GNGTS 2023 - Atti del 41° Convegno Nazionale

Session 3.2 ___ GNGTS 2023 in the same area of the 2021 ERT surveys and its orientation was chosen on the basis of the maximum slope gradient. The survey has been performed with an array of 72 electrodes spaced 1.5 m. The measurements have been collected with a Syscal Pro georesistivimeter (Iris Instruments) using a dipole-dipole configuration with multiple skips (1, 3, 5 and 7). Direct and reciprocals measurements have been acquired and used to estimate a reliable experimental error for the collected datasets (Binley, 2015). Due to the high contact resistances between electrodes and the coarse-blocky surface, usually the ERT data quality in rock glacier environments is low (Hauck & Kneisell, 2008). To partially overcome this problem, we facilitated the injection of the electrical current (Pavoni et al., 2022a) by placing the electrodes between the boulders with sponges soaked with saltwater (Fig.3A). Note that, the sponges have been soaked before each measurement of the ERT time lapse survey to ensure uniform (approximately) contact resistances for each collected dataset. The water for the experiment has been stored on the rock glacier during the previous months, using ten 120-liters bins. The bins have been placed in the point selected for the water injection (violet star in Fig.2D) in early spring when snow cover was still present (Fig.3B), they were filled with snow and covered with nylon sheets pierced at their center to store also rainwater. This way, in mid-June, once the snow cover on the rock glacier has melted, we retrieved more than 1000 liters of water to perform the infiltration experiment. Before the survey, 3 kg of NaCl were added to each bin to create a high electrical conductivity solution. After collecting the time zero dataset, we injected 8 bins of salt water at the selected point in few minutes (Fig.3C). In this way, more than 800 liters of salt water were injected into the frozen subsoil system. The remaining 200 liters of water have been used to wet the sponges before each new data acquisition of the experiment. Four ERT datasets have been acquired in the first hour with a time interval of 15 minutes, followed by five datasets collected at hourly intervals, and a last dataset was measured 24 hours after the water injection. Note that, no rain or uncontrolled water contribution occurred during the experiment but, due to adverse weather, we could not extend longer the survey to check the time required for the subsoil to come back to the pre-injection conditions.