Gas hydrate systems linked to fluid migration within deep-sea fans: results from the SEAGAS project

Abstract submitted and acceptedInternational audienceFluid migration strongly influences gas hydrate occurrences, increasing concentrations in proportion to gas supply. An upward flow of gas-rich fluids is also central to models proposed to account for the formation of venting features within the gas hydrate stability zone (GHSZ), and for the presence or absence at its based of bottom simulating reflections (BSRs). These models are being tested by the SEAGAS project, an EC-funded collaboration of French and Brazilian research groups, which has examined gas hydrate systems within three deep-sea depocentres : the Nile fan in the Mediterranean Sea, and the Amazon fan and the Rio Grande cone on the Atlantic margin of Brazil. In each depocentre, rapid deposition has driven different forms of gravitational collapse above deep detachments. Each contains a different association of BSRs and venting features. On the Nile fan, a faint BSR of limited extent (3000 km 2 , water depths 2000-2500 m) contrasts with more widespread evidence of gas hydrates (in well logs) and fluid vents (pockmarks, mud volcanoes); the relation to pre-and post-Messinian collapse structures is unclear. On the Amazon fan, elongate BSR patches (total area 6700 km 2 , water depths 750-2250 m) coincide with the crests of an arcuate thrust-fold belt on the upper fan, which host mud volcanoes and other seafloor vents; no BSR is observed below the collapse belt on the mid-to lower fan. On the Rio Grande cone, a regional BSR (45,000 km 2 , water depths 500-3500 m) spans paired extensional and compressive belts, the former including fault-controlled fluid vents (pockmarks). In each area, BSR inversion yields temperature gradients elevated up to 5 times background values via long-wavelength (10 0-10 1 km) variations that, on the Brazilian margin, follow structural trends. We argue that these observations can be explained in terms of tectonic-and/or compaction-driven variations in the upward flow of warm fluids containing dissolved and/or free gas, which control the formation (or not) of BSRs and, in places, of gas vents. We hypothesise that temporal variations in fluid flow will change GHSZ thickness from below; bottom-up thinning will release gas and water from hydrate-rich sediments, potentially destabilising large volumes of slope sediments. This hypothesis could account for the occurrence of landslides on depocentres with low seafloor gradients (