Gravimetry for monitoring water mass movements in karstic areas

Karst aquifers represent a significant source of water for about 1/4 of the world’s population. The water circulation in karst occurs mostly underground and it is mainly controlled by alternation of small conduits and large voids present in the rock massif. Such intricate void distribution combined with an irregular recharge provided by the rain results in fast and complex water flows with temporary accumulation of huge water volumes in the voids. The knowledge of the dynamics of such system is usually limited to the areas where a direct access to the vadose zone through speleological exploration is possible. Given the importance of such aquifers and their vulnerability it is important to have a detailed picture of the water dynamics and of the underground water paths. Gravimetry offers a valid complement to classical hydrologic measurements in order to monitor the recharge process. In this thesis, I show an innovative integration of gravimetric and hydrologic observations to constrain a hydrodynamic model of the Škocjan cave system (Slovenia). The Škocjan caves hydrology is mostly governed by the allogenic contribution of the Reka River, which during flood event causes the accumulation of several millions of m3 of water in the cave system for few hours. In 2018 I installed a continuous recording gravimeter nearby Škocjan which allowed the detection of several gravity transients related to the local hydrologic contribution. Gravity observations are sensitive to several other contributions apart the hydrology, such as Earth and marine tides, atmospheric mass redistribution, water mass variations in oceans. All these phenomena superpose their effects and should be carefully evaluated and removed before unveiling the local hydrology contribution. Before discussing the hydrologic gravity signals, the thesis illustrates the efforts in modelling and removing all the non-hydrologic related gravity contributions. The study area is close to the Adriatic Sea, hence global models of tidal and non-tidal ocean (NTO) gravity effects could be inadequate for the correction. I prove that while tidal models are sufficiently accurate to remove the marine tidal influence a dedicated correction of the NTO is required. This was fulfilled by modelling the gravity variations due to a 4D mass model of the NTO constrained by tide gauge observations. The gravity residuals, obtained after reducing the observations for all the non-hydrologic effects, revealed anomalies correlated to the Reka flooding; the transients lasted for 12-24 hours with amplitudes in the range 10-450 nm/s2. I focused my analysis on a large flood event in February 2019 that caused water level variations >90 m inside the caves and gravity variations >400 nm/s2. The gravity and the hydrologic data were used to constrain a hydraulic model of the cave system which approximated the cavity through a series of interconnected conduits with rectangular cross-section. I fitted hydrologic and gravity observations obtaining a 4D model of the water mass variations in the cave system; the model revealed that >3 106 m3 of water were temporary accumulated during the peak’s flood. The inclusion of gravity observations improves water mass budget of the caves, which previously were based relying only on hydrological observations. Finally, the gravity data allowed to draw some general conclusions on the detectability of water storage variations in karst through gravimetry. I assessed the noise level of the Škocjan gravimeter which is about 10 nm/s2 in the diurnal spectral band and which can be taken as representative of the noise level of a typical spring based gravimeter. Relying on realistic water level variations I estimated the expected gravity signals on surface due to temporary water accumulation in other caves of the Classical Karst. For all the considered caves the gravity signal is above the noise threshold, suggesting that a remote monitoring of the storage variations is feasible.