Rifting, magma, and break-up in Central Afar

<p>In Afar, an active mature rift, close to break-up have been described for many decades. However, no data were provided so far to precisely constrain the long-term structure of the rift, its evolution, and the rift maturity, i.e. the documentation of ongoing break-up processes. Here, we synthetize structural, geophysical, and geochemical data that show that break-up is indeed ongoing in Central Afar.</p><p>Geological mapping and dating of volcanic units provided data to build a detailed surface cross-section, showing a dense network of continentward dipping normal faults and two main unconformities. The field cross-section allows to estimate the amount of crustal thinning and stretching (beta factor > 2.5 in the rift center). The crust first stretched quite uniformly during late Oligocene-early Miocene times and then thinned is a more localized manner during late Miocene times. Subsequently, during Plio-Quaternary times the volcanic Stratoïd series emplaced in the thinned area. Eventually, the present-day magmatic axes have been active for about 500 kyr in the thinned continental crust.</p><p>The crust decreases in thickness from 40 km beneath the Ethiopian plateau to about 20-25 km in Central Afar, from receiver functions (RF). New RF data and tomographic models confirm that thickness and show that a necking zone formed at depth most likely during late Miocene times (thinning phase).</p><p>The combination of geological and geophysical data allows to attempt for balancing the cross-section. This balancing suggests that, beneath Central Afar, the crust is much thicker than it should be. This crust in excess is interpreted as representing magma intrusions and/or underplating that occurred in the course of the Cenozoic rifting.</p><p>Trace element analyses suggest that the depth of melting has strongly decreased just after the plume impingement, suggesting an early thermal erosion of the lithosphere base. Geochemical (isotopic) data show that in Central Afar most of the magma sources were in the plume head/tail since its impingement. These data also show that the recent lavas (< 500 kyr) were almost not contaminated by continental crustal rocks, attesting for a transitional crust beneath Afar and suggesting a recent switch from crustal thinning to dyke-related divergence.</p><p>To sum up, we propose that (1) the Afar plume was channelized in the lithosphere, probably since late Oligocene times, (2) a late Miocene thinning event (necking) predated a Quaternary SDR-like volcanic wedge emplacement, and (3) divergence is now accommodated by magmatic accretion.</p>