Open Mind

Methane North and South

August 24, 2009 · 25 Comments

It’s clear that atmospheric methane is on the increase after nearly a decade of stability. We’ve also seen that part of the reason may be the decomposition of methane hydrates on the ocean floor, triggered by the warming of ocean water. As Hank Roberts points out, RealClimate has addressed this issue, highlighting that

It takes decades to centuries to warm up the water 1000 meters down in the ocean, and centuries more to diffuse that heat down into the sediment where the base of the stability zone is. The Arctic Ocean may be a special case, because of the shallower stability zone due to the colder water column, and because warming is expected to be more intense in high latitudes.

Which raises the question, is there a difference between the hemispheres in the amount, and timing, of methane increase in the atmosphere? Rigby et al. (2008) noted that both the timing and magnitude of methane increase was similar for the entire globe. Let’s use more stations, and slightly longer time series, and restrict our comparison to the far north vs. the far south. The World Data Center for Greenhouse Gases has 9 reasonably complete monthly-average time series for stations north of 60 deg.N latitude, and 12 such time series for stations south of 40 deg.S latitude.

The southern stations show remarkably similar numbers since the turn of the millenium:


We can remove the annual cycle from each time series to generate anomaly data (I’ve also added a smoothed curve):


The smoothed curve gives a good picture of the far-southern change, but if we want the most precise timing we should simply take the average anomaly for all the available stations:


From this we see that the recent increase probably started in December 2006. I say “probably” because the increase from Nov. to Dec. 2006 is not so big as to be outside the bounds of normal variation, but it is the start of a long string of months with steady increase over the previous month.

The northern stations show a lot more difference among themselves:


Nonetheless we can still remove the annual cycle (and add a smooth curve):


Better yet we can, just as for the south, take the average anomaly for all the time series:


Once again the increase seems to start in December 2006, but the situation is even more uncertain than with southern data. The northern increase is far less regular and it’s by no means clear whether the upswing from Nov. to Dec. 2006 is the start of the recent trend or simply a return to previous levels. One could argue that the genuine recent trend doesn’t start until June or August of 2007. Also, the northern increase is probably larger but this too is uncertain. The one thing that’s undeniable is that the noise level is greater in the north.

We can plot the average anomaly for northern and southern stations on the same plot:


In my opinion this supports the idea that atmospheric methane concentration started increasing at both hemispheric extremes at about the same time (December 2006) and that the increase has been greater in the north than in the south.

Simultaneity of the increase argues against the source being something peculiar to the Arctic. It could be that the gas hydrate stability zone is warming up along the continental shelves in the southern hemisphere (especially the far south) just as it is in the north. It’s certainly possible that the methane increase is from some other cause, like melting permafrost — but it seems to me that would also require that it start in the artic because there’s so little permafrost area in the southern hemisphere outside of Antarctica (which isn’t melting yet).

All of which goes to show that there’s a great deal about the increasing methane signal that’s unknown, and there’s even more that’s unknown to me.

Categories: Global Warming

25 responses so far ↓

  • climatesight // August 24, 2009 at 5:37 pm | Reply

    I need some help.

    I am not a scientist, and have only been interested in this issue for a few years. I’m fine to point people in the right direction for broad concepts. But I don’t even know any calculus yet.

    I have a commenter who knows a heck of a lot about Steve McIntyre and the Hockey Stick controversy. He’s got a complicated chain of logic and citations which supposedly show that every 1000-year temperature graph ever used by the IPCC is flawed, when the flaws are taken out (specifically bristlecone pine data) the conclusion falls apart, and this has been suppressed by the IPCC which proves they have an agenda.

    I’ve been holding up okay until now. But now I really need someone who knows their climate science well – either to help me out or to (preferrably) take over. I’m not the right person to be taking part in this debate. I’m not a scientist.

    The thread starts here – – and really gets into the specifics around here –

    Any takers?

  • Marcus // August 24, 2009 at 5:48 pm | Reply

    A talk I went to listed five major explanations for historical methane changes:

    Anthropogenic emissions, hydroxyl radical changes, biomass burning (fires), wetlands emissions, and Arctic emissions (permafrost/hydrates).

    The speaker attributed the 2002-2003 methane spike to biomass burning in northern latitudes. 1998 was attributed to el Nino effects in the tropics.

    For 2007-2008 there was still uncertainty: for example, the lack of CO growth argues against biomass burning as a source, but the correlation with ethane argues _for_ biomass burning. 2007 was warm & wet in the Arctic, 2007-8 was warm & wet in the tropics. d13C was “lighter” which according to the speaker would imply wetlands emissions (presumably any biological source would have the same signature, though).

  • Zeke Hausfather // August 24, 2009 at 6:44 pm | Reply

    Interesting. I agree that the lack of a unique Northern Hemispheric signature does somewhat argue against a melting permafrost or clathrate driver, but leaves us with the vexing question of where a Southern Hemispheric methane pulse could have originated (given the lack of land area).

    Anyone have an idea of the general mixing rate of CH4 in the atmosphere? If we are talking about a relatively short period, than perhaps the recent increase in the summer hemisphere is simply lagging a few months behind the northern hemispheric source?

    It also looks like only a single station in the north has data post-2008 so far, which adds considerably to the uncertainty for the latter dramatic oscillations.

  • Hank Roberts // August 24, 2009 at 7:29 pm | Reply

    Wait, the _atmosphere_ doesn’t mix rapidly across the Equator, but the _oceans_ do via the thermohaline circulation.

    And most of the methane that is released goes into solution in the water, the bubbles don’t reach the surface directly.

  • Hank Roberts // August 24, 2009 at 7:34 pm | Reply

    Also, the methane may have already been cycled through biological processes:

    “… The methane plumes observed in the surface and subsurface water differed from each other, suggesting that they are generated independently. The subsurface water in summertime contained methane that was released from sediments during winter, and oxidized over time, leaving the residual methane 13C-enriched. The surface water, on the other hand, contained recently produced, 13C-depleted methane. We propose that methane in situ production occurs during the summer phytoplankton bloom. The concentration of methane increases up to a certain threshold value, above which methane consumption begins. A methane production-removal cycle is established, which is reflected in the varying methane concentrations and δ13CCH4 values. DMSP and methane are inversely correlated suggesting that DMSP could be a potential substrate for the methylotrophic methanogenesis.”

  • rustneversleeps // August 24, 2009 at 7:50 pm | Reply

    Can anyone enlighten me on the following comment, which I have not encountered elsewhere w.r.t. permafrost and clathrate sources for methane:

    “An example of these effects is the melting of the permafrost, which we know is happening because we can isotopically date the helium being released. This He has not been released from the permafrost in at least 40,000 years.” (A good article by the way…)

    The writer is Nate Lewis and the publication is the (Materials Research Society) MRS Bulletin. I’m inclined to think he has researched this… but I’ve never seen it elsewhere in my amateur readings…

    Can we, in fact, isotopically identify the source of increased methane concentrations (as we can with C12 & C13 for fossil-fueled based CO2)? Pointers if so?

  • Marcus // August 24, 2009 at 7:57 pm | Reply

    Well, the lack of a unique NH signature argues against melting permafrost being the _only_ cause of a methane signal, but it is possible to have multiple coincident clauses.

    Do we have sufficient latitudinal resolution to determine if the southern hemisphere response is primarily a high latitude response or a more equatorial response? Looking at the Rigby paper, it looks like the south pole and cape grim have a smaller trend than the more northerly SH measurement stations (and the text states that other high latitude sources behave similarly). (whereas in the north the southerly Mauna Loa observatory has a smaller trend than the more northerly stations… though the Barbados station does not, which may moot that observation) (also, not sure if size of trend is as good a measure of localization as actual temporal lag).

    Rigby et al. theorize an increase in emissions in the north, and a decrease in OH in the south. I’ll note that a Brazilian/sub-saharan Africa wetlands source might be another possible explanation.

    FYI: Seinfeld & Pandis (1998) estimate mixing times in the atmosphere as follows:

    intra-hemisphere: 1-2 months
    inter-hemisphere (trop): 1-2 years
    trop/strat: 1-2 years
    inter-hemisphere (strat): 3-6 years

    (the ITCZ slows the inter-hemisphere mixing time a lot, apparently)

  • David B. Benson // August 24, 2009 at 8:00 pm | Reply

    Generally thought to require two years for well-mixing across both hemispheres.

    Consider the continental shelves off the Antarctic Penninsula and Tierra del Fuego?

  • John Mashey // August 24, 2009 at 8:48 pm | Reply

    Re: Nate Lewis
    I don’t know about that specific paper ( and being at Hot Chips with only an iPhone…)

    but I’ve heard him speak here last year, at GCEP meeting,. He seems very sharp, so indeed he’d be worth reading.

  • Jim Eager // August 24, 2009 at 9:06 pm | Reply

    climatesight, see my response your posting of this question at Deltoid.

  • Phil Scadden // August 24, 2009 at 9:12 pm | Reply

    Hank – “most of the methane goes into solution”.

    I’m a little puzzled by this. My first thought would be that methane in seawater would oxidize pretty rapidly to CO2. Is there any data on methane concentrations in seawater?

    I would also note that southern hemisphere stations appear to be coastal. Not so for all of the northern. Could this be source of greater variability between stations in the north?

  • Marcus // August 24, 2009 at 9:24 pm | Reply

    Huh. Once upon a time I considered working in Nate Lewis’ lab – he was doing neat stuff on “electronic noses” using resistance changes in a set of polymers to develop fingerprints for different gaseous materials. I knew he did solar cell inorganic chemistry too… branching out into helium isotope dating would something different for him (though now that I look at the talk, it looks like he is summarizing other research) (as I noted above, d13C is used for biological signatures… presumably d14C could be used for dating, though the fossil fuel signal would likely dwarf the permafrost signal. I’d guess that helium isotope dating would have to be done at the permafrost site, capturing bubbles or something – I wouldn’t think the signal would be strong enough to differentiate increased permafrost melt from other sources of aged helium on a whole-atmosphere scale)

  • David B. Benson // August 24, 2009 at 9:34 pm | Reply

    Hank Roberts // August 24, 2009 at 7:29 pm — THC is very slow. I gave estimates for various oceans on another thread so won’t repeat those values here, but think several centuries.

  • Hank Roberts // August 24, 2009 at 10:11 pm | Reply

    > concentrations

    This is fairly old, and much cited:

    Nature 327, 226 – 229 (21 May 1987); doi:10.1038/327226a0

    Methane oxidation and methane fluxes in the ocean surface layer and deep anoxic waters

    B. B. Ward*, K. A. Kilpatrick*, P. C. Novelli† & M. I. Scranton†

    Methane is supersaturated in sea water, and is typically at its maximum concentration in near-surface waters, which could support a significant sea-air flux. The magnitude and variability of the flux depends on the mechanisms which produce and consume methane in sea water. Here, we compare measured biological oxidation rates of methane with the diffusional fluxes computed from concentration gradients in the surface layer of the ocean, and show that oxidation of methane in sea water is a mechanism which modulates the flux of methane from marine waters to the atmosphere. Methane fluxes and oxidation rates were investigated in surface waters, at the oxic/anoxic interface and in deep anoxic waters of the Cariaco Basin. Measured oxidation rates were equivalent to 5% of the methane flux into oxygenated waters from the methane-rich deep waters and 10% of the flux into surface waters from the subsurface methane maximum. Thus oxidation was not sufficient to prevent a net sea-air flux. The total methane oxidation rate in the basin amounted to 1.5% of total primary production in the surface layer. Only a small fraction of oceanic primary production would be required to cycle through the methane pool to support the global atmospheric flux from the ocean.

  • Aaron Lewis // August 24, 2009 at 10:56 pm | Reply

    The assumption/rule of thumb that deep water warms slowly should be kept in the same bin as the assumption/rule of thumb that big ice warms and moves slowly. A moment’s thought reminds us that there have been abrupt changes in ocean currents, many times before. All that is required to change ocean currents is an abrupt stimulus, such as the collapse of an ice sheet or change in the local circulation. One example showing how very fast ocean currents can change under the influence of recent climate forcing is: Acceleration of Jakobshavn Isbræ triggered by warm subsurface ocean waters (PDF at

    We now have several reports of recently warming water in the Arctic, at depths known to hold clathrate deposits. We seem to have multiple reports of metahane release from sea beds in this depth range.

    However, we need not worry, because they are going to go take air samples : .

    On the one hand what do we do if they come back “high methane”? On the other, what if this one set of samples come back “moderate methane”? Do we stop worrying if this one set of samples comes back “low methane”?

    What is the action? What is the decision rule?

  • Hank Roberts // August 25, 2009 at 1:50 am | Reply

    “… Scientists are suddenly more interested in thermokarsts because, well, there simply seem to be more of them. From photographs taken from a low-altitude flyover of the entire state in the early 1980s and data gathered in 2006, scientists believe the number of thermokarsts in the Alaskan landscape has doubled in that time. “This is a natural phenomena, but it appears to be accelerated by warming in the Arctic,” Bowden notes….”

    Hat tip to
    H.E. Taylor’s collection, which points to:

    Page down to get to the article, the first “page” is all article links

  • Hank Roberts // August 25, 2009 at 3:07 pm | Reply

    Following that 1987 paper forward, it’s much cited.
    The biological response will surprise anyone who hasn’t looked at the number of microbes per gram of sea water and how fast selection pressure changes the populations. Look into this, whatever happens is going to be dramatic.

    We know almost nothing:

    —excerpt follows—
    No comprehensive studies dedicated to understanding the aerobic methanotrophic biofilter have been conducted and no authoritative works are available on the topic. Several studies of limited applicability have been conducted, and these studies guide our present understanding of this process. A recent review of oceanic methane by Reeburgh (2007) highlights the current understanding of the marine methane cycle and represents the state-of-the-art for marine methanotrophy. An important distinction for the proposed research is separating anaerobic processes of methane production and anaerobic oxidation of methane from the aerobic process of methanotrophy (microbial methane oxidation with oxygen as terminal electron acceptor). The anaerobic processes occur typically within seafloor sediments or in waters of highly restricted
    basins such as the Black Sea and Cariaco Basin (KESSLER et al., 2006a; KESSLER et al., 2006b; REEBURGH et al., 1991). AOM in particular has been the subject of several hundred investigations over the past decade, including several major efforts funded by the European Union. The anaerobic processes often dominate their local environment and as a result are relatively easy to investigate in-situ. In contrast, aerobic methanotrophy occurs broadly in oceanic water columns, and the knowledge base is not as well established as methanotrophy
    occurs as a trace process superimposed over numerous other microbial processes such as
    heterotrophic respiration and photosynthesis. An example of our limited knowledge is that we
    do not currently know the identity of organisms comprising the aerobic methanotrophic biofilter.

  • Simon D // August 25, 2009 at 6:44 pm | Reply

    Attribution is a real challenge with methane [and other more reactive greenhouse gases]. Even if there was a clear hemispheric difference in the methane trends, that would not be enough evidence to conclude that the source had increased (the OH sink could have decreased), let alone the specifics of the source. It’s an important, interesting and frustrating area for research.

  • Hank Roberts // August 26, 2009 at 4:29 am | Reply

    Couple good posts over at
    # Time to unleash seabed methane?
    # China cuts methane emissions from rice fields

  • Didactylos // August 26, 2009 at 12:02 pm | Reply

    I think including data for 2008 based on a single northern station is rather confusing, given the variability in the northern data. The single source can’t possibly be representative, and it causes an odd squiggle at the right side of the graph. It gives the illusion of sudden recent changes (both up and down).

  • Mark // August 26, 2009 at 1:46 pm | Reply

    didactylos, that is why the HadCRUT data doesn’t cover the polar regions whereas GISS does because they think they can get around those issues.

  • Hank Roberts // August 28, 2009 at 4:30 pm | Reply

    hat tip to:


    “… “Our survey was designed to work out how much methane might be released by future ocean warming; we did not expect to discover such strong evidence that this process has already started,” said Professor Tim Minshull of the National Oceanography Centre at the University of Southampton in the UK….”

  • jyyh // September 11, 2009 at 6:54 am | Reply

    I for one would be interested if the effect of methane releases on temperatures can be directly measurable in situ (that is the earth)? This might be possible as the releases are somewhat local on land, and cease during winters. The local wind conditions in permafrost areas when the releases start should be accounted for and the temperature differences around the area could be compared. Thermometers should probably be accompanied with methane analyzers. Likely though the varying weather conditions make this sort of validation (as any more would be needed) of the greenhouse effect extremely difficult.

  • Hank Roberts // October 23, 2009 at 5:21 am | Reply

    (anyone who has paid for the key to the Nature paywall, have a look):

    • Sekerob // October 23, 2009 at 2:53 pm | Reply

      Hank, // October 23, 2009 at 5:21 am not a big surprise. Hydroelectric power-plants are not deemed GHG neutral… the scale of TGD just making it bluntly obvious. There’s allot of slipping and sliding already observed there. The strata not up to the pressures either so a friend Geo-scientist pointed me out 15 years ago.

Leave a Comment