Seawater dynamics and environmental settings after November 2002 gas eruption off Bottaro (Panarea, Aeolian Islands, Mediterranean Sea)

In November 2002 several gas bursts occurred at sea in the caldera within the islets eastward of Panarea (Aeolian Islands), with degassing of CO2 lasting several months. In a depression close to Bottaro Islet (PEG1) the gas flowed violently from the depth of ∼ 14 m to the surface producing a large plume of gas and fluids. The aims of this paper are to report on the morphological modifications, the water and gas fluxes, and the water dynamics and hydrological properties near the PEG1 site. Bathymetric surveys up to 2006 and divers’ observations up to 2008 revealed that the sinkhole had been partially filled by sediments transported by currents or sliding from the rims. Depth reduced to ∼ 12 m and the shape of the depression was becoming similar to the others in the area, suggesting that explosive gas eruptions may have occurred in the past. The gas outflow generated a bubble plume that locally affected the water circulation. Different patterns in the dynamics of the water column were described using rotor current meters, ADCP, ROV and divers’ observations, and high-resolution bathymetry. A divergence about 1 m thick was generated by a surface vortex visible in December 2002 that was not found in September 2003 with reduced gas flow. A sub-surficial layer down to ∼ 1 – 2 m above bottom showed varying speed and directions possibly correlated to tides. A bottom layer of water ∼ 1 m thick flowed continuously toward the crater during both surveys. On the nearby sandy bottoms, sand dunes and 2/4 mm size volcanoclastic gravels rolling toward the emitting area were found. The fluxes of water entering at the seafloor were calculated by current velocities, height of the bottom layer, and area of the degassing vent. The input water fluxes were found to be 4.2 and 0.2×108 l/d, for December 2002 and September 2003. The total vertical output flux was also estimated; considering bubble sizes and voids, gas fluxes were 2.6 and 0.3×108 l/d. These values are similar to the fluxes calculated by geochemical and gas sampling methods. CTD data from summer 2003 and spring 2004 recorded pH anomalies with values as low as 6.3 located on main emissions. The low pH water was also distributed laterally for dozens of meters on the bottom, depicting plumes and bottom layers of altered seawater over remarkable distances. The gas burst generated a plume similar to what is described for double core bubble plumes, where the velocity of seawater entering the system is proportional to the vertical gas flux. A new perspective to monitor vent activity is envisaged, and since measuring seawater is easier than measuring gas fluxes, a continuous and real time monitoring of hydrothermal activity could be designed.