In this paper we employ a combination of gravity and hydrologic data to constrain a hydraulic model of the Škocjan Caves, an allogenic dominated karstic system in Slovenia. The gravity time-series recorded by a spring-based gravimeter, are carefully analyzed to remove tidal and non-tidal effects and unveil the local hydrologic contribution, which is influenced by the temporary accumulation of water in the cave system during the flood events of the Reka river.We make use of a combined analysis of three large flood events with peak river discharge of about 200, 230 and 300 m3/s, that caused significant water level and gravity variations sensed by the pressure transducer and by the gravimeter. By the integration of hydraulic modelling we study the different coupled gravimetric-hydrologic responses to these flood events: we show that, depending on the peak discharge and duration of the event, different flow conditions are present in the cave system. In addition to the information provided by the pressure transducer, the gravimeter is sensitive to the flow dynamics in a different sector of the cave due to the choice of its location; this configuration helps to better constrain the hydraulic model.Moreover, we find that the autogenic recharge by percolating water can significantly affect the gravity time-series and must be considered in related models. By inclusion of both the hydraulic model outcomes and of the modelling of the autogenic recharge, we are able to better explain the gravity transients during the two smaller magnitude events. In particular, during such events the autogenic contribution produces a transient gravity signal, which is about 4 times larger than the allogenic one, while during the largest flood the allogenic contribution drastically overcomes the autogenic effect by a factor 20.By discussing this case, we prove the potential of terrestrial gravity observation to depict the hydro-dynamics of these complex karstic systems as well as the potential of gravimetry to remotely monitor these storage units.
Read full abstract