Abstract The European Marine Observation and Data Network (EMODnet) Project provides freely available data on European seas. The main purpose of EMODnet is to overcome the fragmentation and dishomogeneity of the available data, providing access to a harmonized and interoperable database. The EMODnet Geology Lot includes information at multiple scales on the seabed and its substrate ( http://www.emodnet-geology.eu/ ). The dataset on ‘Geological events and probabilities’ collects information on landslides, earthquakes, volcanic structures, active tectonics, tsunamis and fluid emissions. The Geological Survey of Italy, which coordinates the collation of ‘Geological events and probabilities’ data, provided guidelines to compile layers complemented by comprehensive and detailed patterns of attributes for each feature in order to characterize each type of geological event. Occurrences of events are often associated with each other, particularly in tectonically active areas. Geological events affect both submerged and coastal environments. Data gathered by EMODnet Geology provide a good basis for further studies, contributing to the outlining of different tectonic settings and providing support to the use of marine resources, as well as to the management of marine-coastal areas particularly regarding the identification and assessment of geological and environmental hazards.

Abstract Understanding the behaviour of fluids in hydrothermal systems is a key factor in volcano monitoring. Measuring gas emissions in volcanic areas is strategic for detecting and interpreting precursory signals of variations in volcanic activity. The role of radon as a potential precursor of earthquakes has been extensively debated. However, radon anomalies appear to be better suited to forecast eruptive episodes as we know the loci of volcanic eruptions and we can follow the evolution of volcanic activity. Radon mapping is an effective tool in assessing diffuse and concentrated degassing at the surface. We hereby summarize the in-soil radon emissions collected worldwide and further discuss a collection of data on our key targets. These are closed-conduit and open-conduit volcanoes: Vesuvius (Italy) and La Soufrière (Guadeloupe, Lesser Antilles), Stromboli (Italy) and Villarrica (Chile), respectively. In all the above volcanoes, faults and fracture systems control radon degassing. Automatic and real-time measurements help us to detect major changes in volcanic activity. We present and discuss the radon time series associated with the last effusive eruption at Stromboli. Spectral analyses reveal diurnal and semi-diurnal cycles being probably modulated by atmospheric variations. Multiple linear regression (MLR) analyses have been performed by filtering the radon signals from the effects of local environmental parameters. The residuals do not show particular variations or precursory peaks as the gases have been released from this open-conduit volcano before the onset of the effusive phase (7 August 2014). It is finally emphasized that radon is not the sole precursor, and we should also rely on other geochemical and geophysical parameters. In this perspective, we propose a methodological procedure that can contribute to improving volcano surveillance in an attempt to mitigate volcanic risk.