Résumé
Early MagnetoTelluric (MT) resistivity soundings in Iceland revealed a Deep Conductive Layer (DCL), a layer with anomalously high conductivity, at the depth of 5-20 km (upper seismic layer 3) in the crust under most of the country. When first observed in the 1970s, the DCL was interpreted as partial melt below a thin and hot crust. Later seismic and gravity studies have showed that the crust is 20–40 km thick, dense and relatively cold. The DCL is thus an intra-crustal layer below the brittle/ductile transition and does not correlate with any currently known seismic feature. In this paper we try to narrow down the possible origin(s) and nature of the enigmatic DCL and how its presence can be understood in the framework of the thick and cold crustal model. We review existing electrical conductivity data from early MT measurements done in the 1970s and 1980s and from recent surveys mostly done in the context of geothermal exploration. The DCL domes up beneath active central volcanoes and high temperature geothermal areas. It is found to be electrically anisotropic, being more conductive in the spreading direction than in the tectonic/fracture direction. We review several possible conduction mechanisms that could explain the high conductivity in the light of recent experimental work on electrical conductivity of minerals and rocks and discuss the possible prevailing conduction mechanisms in the light of existing geophysical and petrological observations in Iceland. These investigations show that seismic and electrical surveys are both compatible with the presence of melt in the DCL in the volcanic plumbing system associated to volcanoes. Outside active volcanoes, seismic properties of the DCL, inferred from local tomography, are not compatible with the presence of melt or even aqueous fluid in the DCL but confirm thick, cold and fluid-free lower crust, leaving only solid-state conduction as a viable conduction mechanism. Review of laboratory measurements on various type of minerals leads to the conclusion that the presence of FeTi oxides or hydrated minerals, such as amphibole, could explain the DCL. In Eyjafjörður, North Iceland, the only low-temperature geothermal area studied using MT to date, the DCL locally domes up to shallow depths under the geothermal area. This suggests possible correlation between low-temperature geothermal activity and relatively shallow DCL. Detailed mapping of the DCL might therefore be an important tool in prospection for low-temperature geothermal resources.