The Big Thaw

The blogosphere is buzzing with discussion of new research soon to be published in Geophysical Research Letters, Accelerated Arctic land warming and permafrost degradation during rapid sea ice loss (Lawrence et al. 2008, GRL, in press, scheduled for publication tomorrow), a collaboration between researchers from NCAR (National Center for Atmospheric Research) and NSIDC (National Snow and Ice Data Center). Their research indicates that, in large part due to the rapid reduction of summer sea ice in the arctic, land areas in the far north will warm much faster than the rest of the globe. This would lead to rapid melting of permafrost, which in turn could release massive stores of CO2 trapped in the permafrost.

Last summer witnessed a truly stunning reduction of arctic sea ice:

The ice loss was due to a combination of two factors. First, the arctic is indeed warming faster than the rest of the planet. Second, prevailing wind patterns shifted so that polar ice (which is floating on the arctic ocean) drifted to lower latitudes, making it even more vulnerable to melting. The result was a reduction in arctic sea ice for which the adjective “stunning” is actually rather mild.

The latest monthly sea ice extent data from NSIDC shows that arctic sea ice recovered this winter, and as of May 2008 was above last year’s levels:

However, the most recent daily observations indicate a reduction in late May/early June, so that at present sea ice extent is slightly below last year’s level:

Whether or not this summer’s ice loss will approach or exceed last year’s record-breaking values is an open question. On one hand, the Arctic has indeed been warming faster than the rest of the globe, and a lot of multi-year ice was lost last year, so that far more of this year’s polar ice cap is 1st-year ice, which is much thinner and more vulnerable to melting than multi-year ice; these factors would indicate that we may indeed break the record again. On the other hand, we may not see the same wind patterns as last year, so the ice pack may not be blown about like it was; this indicates that we may not break the record this year. Only time will tell.

But the new research isn’t concerned with single exceptional years as much as it is with the coming decades, and this is presaged by the long-term trend. Then trend has been inexorably downward, and the loss of Arctic sea ice accelerated about mid-2001:

From late 1978 (the beginning of satellite observations) until 2001.4, Arctic sea ice extent declined at a rate of about 37,000 km^2/yr, but from 2001.4 to the present it’s been declining far more rapidly, at a rate of 131,000 km^2/yr. The rapid loss of Arctic sea ice, especially during summer and autumn, can be expected to continue. And it’s this loss of Arctic summer/autumn sea ice that is at the root of forecasts for dramatically rapid warming of north polar land areas, and the consequent rapid melting of permafrost areas.

The research uses the CCSM3 climate model to study both the progress of sea ice change and the changes in Arctic climate. Almost all model runs show a “Rapid sea Ice Loss Event” (RILE) during which sea ice declines dramatically for a 5-10 year period. When sea ice declines, not only does the open water absorb more incoming solar energy (albedo change), but the open water can exchange heat between the warmer ocean and the colder air, dumping lots of heat in the Arctic region. This warms the Arctic land climate dramatically, both air and ground temperature rising, and the rising ground temperature leads to rapid degradation of permafrost. This is summed up in the abstract:

Coupled climate models and recent observational evidence suggest that Arctic sea ice may undergo abrupt periods of loss within fifty years. Here, we evaluate the impact of rapid sea ice loss on terrestrial Arctic climate and ground thermal state in the Community Climate System Model. We find that western Arctic land warming trends during rapid sea ice loss are 3.5 times greater than secular 21st century climate-change trends outside these periods. The accelerated warming signal extends up to 1500km inland and is apparent throughout most of the year, peaking in autumn. Idealized experiments using the Community Land Model, with improved permafrost dynamics, indicate that an accelerated warming period substantially increases ground heat accumulation – the earlier the event the greater the long-term impact. For warm permafrost, enhanced heat accumulation can lead to rapid degradation. For colder ground, heat accumulation preconditions permafrost for earlier and/or more rapid degradation under continued warming.

The Arctic is already warming twice as fast as the rest of the globe, but this is greatly amplified during RILE events. Here’s the simulated change in September Arctic sea ice extent, and October-December Arctic surface air temperature, during a typical RILE (the solid line shows sea ice extent, the dashed line surface air temperature)

The additional warming due to sea ice loss is greatest in exactly the regions where permafrost exists presently:

According to the study, “We find that the secular 21st century land warming trend is augmented by a factor of 3.5 during RILEs, which is likely to have adverse impacts on permafrost.” In fact they conclude that permafrost is greatly affected, “may trigger rapid degradation of currently warm permafrost and precondition colder permafrost for subsequent degradation under continued warming. This sea ice loss – land warming relationship may be immediately relevant given the record low sea ice extent in 2007.”.

To test whether the rapid Arctic warming events are truly the result of sea ice loss, the authors compared Arctic climate from model runs in which climate forcing factors were held constant except for the extent of Arctic sea ice:

Clearly the loss of sea ice, with its consequent albedo decrease and release of ocean heat to the atmosphere, has a profound impact on temperature changes in the Arctic.

Why should we care about the melting of Arctic permafrost? When permafrost soil melts, it releases frozen water which can discharge fresh water into the oceans. It also destabilized the ground, with important impacts on infrastructure in populated Arctic areas. And of course, the changes in temperature, solidity of the ground, and the water cycle cause significant changes for Arctic ecosystem.

But an even more dramatic consequence of permafrost melting may be in store for the next generation — or even for our generation. Permafrost soil holds about 200 to 800 Gt (gigatonne) of carbon, and its melting can release this frozen carbon store to the atmosphere. The carbon can be released as CO2, or with anaerobic bacterial processes it can be transformed to methane (an even more potent greenhouse gas). Considering that the entire atmosphere presently has about 770 Gt carbon (almost entirely CO2), the amount of additional carbon in permafrost is enough to double earth’s atmospheric carbon load. In fact it has been estimated that rapidly melting permafrost could add 1 to 4 Gt carbon release per year, as much as half the present emissions from human activity — a substantial increase in carbon emissions and a very unpleasant feedback on an already stressed global climate system.

The atmospheric release of a massive store of sequestered carbon is one possible hitherto-unanticipated feedback in the climate system, one which may quite literally be disastrous, not just for the Arctic but for the entire planet; CO2 is a well-mixed atmospheric gas and the extra CO2 from permafrost loss will reach to all corners of the globe. We simply can’t afford an extra 4 Gt/yr carbon emissions which will sustain long enough to double the atmospheric CO2 load. It’s an example of how our present actions may lead to unanticipated and very unpleasant consequences. It also emphasizes the importance of drastic reductions in human greenhouse-gas emissions now, not later. If we can eliminate our own climate-changing activities, we can delay and/or reduce, or possibly even avoid, a very nasty feedback in the carbon cycle due to permafrost. If we continue to drag our feet, sacrificing the long-term good for short-term comfort, we may well be headed for a future which is far worse than even the most dire predictions of the doomsayers.