Editors note: This is a condensed version of a post that originally appeared on  RealClimate.org

It’s the unknown that grabs attention.

We don’t know the total amount of methane frozen deep beneath the ocean, but we suspect it could rival the rest of fossil fuels combined. And we don’t know how much is frozen in the Arctic’s thawing permafrost and lake sediments.

We do know those methane deposits are seeping into the atmosphere, however. And the possibility of a catastrophic release is of, course, what gives methane its power over the imagination. Journalists in particular seem susceptible to doomsday predictions from such a scenario.

We have seen methane bubbling from the sea floor in the Arctic. Lakes provide an escape path for the methane by creating “thaw bulbs” in the underlying soil, and lakes are everywhere appearing and disappearing in the Arctic as the permafrost melts.

Yet so far we haven’t seen ironclad evidence of greater methane releases due to anthropogenic warming, though such an event is certainly believable for the coming century. This brings us to the key question: What effect would a methane release have on climate?

The impact depends on whether methane is released all at once or in an ongoing, sustained manner.

Let’s pick the most likely scenario: A slow ongoing release.

Methane is a powerful greenhouse gas, trapping 72 times the amount of heat per molecule in the short term compared to carbon dioxide but “burning” to CO2 in the atmosphere in about a decade. We suspect large stores of methane, known as methane hydrates, lie frozen at the bottom of ever-warming oceans, particularly in the Arctic. On land, decomposing organic matter in thawing permafrost is another potential methane source.

I’ve modeled what happens if that methane is released continually for several decades: The atmospheric impact from methane itself only persists for about a decade beyond the methane release, whereas the extra CO2 in the atmosphere persists throughout the simulation of 100,000 years.

The possibility of a catastrophic release is more remote, but it’s a subject that, as journalists say, has legs. A submarine landslide might release a gigaton of carbon as methane, but the effect of that would be small, about equal in magnitude to – but opposite in effect of – a volcanic eruption. Detectable, perhaps, but not the end of humankind as a species.

So what could happen to methane in the Arctic?

The methane bubbles coming from the Siberian shelf are part of a system that takes centuries to respond to changes in temperature. The methane from the Arctic lakes is also potentially part of a new, enhanced, chronic methane release to the atmosphere. Neither of them could release a catastrophic amount of methane – hundreds of gigatons – within a short time-frame of a few years or less. There isn’t some huge bubble of methane waiting to erupt as soon as its roof melts.

And so far, the sources of methane from high latitudes are small, relative to the big players: wetlands in warmer climes and human emissions. It is very difficult to know whether the bubbles are a brand-new methane source caused by global warming, or a response to warming that has happened over the past 100 years, or whether plumes like this happen all the time. In any event, it doesn’t matter very much unless they get 10 or 100 times larger, because high-latitude sources are small compared to the tropics.

So maybe by century’s end, perhaps 2,000 gigatons of carbon could be released into the atmosphere by humans burning fossil fuels and other activities under some sort of business-as-usual scenario. And we might see another 1,000 gigatons of carbon from soil and methane hydrate release, as a worst case.

Can we get some sort of a doomsday, runaway greenhouse effect scenario from that?

I tweaked my models to try. If the methane hydrates released too much carbon, say two carbons from hydrates for every one carbon from fossil fuels, on a time scale that was too fast – say 1,000 years instead of 10,000 years – the system could run away.  But the fact that ice core and sediment records do not seem full of methane spikes makes it seem like the real world is not as sensitive as we were able to set the model up to be.

On the other hand, the deep ocean could ultimately – after a thousand years or so – warm up by several degrees in the business-as-usual scenario, making it warmer than it has been in millions of years. It takes millions of years to grow the hydrates; they have had time to grow in response to Earth’s relative cold of the past 10 million years or so. Also, the climate is very sensitive to changes in CO2 when its concentration is low, as it is today relative to what it was 50 million of years ago.

In short, if there was ever a good time to provoke a hydrate meltdown it would be now. But “now” is in a geological sense, thousands of years into the future, not really “now” in a human sense. Thawing methane hydrates in the ocean and permafrost peats could be a significant multiplier of the long tail of the carbon dioxide, but it will probably not be a huge player in climate change in the coming century.

The real point of no return in this adventure is our ongoing release of carbon dioxide.

The only way back to a natural climate in anything like our lifetimes would be to physically extract carbon dioxide from the atmosphere. The carbon dioxide that has been absorbed into the oceans would degas back to the atmosphere to some extent, so we’d have to clean that up too.

And if methane hydrates or permafrost peats contributed some extra carbon into the mix, that would also have to be part of the bargain, like paying interest on a loan.