Or Methane clathrate from the ocean floor. The Gulf of Mexico has a LOT of those deposits.
Clathrates as a cause of tsunamisDuring the formation of gas hydrates, methane and water become immobilized within the sediment pore spaces. Because of the presence of these solids (instead of pore waters and gas), the sediment can not become consolidated because the water can not be expulsed with increasing overburden as more sedimentation occurs. Cementation of the sediments does not occur when pore spaces are filled with hydrates (solid ice) rather than with water, from which minerals such as calcite can be precipitated. Gas hydrate rich sediments are thus cemented by the hydrates, which may occupy much of the sedimentary section, but which are not stable when the temperature rises or the pressure falls.
This may lead to problems during continued sedimentation and further burial of the gas hydrates: the hydrates will become buried so deeply that the temperature will increase according to the regional geothermal gradient. The hydrates will then no longer be stable, and will disintegrate into a liquid water and gas mixture. The basal zone of the gas hydrate becomes under-consolidated, possibly over-pressured because of the release of the methane, leading to the development of a zone with low shear strength where failure could be triggered and massive landslides could occur. With the landslides, more gas could escape.
Several examples of possibly gas-hydrate linked extremely large slumps have been described, e.g., on the Norwegian continental margin (Bugge et al., 1987), where debris from the giant, three-part Storegga slide, over 450 m thick, is spread over a distance of 800 km. One of the Storegga slides caused a tsunami to deposit sediment up to 4 m above the high water line in Scotland (Nisbet and Piper, 1998). There are more of these mega slides in the same region (Laberg et al., 2000).
More about clathrates:
http://ethomas.web.wesleyan.edu/ees123/clathrate.htm