Vol.:(0123456789) 1 3 Environmental Earth Sciences (2020) 79:188 https://doi.org/10.1007/s12665-020-08928-1 ORIGINAL ARTICLE A Messinian model explains the salt contamination of the Mediterranean Coastal Springs Eric Gilli 1 Received: 15 October 2019 / Accepted: 3 April 2020 © Springer-Verlag GmbH Germany, part of Springer Nature 2020 Abstract Mediterranean submarine karst springs could feed millions of persons in areas that sufer from scarcity of water but they are brackish which limits their use. The water comes from deep karst galleries which were explored by cave divers. There were many attempts to catch the water by building dams or bells to prevent sea intrusion in the cave, or to augment the water head for lowering the haline interface, according to the Ghyben–Herzberg relation. They failed as the water remained brackish upstream to the works. Recent studies show that salinity is acquired at a very important depth (> 200 m) by mixing of fresh water and sea water in deep zones where karst galleries were dug during the Messinian salinity crisis (5.9–5.3 Ma), when the sea level dropped down to 1500–2500 m. The validation of this Messinian model is a starting point for the use of this water resource by drilling deep wells, upstream of the galleries junction or by blocking the intrusion in the Messinian galleries. Keywords Karst · Submarine springs · Messinian · Salinity · Cassis Introduction Coastal karst springs are common on the Mediterranean sea mainly on its northern bank, but most of them are brackish, which limits their use for water supply. They are either on the sea shore (e.g. Kiveri, Greece), or inland (e.g. Herak- lion, Crete) or under the sea (e.g. Mortola, Italy) (Fleury et al. 2007). As they are located in areas that sufer from scarcity of water, their catchment is a challenge since dec- ades. Numerous attempts (dams, fexible or rigid artifcial insulations, pumping, etc.) were conducted at the outlet of the springs, either inland, as for the Almyros spring (Her- aklion, Greece) (Arfb 2001) or ofshore, at Mortola and Galeso springs (Italy) (Stefanon 1984); Cabbe (France) (Gilli 1999a) and in several places along the Dalmatian coast (Breznik 1998) (Fig. 1). The strategy was based on the Ghyben–Herzberg formula, which defnes the position of the salt interface in a porous aquifer: Z =  f  s - f ⋅ H (Z interface depth, H aquifer water- head,  f fresh water density,  s seawater density). In Mediter- ranean Sea, Z is roughly 40 H. The augmentation of the water head was supposed to provoke the migration of the saltwater interface far from the karst conduit. Unfortunately, they all ended in failure (Gilli 2003). One of the most important catchment system was made in Heraklion Almyros (Crete-Greece) a spring that is brackish in summer season. A circular dam was built to rise the water level up to 6 m above the spring. However, it was impossible to decrease the salinity signifcantly (Arfb 2001). Same problem occurred in Cassis (South-eastern France) where the brackish springs of Port-Miou and Bestouan dis- charge from submarine karst galleries. An underground dam was built in the 1970s to totally block the gallery (SRPM 1974). The principle was to prevent a sea intrusion in the cave entrance and to artifcially increase the fresh water head up to 3.7 m with the aim of lowering the salt interface for reducing the salinity. The salinity dropped but was never less that 3 g/L. The project was abandoned in 1974 but later studies from 1990 to 2018 explains the cause of the failure. * Eric Gilli e.gilli@wanadoo.fr 1 Geography Department, University Paris 8, 8 Place Garibaldi, 06300 Nice, France