Pan-Cryosphere Feedbacks: Albedo, Permafrost Melt & Sea Level Rise

Some impacts of rapid climate change in the cryosphere occur across the different regions, with contributions from each region.


When highly reflective snow and ice surfaces melt away, they reveal darker land or ocean surfaces that absorb more of the sun’s energy.  This process becomes especially apparent at the edges of mountain glaciers (though extremely debris-covered glaciers seem to show slower rates of loss due an insulation effect when the debris field is sufficiently thick).  The result is enhanced warming of the Earth’s surface and the air above it.  The impact of potentially greater cloud cover in these regions remains a matter of some debate, since clouds both insulate and warm the surface — and also reflect sunlight.

A growing body of evidence nevertheless indicates that albedo-associated warming is happening over the Arctic as the extent of sea ice decreases and snow cover retreats earlier in the spring (Lawrence et al. 2008).  Recent satellite observations of reflectivity (Riihelä et al. 2013) have concluded that albedo during the summer months has decreased in every month expect May (when cover remains thick) over the past 28 years. The highest albedo loss has occurred in August, (the month of the highest sea ice melt rates), with a decrease in reflectivity of about 3 percent per decade.  The loss of albedo also speeds processes related to permafrost and sea-level rise.


Temperatures in some parts of the Arctic permafrost have risen by up to 2 degrees over the past 30 years, faster than surface air temperatures.  In AR5, the IPCC noted with “high confidence” that permafrost temperatures have increased in most regions since the early 1980s, with observed permafrost warming of up to 3°C in parts of Northern Alaska, and up to 2°C in parts of the European Russia, with a considerable reduction in thickness and extent of permafrost in both regions.

The SWIPA report (2011) attributed much of this permafrost temperature rise to an 18-percent decrease in snow cover since the 1960s. Although most permafrost exists in the Arctic region, some occurs in most alpine systems and a large area exists on the Tibetan Plateau.  Globally, permafrost together with deposits in frozen near-shore seabeds are thought to hold about 1,700 Gt of carbon, compared to 850 Gt of carbon currently in the Earth’s atmosphere.  Release of even a portion of this carbon into the atmosphere could drastically compound the challenge already presented from anthropogenic sources, potentially wiping out any hard-won mitigation gains.  In regions such as Siberia, Alaska, and Tibet, loss of permafrost threatens infrastructure from homes, roads, and trains to oil and gas pipelines that may leak when placed under stress.

Permafrost carbon is released as CO2 under dry conditions, but some portion of the release will occur as methane under wet conditions (i.e., swamplands or methane hydrates coming from coastal seabeds).  Permafrost scientists estimate that release of just one percent of stored carbon in the form of methane will double current rates of warming due to methane’s more powerful near-term forcing effects.  About 12 percent, or 190 Gt of permafrost carbon, is stored in the upper 30 cm of permafrost layers considered most vulnerable to permanent melting (Zimov et al. 2006).  The IPCC estimated in AR5 that anywhere between 50-250 Gt carbon (5-30 percent of current atmospheric carbon) could be released by the end of this century; great uncertainty surrounds even this wide range, and it does not include potential releases of near-shore methane hydrates.

Some of the more dramatic observations in the past few years have involved the release of large bubbles of methane hydrates off the coast of Siberia in 2010 and 2011 (Shakhova, Alekseev, and Semiletov, 2010).  Some arctic methane researchers estimate that 50 Gt of carbon could be released in the east Siberian sea in the coming few decades, and cannot rule out this occurring in very brief timeframes, as a “pulse” over a few years.  Siberian shelf methane is essentially flooded permafrost in shallow waters, and may be highly sensitive to warmer waters arising from lower sea ice extent in the past decade (Shakhova et al. 2010).  Recent modeling (Whiteman et al. 2013)  estimates that such a  release from coastal seabeds, or emission from land permafrost along the same  scale (attributed in part to loss of sea ice albedo, per the above) could raise Arctic temperatures by 0.6°C by 2050.  This could add perhaps $60 trillion to the cost of adaptation (mostly in developing countries) by bringing forward the date when the globe exceeds 2°C of warming over pre-industrial levels to as early as 2035-40.  Keeping as much carbon as possible within the permafrost in the near-term, by maintaining lower temperatures in the Arctic and alpine permafrost regions, is therefore an issue of global importance.

Sea-level Rise

A rise in the level of the world’s oceans occurs from melting land ice (not sea ice), as well as thermal expansion as overall global temperature rises: warmer water takes up more space than colder water.  Various factors such as the earth’s spin, ocean currents, and gravity also cause sea-level rise to be non-uniform across the globe. The greatest relative increase is expected to be near the equator (especially Western Australia, Oceania, and small atolls and islands, including Hawaii and Micronesia) (Spada et al. 2013).   Recent observations have noted accelerated sea-level rise in the northeastern United States at 3-4 times that of the global mean (Sallenger et al. 2012).

The 2011 SWIPA report revised earlier estimates of sea-level rise based on observations of accelerated ice loss from Greenland and land glaciers; these are factors the previous IPCC Assessment Report (AR4, 2007) could not take into account.  SWIPA noted that accelerating melt from Arctic glaciers and ice caps, at approximately 40 percent of the total sea-level rise, was contributing much more than previously thought.  SWIPA  therefore revised AR4 estimates upward to between 0.9-1.6 meters in sea-level rise by 2100 (SWIPA 2011).

In its recently-released AR5 report, the IPCC increased its estimates from 2007, projecting about 0.5-1meter sea-level rise by 2100; this is still lower than SWIPA.  This is primarily because SWIPA and similar estimates rely on more empirical approaches, whereas the IPCC ultimately found such approaches not sufficiently tested to use in its latest assessment. [1]

Neither SWIPA nor IPCC estimates take into account the possibility of rapid disintegration (dynamical or non-linear ice discharge) of the West Antarctic Ice Sheet (WAIS).  Like all sea-based ice sheets, it is inherently unstable and subject to rapid changes, as seen in the collapse of the far smaller Larsen B Ice Shelf on the Antarctica Peninsula, which lost over 2,500 km2 (1,000 square miles) in the space of a few days in March 2002.  Although an increase in discharge from West Antarctica has been observed in recent years, too little is understood about the processes around the possible disintegration of the approximately 2-million km2 WAIS to include in current projections.  Such disintegration has, however, fairly clearly occurred in the geologic past (Pollard and DeConto 2009).  The WAIS could therefore contribute significant global sea level rise — or very little — over the next century (Bamber et al. 2009).  A total disintegration would raise global sea level from 3.3-5 meters, but in highly uncertain time spans of 100-1000 years.[2]  AR5 did note that such a collapse could result in up to “several” decimeters in additional sea-level rise during this century.

Recent studies have indicated that sea levels may have peaked at 4-9 meters higher[3] than today’s levels in the Eemian (125,000 years ago), the geologic period closest to current temperatures and CO2 levels[4].  Much of that sea-level rise appears to have come from West Antarctica, rather than Greenland as believed earlier (Bamber and Aspinal 2013).

In the case of both Antarctica and Greenland, some research indicates that today’s melting could set in motion processes that may prove difficult to stop, even with stable or falling temperatures globally by the end of this century. Some simulations show a loss of the Greenland ice sheet at ranges within the current 400ppm of CO2 (Stone et al. 2010).  AR5 noted with high confidence that some threshold exists beyond which a near-total loss of the Greenland ice sheet would occur, which could result in up to six meters of sea-level rise occurring over a 1000-year period.  AR5 set that threshold between one degree (which has already passed) and four degrees, but with lower confidence at the lower temperature ranges.

Sea-level rise thus may occur slowly, but failure to slow rapid warming in these regions soon enough may commit us to future sea-level rise that would threaten many of the world’s largest population centers — especially in developing countries.  This threat comes not just from sea-level rise but also from sea-level rise combined with extreme events, including offshore winds and tidal surge (such as that of Hurricane Sandy in 2012). These events could lead to enormous and costly infrastructure damage.

[1] AR5’s 2013 figure took into account more factors than AR4, but chose not to include others included by SWIPA due to continuing difficulty in estimating their scale of contribution.  AR5’s estimate should therefore not be seen as a “lower” estimate but rather one that chooses not to include certain elements until they can be more precisely quantified.

[2] O’Reilly, Oreskes, and Oppenheimer (2012). The Rapid Disintegration of Projections: the West Antarctic Ice Sheet and the Intergovernmental Panel on Climate Change. Social Studies of Science. 42(5):709-731.

[3] Dutton and Lambeck (2012). Ice Volume and Sea Level During the Last Interglacial.  Science 337 (6091):216-219

[4] Today’s CO2 levels over 400ppm are now actually higher than in the Eemian.