreover, the whole stratigraphical model proposed in Margiotta et al. (2012) appears based on subjective considerations. Such Authors affirmed that their “research involved […] the critical revision of more than 40 well-core stratigraphies”, without describing what is such a “critical revision”. Their subsoil model seems mainly based on interpretation of geophysical measures. Consequently the controversy between the models embraced also methodological aspects regarding data processing and interpretation. We have to observe that for geophysical data Margiotta et al. (2012) associate individual radar signal reflections to specific geological boundaries and even associate a radar signal reflection to a prediction of subsidence. However, it is common practice to perform the radar profiles in correspondence of stratigraphical columns or trenches to be able to associate a specific radar reflections to geological stratigraphy (Conyers and Goodman, 1997; Conyers, 2004). A confrontation with the geophysical data of Margiotta et al. (2012) is really difficult because the figures were not possible to read the metric scales, time scales, and resistivity scales. Their radar sections are surely affected by ringing (Conyers and Goodman, 1997; Conyers, 2004) probably related to low penetration of radar signal. Seems that the radar signal do not propagate below 20ns. This is due to the high conductive materials presents in the subsoil as confirmed by ERT results presented by the same Authors. In fact in the surveyed area electrical resistivity seems to ranging from 0.5 to 20 Ω·m, index of highly conductive subsurface. Therefore the anomalous zones below 20 ns could probably be an artifact due to the application of the back ground removal filter. For ERT profiles Margiotta et al. (2012) associate resistivity values to specific geological formations. For example 0.5–2 ohm m were related to clay sediments without any reference to a nearby stratigraphycal core or a bibliography. In fact Reynolds (2011) at page 291 done table in which correlates various types of geological materials with a range of values of electrical resistivity (for clay materials resistivity value ranging from 1 to 150 Ω·m). Also in Loke (2011, page 6) for clay materials, resistivity value ranging from 1 to 100 Ω·m, while for the sea water resistivity value is about 0.2 Ω·m (due to the relatively high salt content). Given the above, the stratigraphical features of the marsh deposits described by Margiotta et al. (2012) appear inconsistent and the result of a wrong interpretation of geophysical data. Consequently, their model of sinkhole susceptibility is questionable. Conclusions. The case of Casalabate is a good example of how geological uncertainties can worse the hazard assessment. In consideration of the complexity of the problem and the concerns of the collectivity, an efficient predictive model of the hazard should be defined. In order to select the most appropriate geological-geophysical model, confrontations among all the researchers working on this issue would be desirable. Especially the knowledge exchange between geologists and geophysicists must be improved. As revealed the ERT survey exposed in this report, dangerous cavities could be placed also out of the area repeatedly interested 111 GNGTS 2013 S essione 3.2