[Vision2020] 9-5-2012: "Record Arctic Sea Ice Melt to Levels Unseen in Millennia"

[Ted Moffett]

This just posted article on the 2012 record Arctic sea ice decline, from
Skepticalscience.com, offers a superb summary of critical science on this
issue, noting at the bottom that several of the models used to predict
Arctic ice responses to anthropogenic climate change, rather than biased
towards large impacts, which is what many in the anthropogenic climate
change denialasphere insist, actually underestimated the observed rate of
Arctic sea ice decline we now see.  The very important graphs omitted from
the text below can be viewed at the following website:

http://www.skepticalscience.com/record-arctic-sea-ice-melt-to-levels-unseen-in-millennia.html
Record Arctic Sea Ice Melt to Levels Unseen in Millennia Posted on 5
September 2012 by dana1981

The record Arctic sea ice
decline<http://skepticalscience.com/why-arctic-sea-ice-shouldnt-leave-anyone-cold.html>this
year has predictably and deservedly received a fair amount of media
attention. Jonathan Leake of the *Sunday Times* recently penned an article
on the impending sea ice
record<http://www.skepticalscience.com/pics/Leake-ST-ice-article.jpg>.
The bulk of the article was quite good, but at the end succumbed to the
standard mainstream media practice of seeking "balance," thus including
some comments by John
Christy<http://www.skepticalscience.com/John_Christy_blog.htm>.
Christy has become very reliable for arguing that anything and everything
related to climate change probably just boils down to natural variability, as
he recently told US
Congress<http://www.skepticalscience.com/christy-once-again-misinforms-congress.html>was
the case with regards to the frequency of extreme weather events,
contrary
to the body of peer-reviewed scientific
literature<http://www.skepticalscience.com/extreme-weather-global-warming-intermediate.htm>
.

As we will see in this post, Christy once again misrepresented the body of
scientific literature with regards to Arctic sea ice extent in his efforts
to paint the Arctic sea ice death
spiral<http://www.skepticalscience.com/lessons-from-past-predictions-vinnikov-arctic-sea-ice.html>as
nothing out of the ordinary.
2012 vs. 1940

In Leake's article, Christy was paraphrased as saying that there is

"...anecdotal and other evidence suggesting similar melts from 1938-43 and
on other occasions."

Christy's comments to Leake via email slightly differed from Leake's
paraphrasing, as Christy claimed that evidence suggests summer melts during
1938-43 were "very low extent." This is a rather vague and subjective
statement - very low relative to what? Given the context, Leake
understandably appears to have assumed that Christy meant very low relative
to recent years, and perhaps he did, but it is also possible that he meant
'very low' relative to the early 20th Century, for example.

This begs the obvious question - in the scientific literature, how does
Arctic sea ice extent during the period 1938-43 compare to the rest of the
20th Century and current levels? One of the most widely used long-term
estimates of Arctic sea ice extent comes from Walsh and Chapman
(2001)<http://www.ingentaconnect.com/content/igsoc/agl/2001/00000033/00000001/art00071>,
whose data are available from the University of
Illinois<http://arctic.atmos.uiuc.edu/SEAICE/timeseries.1870-2008>(updated
through 2008). A description of the vast array of data used by
Walsh and Chapman is available via tamino
here<http://tamino.wordpress.com/2011/09/17/more-cherry-ice-from-joe-daleo/>,
and the data are plotted in Figure 1.

Clearly the extent of Arctic sea ice during 1938-43 was nowhere near as low
as current levels, based on these data. According to this reconstruction,
the minimum extent during that timeframe (9.8 million square kilometers in
1940) was higher than it has been at any time since 1979. In other words,
Arctic sea ice extent has been lower than it was in 1938-43 during the
entire satellite record, and the current average summer extent is
approximately 4.3 million square kilometers lower than the 1940 minimum.

It's true that according to this dataset, 1940 was a local minimum - the
lowest Arctic sea ice extent of the 20th Century up to that point, and a
minimum that was not repeated again for another 20 years. In that sense one
could argue that Arctic sea ice extent at least in 1940 was "very low"
compared to the early and mid 20th Century, but compared to the past 20
years it was actually very high. This is also clear from a visual
comparison of sea ice extents in 1938 and 2012 (Figure 2).

**Henry Larsen, who sailed the Arctic from 1922 to 1948 (including 12 times
surviving being stuck in the Arctic ice all
winter<http://www.vancouvermaritimemuseum.com/page216.htm>),
tried to sail the St. Roch through the Northwest
Passage<http://www.thecanadianencyclopedia.com/articles/northwest-passage>in
1940. The voyage took more than two years to complete, as the ship
struggled through the ice. Larsen gave this firsthand
account<http://www.skepticalscience.com/StRoch.html>of the state of
Arctic ice at the time:

"The three seasons of the short Arctic Summers from 1940-42 had been
extremely bad for navigation, the worst consecutive three I had experienced
as far as ice and weather conditions were concerned, and in my remaining
years in the Arctic I never saw their like. Without hesitation I would say
that most ships encountering the conditions we faced would have failed."

On the other side of the Arctic, Russian maritime operations using
icebreakers on the Northern Sea Route began in 1932 and give no evidence
for improving ice conditions in this period; rather the opposite, as 1937
and 1940 were noted for heavy ice in the Laptev
Sea<http://www.skepticalscience.com/pubs.aina.ucalgary.ca/arctic/Arctic33-1-3.pdf>
.
Greenland Temperature and Arctic Dominoes

So what was the basis of Christy's claim of "very low extent" in 1938-43?
Christy provided two references to support his assertions, Box et al.
(2009)<http://journals.ametsoc.org/doi/abs/10.1175/2009JCLI2816.1>and
Kobashi
et al. (2010) <http://escholarship.org/uc/item/1zx1v60n#page-1>. However,
neither of these papers involve reconstructions of Arctic sea ice extent;
rather, they deal with reconstructing Greenland temperatures, which are not
necessarily representative of Arctic temperature as a whole - after all,
the continent is covered in a large ice sheet, and the reconstructions are
from the summit of the Greenland ice cap, not at sea level.

We contacted Jason Box <http://www.meltfactor.org/blog/> (lead author of
Box et al. 2009), who also noted that local temperatures are not the only
factor at play in determining Arctic sea ice extent. During the past two
decades, Greenland temperatures have climbed rather steeply, surpassing
local temperatures in the
1930s<http://www.skepticalscience.com/Greenland_ice_sheet_summer_temperatures_highest_in_172_years.html>(Figure
3), and even
moreso over the whole
Arctic<http://www.skepticalscience.com/Prudent_Path_Polar.html>(Figure
4). These rising temperatures have been accompanied by unprecedented
Greenland ice sheet surface
melt<http://skepticalscience.com/unprecedented-greenland-ice-sheet-surface-melt.html>.
Greenland glaciers have declined (Bjørk et al.
2012<http://www.nature.com/ngeo/journal/v5/n6/full/ngeo1481.html>)
as have Greenland's ice shelves (Falkner et al.
2011<http://www.agu.org/pubs/crossref/2011/2011EO140001.shtml>).
For example, the Ward Hunt Ice
Shelf<http://earthobservatory.nasa.gov/Features/wardhunt/>,
which was a least 3,000 years old, split off in 2002 (Mueller et al.
2003<http://www.cen.ulaval.ca/warwickvincent/PDFfiles/175.pdf>,
England et al. 2008 <http://dro.dur.ac.uk/6877/1/6877.pdf>, and Antoniades
et al. 2011 <http://www.pnas.org/content/108/47/18899.full.pdf+html>).

All of this regional ice loss has decreased the local surface reflectivity
(albedo), causing the Arctic to absorb more solar radiation, and thus we
can expect that similar temperatures now will have a larger impact on sea
ice extent than in the past.

**In other words, there is a domino effect at play. Human-caused global
warming contributes to the summer Greenland warming (Figure 3), which
causes snow to melt
earlier<http://www.skepticalscience.com/record-snow-cover.htm>,
which causes decreased local albedo, which contributes to record Greeland
ice sheet decline, which further decreases local albedo, which in turn
contributes to the Arctic sea ice decline.
2012 vs. the More Distant Past

Christy also claimed that sea ice extent was low 1,000 years ago, as well
as in the more distant past. However, his reference for the claim of low
sea ice extent 1,000 years ago was again Kobashi et al. (2010), which as
noted above<http://www.skepticalscience.com/record-arctic-sea-ice-melt-to-levels-unseen-in-millennia.html#kobashi>,
dealt with Greenland temperatures rather than Arctic sea ice extent.

It's important to note that we expect the Arctic to have been cooling over
the past ~6,000 years due to the Earth's orbital cycles. Thus if we look
back far enough in the past, we can certainly find a period during which
the Arctic was hotter and Arctic sea ice extent was lower. However, this
actually contradicts John Christy's argument that the current sea ice
decline could be natural, because that long-term orbital forcing has not
reversed, and thus cannot account for the sudden and rapid Arctic warming
and concurrent sea ice decline.
<http://www.see.ed.ac.uk/~shs/Climate%20change/Data%20sources/Kaufman%20Schneider%20recent%20warming.pdf>

Kaufman et al. (2009)<http://www.see.ed.ac.uk/~shs/Climate%20change/Data%20sources/Kaufman%20Schneider%20recent%20warming.pdf>reconstructed
Arctic temperatures even further back in time than shown in Figure
5<http://www.skepticalscience.com/record-arctic-sea-ice-melt-to-levels-unseen-in-millennia.html#Fig5>,
and confirmed that the Arctic had been cooling for at least the past 2,000
years prior to the 20th Century, and found an Arctic temperature 'hockey
stick' <http://www.skepticalscience.com/Prudent_Path_Polar.html> (Figure 5).

Perhaps the authoritative paper on Arctic sea ice extent over the past
1,450 years is Kinnard et al.
(2011)<http://www.skepticalscience.com/Arctic-sea-ice-hockey-stick-melt-unprecedented-in-last-1450-years.html>,
which used a combination of Arctic ice core, tree ring, and lake sediment
data to reconstruct past Arctic conditions. The results are shown in Figure
6.

Based on the Kinnard results, Arctic sea ice extent is currently lower than
at any time in the past 1,450 years.

Polyak et al. (2010)<http://www.skepticalscience.com/bprc.osu.edu/geo/publications/polyak_etal_seaice_QSR_10.pdf>looked
at Arctic sea ice changes throughout geologic history and noted that
the current rate of loss appears to be more rapid than natural variability
can account for in the historical record.

"The current reduction in Arctic ice cover started in the late 19th
century, consistent with the rapidly warming climate, and became very
pronounced over the last three decades. This ice loss appears to be
unmatched over at least the last few thousand years and unexplainable by
any of the known natural variabilities."

Is the Sea Ice Decline Human-Caused?

The evidence above certainly suggests that humans have played a role in the
Arctic sea ice decline, but what does the scientific literature say on the
matter?

Vinnikov et al.
(1999)<http://www.skepticalscience.com/lessons-from-past-predictions-vinnikov-arctic-sea-ice.html>estimated
the probability that the Arctic sea ice decline could simply be
natural. The authors used very long control runs of both the Geophysical
Fluid Dynamics Laboratory (GFDL) and Hadley Centre climate models (5,000
years for the GFDL model) to assess the probability that the observed and
model-predicted trends in Arctic sea ice extent occur by chance as the
result of natural climate variability. They found that large trends in sea
ice extent only appeared over short time intervals in the control run, due
to natural variability alone. This suggests that natural variability will
not cause large long-term Arctic sea ice trends.

Updating this analysis using observational data through 2011 (not even
including the 2012 record low sea ice extent), the 32-year trend
(1979-2011) is -530 thousand square km per decade, and the 20-year trend is
-700 thousand square km per decade. Using the Vinnikov et al. results,
these trends both correspond to *probabilities of well under 0.1% of being
due solely to natural variability*.

Day et al. (2012) <http://iopscience.iop.org/1748-9326/7/3/034011> used
five climate models to try and quantify the contribution of natural
variations in Arctic sea ice changes. They found that between 5% and 30% of
the Arctic sea ice decline from 1979 to 2010 could be attributed to the
natural cycles of the Atlantic Multidecadal Oscillation (AMO) and Arctic
Oscillation (AO), and even less can be attributed to natural cycles since
1953, since these natural cycles tend to average out over longer timeframes
(as Vinnikov also
found<http://www.skepticalscience.com/record-arctic-sea-ice-melt-to-levels-unseen-in-millennia.html#large-short>
).

"despite increased observational uncertainty in the pre-satellite era, the
trend in [Arctic sea ice extent] over this longer period [1953–2010] is
more likely to be representative of the anthropogenically forced component."

Stroeve et al. (2011)<http://www.agu.org/pubs/crossref/2011/2010GL045662.shtml>noted
that in 2009-2010, the AO was in a state which should have resulted
in a large sea ice extent; the fact that 2010 was a year of relatively low
sea ice extent is indicative long-term human-caused sea ice decline.

"Based on relationships established in previous studies, the extreme
negative phase of the Arctic Oscillation (AO) that characterized winter of
2009/2010 should have favored retention of Arctic sea ice through the 2010
summer melt season. The September 2010 sea ice extent nevertheless ended up
as third lowest in the satellite record, behind 2007 and barely above 2008,
reinforcing the long-term downward trend."

Notz and Marotzke
(2012)<http://www.agu.org/journals/gl/gl1208/2012GL051094/>also found
very poor correlation between the AO and Pacific Decadal
Oscillation (PDO) and Arctic sea ice extent (yellow and green in Figure 7),
concluding:

"the available observations are sufficient to virtually exclude internal
variability and self-acceleration as an explanation for the observed
long-term trend, clustering, and magnitude of recent sea-ice minima.
Instead, the recent retreat is well described by the superposition of an
externally forced linear trend and internal variability. For the externally
forced trend, we find a physically plausible strong correlation only with
increasing atmospheric CO2 concentration. Our results hence show that the
observed evolution of Arctic sea-ice extent is consistent with the claim
that virtually certainly the impact of an anthropogenic climate change is
observable in Arctic sea ice already today."

Global Climate Models Struggle to Account for the Death Spiral

Arctic sea ice has declined at a rate significantly faster than global
climate models have predicted. Vinnikov et al.
(1999)<http://www.skepticalscience.com/lessons-from-past-predictions-vinnikov-arctic-sea-ice.html>used
the aforementioned GFDL and Hadley Centre climate models, forced by
greenhouse gases and sulfate aerosols, to project how Arctic sea ice extent
would change in the future. As is the case with most climate models, they
under-predicted the ensuing decline (Figure 8). In fact, the Arctic sea ice
decline is already 27 years ahead of Vinnikov's projections.As Stroeve et
al. (2012) <http://www.agu.org/pubs/crossref/2012/2012GL052676.shtml>discuss,
newer climate models have made some progress in this area, but
still cannot account for the full extent of the Arctic sea ice decline.

"Previous research revealed that the observed downward trend in September
ice extent exceeded simulated trends from most models participating in the
World Climate Research Programme Coupled Model Intercomparison Project
Phase 3 (CMIP3). We show here that as a group, simulated trends from the
models contributing to CMIP5 are more consistent with observations over the
satellite era (1979–2011). Trends from most ensemble members and models
nevertheless remain smaller than the observed value."

This may be due to the difficulty in accounting for natural variations, or
the physics associated with the domino effect discussed
above<http://www.skepticalscience.com/record-arctic-sea-ice-melt-to-levels-unseen-in-millennia.html#domino>,
or both. There are some regional models (i.e. see the model of Maslowski et
al.<http://www.ametsoc.org/atmospolicy/documents/May032006_Dr.WieslawMaslowski.pdf>,
also discussed here<http://www.climatecodered.org/2012/08/big-call-cambridge-prof-predicts-arctic.html>)
which have had success in accounting for and predicting Arctic sea ice
changes, but climate models overall have been too conservative in
projecting the rapid decline in Arctic sea ice.
Human-Caused Arctic Sea Ice Death Spiral

The scientific literature is clear that the current record Arctic sea ice
decline is beyond what has occurred due to natural variability for at least
the past several millennia, certainly beyond what occured circa 1940, and
that human influences are primarily responsible for the rapid rate of the
death spiral. The rapid Arctic warming appears to have caused a domino
effect by resulting in record ice melt and thus a significant decrease in
local albedo, and therefore an increase in absorbed solar radiation. The
rapid warming and increased solar radiation absorption have combined to
result in younger, thinner Arctic sea ice, which therefore melts more
easily, making record low extents more likely to occur.

It is wishful thinking to believe that the Arctic sea ice death spiral
could simply be due to natural variability. The scientific literature
clearly shows that human-caused warming is the main driver behind this
exceptionally rapid decline.

-------------------------------------------

Vision2020 Post: Ted Moffett
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.fsr.com/pipermail/vision2020/attachments/20120905/b8049c52/attachment-0001.html>