The sources of atmospheric methane are many and varied:
Some are natural (i.e., wetlands) and some are anthropogenic. Methane levels have doubled since the pre-industrial era, so the anthropogenic emissions, which have increased, are having a significant impact. That doubling, to review, adds about half again to the forcing of the excess CO2 in the atmosphere (while methane is a more potent greenhouse gas than CO2, its concentration in the atmosphere is much smaller.) Anthropogenic emissions break down as follows:
Much of the excitement about methane, and the source of the recent flurry of interest in methane (see Climate Progress, Dot Earth, and Realclimate) is the stored methane around the world, locked up in permafrost or deep under the ocean.
Natural processes have created carbon-based deposits of many kinds; coal, oil, "natural gas" (which is methane) and others. Methane hydrates, which I promised to explain last time, are deposits in cold water and usually under great pressure, composed of methane molecules (original released by aquatic bacteria) trapped in a matrix of frozen H2O (methane clathrates are the same thing -- a "clathrate" is just the generic term for anything frozen inside a matrix of anything -- a "hydrate" is a clathrate in which water is the matrix. Methane hydrates is the more usual term today, but "clathrates" come up, as we see below.")
The reason these deposits get people excited is that, since obviously they depend on the cold in order to stay frozen, that they may be destabilized by climate change. Estimates of how much of this stuff there is vary widely, but the lowest estimates are very high; 500 to 2,500 billion tons.
Suppose 1% of that leaked out per year. How bad would that be? Taking a middle estimate of 1,500 billion tons, that would be a release of 15 billion tons. That would approximately double the amount of methane in the atmosphere in a single year, doubling the forcing of methane (which has already doubled once). This would continue in the following year, and the year after that. Assuming no acceleration, and assuming methane decays after ten years, on average, we would end up with 10-15 times at much methane in the atmosphere as we have now, which would lead to an almost instantaneous warming of 1-2 degrees C, plus an additional 1-2 degrees as the long-term feedbacks kick in.
Unfortunately, that methane does not just vanish as it decays -- it turns into CO2 and stratosphere water vapor (a powerful greenhouse gas), such that after ten years, in addition to the warming from the methane, we've also added water vapor and significant amounts of CO2 (roughly 50% over and above our anthropogenic emissions).
The above is probably much too simple, as it ignores the local effects (the most likely location for massive methane releases is in the Arctic, which could thereby lose more sea ice, changing the earth's albedo) and also ignores methane sinks. However, it is enough to give a flavor of why people are concerned.
Some are more concerned than others, pointing, for example, to mass extinctions in which the release of frozen methane deposits may have been the culprit (the clathrate gun hypothesis). On the other hand, climate scientists point out that we really don't have evidence or a convincing mechanism for the sudden release of even 1% of these deposits -- yet. All of which leads us to the reason this debate is heating up right now: scientists who went to the Arctic to study the possibility that methane frozen in undersea permafrost might someday be release were shock to discover hundreds of methane plumes across the Arctic seabed. Their findings quickly settling the question of "Will this happen?" and made the key questions "How long, how fast, how bad?"
I'll talk about those key questions in part three.