Oceanography

The Official Magazine of the Oceanography Society

CITATION
Walsh, J.E. 2013. Melting ice: What is happening to Arctic sea ice, and what does it mean for us?
Oceanography 26(2):171–181, http://dx.doi.org/10.5670/oceanog.2013.19.

DOI
http://dx.doi.org/10.5670/oceanog.2013.19

COPYRIGHT
This article has been published in Oceanography, Volume 26, Number 2, a quarterly journal of
The Oceanography Society. Copyright 2013 by The Oceanography Society. All rights reserved.

USAGE
Permission is granted to copy this article for use in teaching and research. Republication,
systematic reproduction, or collective redistribution of any portion of this article by photocopy
machine, reposting, or other means is permitted only with the approval of The Oceanography
Society. Send all correspondence to: [email protected] or The Oceanography Society, PO Box 1931,
Rockville, MD 20849-1931, USA.

downloaded from http://www.tos.org/oceanography

 F o u r t e e n t h A n n u a l R o g e r R e v e l l e c o m m e m o r at i v e L e c t u r e

Melting Ice
By J o h n E . Wa l s h

What Is Happening to Arctic Sea Ice,

and What Does It Mean for Us?
The Roger Revelle Commemorative Lecture Series was created by the Ocean Studies Board of the
National Academies in honor of Roger Revelle to highlight the important links between ocean sciences
and public policy. John Walsh, the fourteenth annual lecturer, spoke on March 20, 2013, at the
Baird Auditorium, Smithsonian Institution, National Museum of Natural History.

Abstr ac t. Sea ice has emerged as the canary in the coal mine of climate Introduc tion: Is the
change. Its summer extent in the Arctic has decreased by about 50% over the past Arc tic a Bellwether of
decade, and the Arctic Ocean has undergone a regime shift from a cover of thick Global Climate Change?
multiyear ice to a largely seasonal and much thinner ice cover. The recent loss is Over the past few decades, environmen-
unprecedented in the periods of satellite and historical records of sea ice, and it tal changes in the Arctic have attracted
also appears to be unique in paleo reconstructions spanning more than a thousand the attention of scientists, residents of
years. A “perfect storm” of warmer atmospheric and oceanic forcing, together with Arctic communities, policy and decision
a boost from natural variability of wind forcing in some years, drove the change. makers, and, more recently, the broader
However, the reduction of ice coverage is not apparent in some sub-Arctic regions public. This region that was previously
during the winter, nor has it occurred in the Antarctic region. of little interest to outside residents has
Signals of a response to the loss of sea ice are emerging in the ocean and the now become a focal point of concern
atmosphere. Ocean heat storage during the ice-free season not only contributes about global climate change. It can be
to a later freeze-up than in the past, but it also reduces the thickness to which argued that the Arctic is now awakening
first-year ice can grow. The vulnerability of this thinner ice to rapid spring melt the “sleeping giant” of public awareness
is a manifestation of the ice-albedo-temperature feedback that has long been of climate change and a growing accep-
postulated as a contributor to polar amplification of climate change. More notably tance of its reality.
for middle latitudes, the loss of sea ice appears to be triggering a reduction of the
large-scale westerlies that characterize atmospheric circulation in middle and sub- John E. Walsh ([email protected])
polar latitudes. This response is consistent with increased persistence of departures is Chief Scientist, International Arctic
from normal temperature, precipitation, and extreme weather during autumn and Research Center, University of Alaska
winter in heavily populated areas of the Northern Hemisphere. Fairbanks, Fairbanks, AK, USA.

Oceanography | June 2013 171

 But what is driving the changes in These effects, which may be quite con- temperature. The temperature recon-
the Arctic environment? And what are sequential for heavily populated areas struction in Figure 1 is primarily from
the consequences of these changes for of Eurasia and the United States, can be terrestrial sources, including lake sedi-
regions outside the Arctic? While these counterintuitive, but are nevertheless ments, pollen records, diatoms, and tree
questions are still at the core of on­going scientifically plausible. rings, which provide information on
research in many countries, some hints warm-season temperatures. It shows
of the answers are emerging. Of all the Cause s of the Recent slow (summer) cooling in the Arctic for
variables that must be included in diag- Se a Ice Retre at most of the past 2,000 years. This cooling
nostic studies of Arctic environmental What Is the Evidence for is consistent with known slow variations
change, sea ice is perhaps the most Sea Ice Retreat? in Earth-sun orbital parameters that
prominent. Sea ice plays a key role in The most striking feature of recent affect the solar radiation reaching the
climate by modifying the exchanges changes in the Arctic, particularly in sea Arctic in the sunlit portion of the year.
between the ocean and the atmosphere, ice, is how quickly the Arctic is warm- However, the recent warming since the
but it also has many other complex con- ing and ice is melting relative to changes 1800s, confirmed by direct temperature
nections to the climate system. It appears seen in the long-term climate record. measurements (red line in Figure 1), has
to respond to global influences, and Figures 1 and 2 show reconstructions of left the Arctic warmer than at any time
recent decreases in sea ice may already Arctic summer temperatures (Kaufman in the preceding 2,000 years by a consid-
be affecting the larger climate system et al., 2009) and Arctic sea ice (Kinnard erable margin. The recent instrumental
through a variety of physical, dynamical, et al., 2011) over time frames of the past temperatures are outside the envelope of
and ecological processes (AMAP, 2011). 1,500–2,000 years. These reconstructions the natural variability seen in the recon-
Moreover, sea ice is changing faster than are based on proxy information—the struction; for example, the warming is far
other Arctic environmental variables. history of past climate shifts preserved more than simply a recovery from the so-
For these reasons, Arctic sea ice has in ancient sediment deposits, in ice called Little Ice Age, which is apparent
been referred to as the “bellwether” of sheets, or in annual growth rings of trees. from the late 1500s through the 1800s in
global climate change. Is such a notion Natural systems change in response Figure 1. The sea ice reconstruction in
justified? The answer to that question to environmental changes and hence Figure 2 is based on high-resolution ter-
requires an understanding of the reasons “record” shifts in parameters such as restrial proxies from the circum-Arctic
for the recent dramatic changes in Arctic
sea ice. The current understanding of
those reasons is one focus of this article.
A second focus is the impact of
retreating Arctic sea ice on the broader
climate system, particularly in mid-
latitudes. If sea ice truly provides an early
indication of changes in global climate,
then loss of sea ice could already be
influencing climate in regions outside of
the Arctic. There are scientific reasons
to expect that such a mechanism exists,
and modeling studies (Honda et al.,
2009; Liu et al., 2012) and observational Figure 1. A reconstruction of Arctic summer temperatures. The blue line shows estimates
analyses (Francis and Vavrus, 2012) both of Arctic temperatures over the last 2,000 years, based on proxy records from lake sedi-
ments, ice cores, and tree rings. The green line shows the long-term cooling trend. The red
provide some intriguing suggestions for line shows the recent warming based on actual observations. From Kaufman et al. (2009),
effects on climate in the mid-latitudes. modified by UCAR

172 Oceanography | Vol. 26, No. 2

 domain: ice cores, tree rings, and lake averages is typical of variations in many the darker water surface absorbs more
sediments. As with the temperature climate variables (IPCC, 2007). of the sun’s energy as heat, resulting in
reconstruction described above, these further warming and melting of ice and
proxies respond primarily to changes Why Is the Arctic Warming snow. Conversely, an expansion of sea
in the warm season, so the reconstruc- So Quickly? ice results in greater reflection of solar
tions in Figure 2 depict summer sea Figure 3 shows the geographic varia- radiation and reduces the amount of heat
ice variations. The pan-Arctic time tion in the recent unprecedented rise absorbed. This positive feedback phe-
series (red line) shows that the recent in Arctic temperature. Over the past nomenon is called the temperature-ice-
decline of sea ice is unprecedented in the 60 years, the Arctic has warmed by albedo coupling. Two other factors that
1,450-year reconstruction, and mirrors more than 2°C, more than double the appear to have contributed to the recent
the recent, rapid warming evident in global average warming over the same polar amplification of warming include
Figure 1. The abrupt decline in the Fram period. Figure 3 illustrates the poleward an increase of atmospheric water vapor
Strait region during the past several increase, or “polar amplification,” of this (a strong greenhouse gas) in the Arctic
decades is also unique in the 1,450-year warming. Polar amplification is also seen (Serreze et al., 2012) and an increase of
time series for that region. However, the in periods of cooling in the historical poleward transports of heat by the ocean
reconstructions for the Chukchi Sea in record, and it is attributable in part to and the atmosphere (Shimada et al.,
Figure 2, and also for the Barents Sea the role of sea ice and its overlying snow 2006; Beszczynska-Möller et al., 2011).
(not shown), indicate that sea ice cover- cover (Serreze and Francis, 2006; Serreze Paired mid-September satellite
age in these regions was comparable to et al., 2009). Specifically, there is a posi- images from 1992 and 2012 show the
the present during the late 1500s and tive feedback that amplifies both warm- dramatic loss of sea ice in recent decades
early 1600s. This tendency for smaller ing and cooling trends because of the (Figure 4). The maximum seasonal
regions to show greater variability and change in the amount of solar radiation retreat of sea ice usually occurs in mid-
behave differently from hemispheric reflected by sea ice. When sea ice melts, September, so these images capture

800 1200 1600 2000

Figure 2. Recon-
T3 T2 T1
struction of Arctic
a
11 summer sea ice
variation. (a) Forty-
Arctic ice extent (× 10 6 km2 )

year smoothed
reconstructed late-
10
summer Arctic sea
ice extent, with 95%
12 confidence interval,
9 and yearly ice duration
Ice cover > 50% (months per year)

b in the (b) Chukchi Sea

10
and (c) Fram Strait.
Chukchi Sea Reprinted by permis-
8
8 sion from Macmillan
Publishers Ltd.: Nature,
6 Kinnard et al. (2011),
c copyright 2011 –4.1 –2.0 –1.0 –0.5 0.2 0.5 1.0 2.0 4.0 4.1

4 Figure 3. Geographic variation in the recent

unprecedented rise in Arctic temperature. The
Fram Strait 2 figure illustrates the poleward increase, or “polar
amplification,” of warming. Courtesy of NASA GISS
0

800 1200 1600 2000

Year (AD)

Oceanography | June 2013 173

the ice cover that survived the summer What Has Caused the Dramatic evidence for several of these drivers of
melt and hence can undergo additional Loss of Ice in Recent Decades? Arctic sea ice loss.
growth in thickness during the fol- Warming of the Arctic has undoubtedly The mean pattern of currents in
lowing winter. The extent of sea ice in contributed to the loss of sea ice, but the upper ocean of the Arctic and the
September 2012 was approximately half higher air temperatures alone cannot subpolar regions includes warm, saline
of the extent in 1992. The rapidity of this explain the rapid decrease over the past water from the North Atlantic that enters
decrease is unprecedented with respect few decades (Stroeve et al., 2011). Other the Arctic Ocean through the Barents
to the paleo-reconstruction shown in factors include: Sea and Fram Strait, then descends
Figure 2, as well as other reconstruc- • Periods of increased wind-driven to depths of 100–400 m and circu-
tions based on additional types of sea transport of older, thicker ice from the lates in a generally counterclockwise
ice information such as ship reports and Arctic into the North Atlantic direction around the Arctic Ocean.
coastal observations. Within the period • Increased flow of warmer ocean This water transports heat acquired at
of satellite observations (1979–2012), waters into the Arctic from the North lower latitudes to the Arctic. Some of
the loss of sea ice has occurred most Atlantic and the North Pacific this warmer water reaches the base of
abruptly in 2007 and 2012 (Perovich • Increased atmospheric warming as a the sea ice and contributes to bottom
et al., 2013). In both of these years, the consequence of increased humidity in melt. Measurements from ocean moor-
summer minimum was more than a the Arctic, and perhaps also because ings and research vessels show that the
half million square kilometers below the of variations of cloudiness inflowing Atlantic water has gotten
previous record. Interestingly, the extent • The amplified loss of sea ice due to warmer, albeit irregularly, over the past
of winter sea ice shows a much smaller increased absorption of solar radia- two decades (Alexeev et al., 2013), as
decline in recent decades. The result tion by the darker ocean surface, as can be seen in Figure 5. This warming
is a much greater area of seasonal ice described earlier has been detected not only where the
(i.e., ice that forms during the autumn/ It is difficult to rank the relative impor- Atlantic water flows in, but also along
winter and melts the following spring). tance of each of these driving forces, the continental shelf break north of the
Because this ice has only a few months to but the emerging consensus is that Siberian coast (Polyakov et al., 2010).
grow, it is thin and readily deformed by together they have resulted in the “per- The mechanisms by which heat is trans-
the wind—a force that is responsible for fect storm” of forcing responsible for the ferred from the deeper Atlantic waters to
much of the short-term ice movement. rapid sea ice loss. I next highlight the the surface is unclear, but double diffu-
sion and mesoscale (~ 10 km diameter)
eddies have been suggested as possible
mechanisms. Double diffusion refers to
the ability of heat to move more read-
ily than salt through seawater, while
eddies can move water and heat verti-
cally. By using a heat budget approach,
scientists have estimated that the warmer
Atlantic water can account for around
several tenths to a meter of bottom melt
over the past decade.
On the Pacific side, Bering Strait is
the entry corridor for warmer water
of sub-Arctic origin. After entering
Figure 4. A pair of mid-September satellite images from 1992 and 2012 shows the dramatic loss of sea the Chukchi Sea, this water generally
ice in recent decades. Courtesy of University of Illinois, The Cryosphere Today moves eastward offshore of the northern

174 Oceanography | Vol. 26, No. 2

Alaskan coast. There are indications of attempted to place the ice-albedo feed- Impac ts of Se a Ice Loss
positive feedback whereby heat entering back into a quantitative framework The most direct and obvious impacts
from the Pacific reduces the thickness (e.g., Perovich and Richter-Menge, to date are in the Arctic, where sea ice
and coverage of sea ice north of Alaska 2009). The trend in the solar heat input loss is affecting people, marine life,
(Shimada et al., 2006). This thinner, to the Arctic Ocean over 1979–2005 and Arctic climate. However, there are
looser ice is more mobile and susceptible exceeds 2% per year in certain areas, emerging signs of impacts that extend
to movement by winds that drive the including the Beaufort-Chukchi-East into the mid-latitudes. Here, I highlight
Beaufort Gyre, thereby facilitating the Siberian Sea sector noted above, sug- both local and distant impacts, begin-
transport of the warmer Pacific water gesting a 50% increase over the 26-year ning with the Arctic and then addressing
from the Alaskan shelf to the deeper period. This increase precedes the rapid impacts on other regions. The discussion
waters offshore. This transport leads acceleration of ice retreat that began in is limited to impacts of diminished sea
to a reduction of the ice cover farther 2007, so the percent increase would be ice, acknowledging that Arctic warming
offshore, and the reduction then may be even higher if evaluated through 2012. also has other important impacts such as
further enhanced by the albedo effect of The increasing heat input, even prior to the contribution of melting glaciers and
reduced reflection of solar radiation. This 2007, is far greater than the 1 to 2 W m–2 ice sheets to rising sea level.
mechanism is especially relevant to the of surface radiative warming from the
ice loss of the past decade because the increase in greenhouse gases. The more How Does Sea Ice Loss Impact
sector containing the Beaufort, Chukchi, than 50% increase in solar absorption People in the Arctic?
and East Siberian Seas has experienced demonstrates the importance of the tem- Coastal communities in Alaska and
the greatest sea ice loss (Figure 4). perature-ice-albedo feedback in acceler- Siberia are experiencing increased flood-
Recent heat budget studies have ating Arctic warming. ing and coastal erosion as a result of the
loss of the sea ice buffer that previously
protected the coast from wind-driven
waves during summer and autumn
storms. As a result, several communities
in Alaska are facing costly relocation
away from the coast. An increase in ship
traffic is another impact of the retreating
sea ice cover, as the lengthening open-
water season presents opportunities
for offshore resource extraction, tour-
ism, and shortened transit times for the
marine transport industry. The oil and
gas industry is a particular beneficiary
of the diminished ice cover, as seen
by the recent increase in exploratory
activity over the shelf seas north of
Alaska and Russia. Such activity brings
potential benefits as well as risks to
northern communities.
Figure 5. Measurements show that inflowing Atlantic water has gotten warmer over the past two
decades. This image shows concentration of multiyear ice in 2004 and 2008 (two upper insets)
and temperature in the Atlantic water core (main graph) measured in Fram Strait (yellow circle,
upper left inset) and from transects of different extent (three lower insets) made in September
2004 and 2006 and October 2008 at 31°E, 80°N (location marked by yellow line in the upper left
inset). From Alexeev et al. (2013)

Oceanography | June 2013 175

How Does Rapid Loss of Arctic from sea ice retreat (Figure 6). the model successfully reproduced the
Sea Ice Impact Marine Life? Assessments of changing sea ice observed levels of sea ice loss and also
In the Bering Sea, there is some evidence impacts on marine life, particularly the showed increases in primary productiv-
that individual species as well as eco- lower trophic levels, are largely reliant ity (photosynthesis by algae and plank-
systems respond to climate variations on modeling studies. Marine ecosys- ton at the base of the food web) con-
(Grebmeier et al., 2006). But farther tem modeling is a key element of the sistent with satellite-derived estimates.
north, in the Arctic Ocean, the dynam- US-supported Bering Sea Ecosystem The simulated primary productivity
ics of ecosystems and the food web are Study (BEST), and biogeochemical mod- increased at various depths in the water
much less known. This lack of knowledge eling is just now being applied to the column, including areas under sea ice.
is reflected in the decision of the federal Arctic Ocean, where the magnitude of The under-ice increases are consistent
government to impose a moratorium on sea ice retreat is greatest. One such mod- with the greater penetration of light
commercial fishing in US waters north eling study incorporated marine biogeo- when ice is thinner. In the model, dia-
of Bering Strait. The US Department of chemical cycling into a state-of-the-art toms and flagellates increased, as well as
the Interior has listed the polar bear as Arctic Ocean sea ice model (Zhang et al., two types of zooplankton. Although the
a threatened species, based in part on 2010). Tested by simulating a two-decade simulations do not extend to the most
changes in polar bear habitat arising period in the recent past (a “hindcast”), recent years of greatest sea ice retreat,
the increased productivity of lower tro-
phic levels has profound implications
for higher trophic levels in the food web,
including fish and marine mammals,
with the potential to alter the ecologi-
cal structure of large areas of the Arctic
Ocean that have historically been cov-
ered by perennial sea ice.

What Are the Impacts of

Rapid Loss of Arctic Sea Ice
on Climate Change?
The continental shelves of the Russian
seas are among the largest in the world,
and much of the seafloor in these seas
contains relict permafrost. This perma-
frost and the underlying layers contain
large stores of methane, a powerful
greenhouse gas, in the form of methane
hydrates. Recent measurements from
these areas (Shakhova et al., 2010) have
detected releases of methane (methane
flares) consistent with perforations in
Figure 6. Projected changes in spatial distribution and integrated annual area of optimal polar bear
the permafrost above the hydrate stores
habitat. The base map shows the cumulative number of months per decade where optimal polar
bear habitat is either lost (red) or gained (blue) from 2001–2010 to 2041–2050. Offshore gray shading (Figure 7). The extent to which warming
denotes areas where optimal habitat is absent in both periods. Insets show the average annual cumula- of the shelf waters, enhanced by sea ice
tive area of optimal habitat (right y-axis, line plot) for four 10-year periods in the twenty-first century
(x-axis midpoints) and their associated percent change in area (left y-axis, histograms) relative to the
retreat in this region, has accelerated the
first decade (2001–2010). Courtesy of USGS subsea permafrost thaw and the release

176 Oceanography | Vol. 26, No. 2

of the methane stores is not known.
However, preliminary estimates sug-
gest that the amount of methane being
released from the East Siberian Shelf
region may be comparable to the amount
released from the remainder of the
global ocean. Given the plausibility of
further acceleration of subsea permafrost
thaw in areas of sea ice loss, together
with the potency of methane as a green-
house gas, this region bears watching
for its potential to contribute to future
global warming.
The most direct impact of sea ice
retreat on climate is the warming of the Figure 7. The subsea permafrost of the East Siberian Arctic Shelf (an area of about 2 million square
Arctic atmosphere. The warming would kilometers) is more porous than previously thought. The ocean on top of the permafrost and
the heat from the mantle below the permafrost warm it and make it permeable, allowing meth-
be expected to be strongest in autumn,
ane gas stored beneath it under pressure to escape into the atmosphere. The amount leaking
when the additional heat absorbed by the from this locale is comparable to all the methane from the rest of the world ocean put together.
newly open ocean delays freeze-up and is Methane is a greenhouse gas more than 30 times more potent than carbon dioxide. Courtesy of
Zina Deretsky, National Science Foundation, based on Shakhova et al. (2010)
released back to the atmosphere. Because
the air normally tends to cool in autumn,
the impact of the heat released from
the ocean is greatest in the September 90

to November period. This heat release

60
from the ocean continues even after
freeze-up because the ice is thinner and
30
less insulating than in previous decades.
This ocean-to-atmosphere heat transfer 0
affects the distribution of atmospheric
pressures that, in turn, drive atmospheric –30
circulation (Overland and Wang, 2010).
Figure 8 displays evidence that sea ice –60

loss is already affecting the atmosphere

–90
in autumn and in winter. The figure Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
shows the 2007–2012 warmth relative
to the 1971–2000 “normal” as a func- –3.0 –1.9 –1.3 –0.8 –0.5 –0.1 0.1 0.5 0.8 1.3 1.9 3.0
°C
tion of latitude and calendar month. The
Figure 8. The warmth of 2007–2012 relative to the 1971–2000 “normal” as a function
pattern in Figure 8 not only highlights
of latitude and calendar month. The data provide evidence that the loss of sea ice is
the polar amplification discussed earlier already affecting the atmosphere in autumn and winter. Courtesy of NASA GISS
but also shows that the relative increase
in Arctic warming is greatest in autumn
and in early winter, precisely the sea-
sonality expected from the loss of sea
ice. Figure 9 shows that the warming is

Oceanography | June 2013 177

 strongest near the surface, consistent (October) snow cover over Eurasia. expands vertically. Because air pres-
with the idea that such changes are The increase since the late 1980s has sure is the weight of the overlying air,
driven by changes in sea ice. been more than 1.4 million square kilo- this expansion increases the altitude
meters of snow cover per decade. The at which a particular pressure will be
How Are the Mid-Latitudes correlation between autumn ice extent found. Figure 9 shows that there has
Impacted by Rapid Loss of in the Arctic and winter snow cover been an increase in elevation (the geo-
Arctic Sea Ice? over the Northern Hemisphere is even potential height) of the pressures in the
The fact that the warming is strongest in more noteworthy. Reduced Arctic sea Arctic atmosphere, as would be expected
autumn and at the Arctic Ocean’s sur- ice extent in autumn is associated with with warming. Corresponding to these
face (Figures 8 and 9) is consistent with increased winter snow cover in large increases of geopotential height are
the delayed freeze-up noted above. The areas of eastern Asia, central Europe, and increases of pressure at all elevations,
delayed freeze-up means that an ice-free the northern half of the United States with the largest increases at the highest
ocean underlies the atmosphere at a time (Liu et al., 2012). But why should sea elevations (as in Figure 9). Higher pres-
of year when reduced solar radiation ice in autumn affect wintertime snow sures in middle and upper levels of the
favors strong atmospheric cooling. The cover in middle latitudes? The proposed Arctic atmosphere favor a weakening
expanded areas of open water during explanation for this relationship is based of westerly winds (or a strengthening of
autumn and early winter represent not on reasoning about the pressure field easterly winds) at lower latitudes. Francis
only a source of heat to the lower atmo- that drives the primary feature of the and Vavrus (2012) show that there has
sphere but also a source of moisture. Northern Hemisphere atmospheric cir- indeed been a weakening of the west-
This additional moisture increases the culation—the west-to-east flow at middle erly winds in the middle troposphere
amount of precipitation falling over the and upper levels of the mid-latitude over the past two decades. The season-
Arctic Ocean and adjacent land areas atmosphere. This airflow includes the ality of this weakening westerly flow
during autumn and early winter. Not jet stream, with its wavelike meanders shows agreement with the loss of sea ice
surprisingly, recent decades have seen around the hemisphere. (i.e., the westerly winds weakened pri-
a highly significant increase in autumn When a column of air warms, it marily in autumn and winter), and the

a b
500 NOAA/ESRL Physical Sciences Division NOAA/ESRL Physical Sciences Division
550
600
650
Pressure (hPa)

700
750
800
850
900
950
1000
60°N 63°N 66°N 69°N 72°N 75°N 78°N 81°N 84°N 87°N 60°N 63°N 66°N 69°N 72°N 75°N 78°N 81°N 84°N 87°N 90°N

–5 –4 –3 –2 –1 0 1 2 3 4 5 –35 –25 –15 –5 5 15 25 35

Oct to Dec; 2002 to 2008; 135°E to 225°E Oct to Dec; 2002 to 2008; 135°E to 225°E

Figure 9. (a) Vertical cross-section composite plot of air temperature anomalies (ºC) for the section covering the East Siberia Sea, Chukchi Sea, and
Beaufort Sea from Bering Strait to the North Pole for October to December 2002–2008. (b) Vertical cross-section composite plot of geopotential height
anomalies (dynamic meters) for the section from Bering Strait to the North Pole for October to December 2002–2008 over the Siberia Sea to Beaufort Sea
area. The data show that Arctic warming is strongest at the ocean surface, and that there has been an increase of elevation (geopotential height) of the
pressures in the Arctic atmosphere, as would be expected with warming. From Overland and Wang (2010)

178 Oceanography | Vol. 26, No. 2

timing of the wind-pattern changes over The study by Francis and Vavrus (2012) Wallace, 1998). The Arctic Oscillation is
the past few decades also agrees with the and other recent studies suggest that strongly correlated with the strength of
loss of sea ice (i.e., the largest decreases blocking is becoming more common the zonal (west-to-east) winds in middle
of westerly winds occurred after 2007). during autumn and winter. Extreme win- and high latitudes. When the Arctic
A general weakening of the prevail- ter anomalies, such as the extended cold Oscillation enters its negative phase, the
ing westerlies means more meandering periods in Europe during the 2010/11 west-to-east flow weakens and north-
of the airflow, including the jet stream. and 2011/12 winters, and the cold, south meanders of the airflow (including
The jet stream typically has three to snowy winter of 2010/11 in the United the jet stream) become more prominent,
seven waves (meanders) around the States, are consistent with this notion of especially in the North Atlantic sector.
hemisphere at any time (Figure 10), increased blocking. Even the extremely This is an example of a blocking pat-
with northward bulges (referred to as mild winter of 2011/12 in the United tern, discussed earlier. Extensive autumn
ridges) and southward dips (referred to States can be viewed as an example of snow over Eurasia has been linked to a
as troughs). As the Arctic warms rela- blocking, although the persistent char- negative phase of the Arctic Oscillation
tive to lower latitudes, these waves are acteristic in that case was the absence during winter, through a complex
predicted to increase in amplitude. These of deep troughs and their associated dynamical mechanism (Cohen et al.,
predictions are confirmed in an analysis cold air masses. 2012). This linkage is consistent with the
of observational data by Francis and Extreme winter weather, as well as previously described effects of sea ice on
Vavrus (2012) who provide evidence that other weather events, has been linked atmospheric wind patterns because sea
ridges have indeed strengthened more to a particular mode of variability in ice retreat contributes to the increase
than troughs have weakened, increas- atmospheric pressure systems called the of Eurasian snow cover, which in turn
ing wave amplitudes in the Northern Arctic Oscillation. It has a positive phase, favors a negative (blocking) phase of the
Hemisphere. However, recent work by with relatively high pressure over the Arctic Oscillation.
Screen and Simmonds (2013) indicates polar regions and low pressure at mid- A topic of recent interest is the
that conclusions about recent changes latitudes, and a negative phase in which extent to which individual storm events
in the amplitude of jet stream waves are this pattern is reversed (Thompson and can be tied to the atmospheric signals
quite sensitive to the method by which
wave amplitude is evaluated.
How do these changes in wave Figure 10. The jet stream on
amplitude affect weather and climate in January 30, 2013, immediately
mid-latitudes? With weaker westerlies prior to a major blizzard that
affected the northeastern United
and larger-amplitude waves, the normal States. The jet stream typically
west-to-east progression of waves in the has three to seven waves (mean-
ders) around the hemisphere
atmosphere is slowed. This increases the
at any time, with northward
persistence of departures from normal bulges referred to as ridges
surface weather associated with the and southward dips referred
to as troughs. Courtesy of
waves—for example, cold surface condi-
University of Wisconsin-Madison,
tions beneath troughs, warm surface Department of Atmospheric and
conditions beneath ridges. In extreme Oceanic Sciences

cases, features can lock into place for

weeks, a situation known meteorologi-
cally as “blocking.” The extended dura-
tion of anomalous weather can contrib-
ute to large departures from normal over UW-Madison, Dept of AO
monthly or even seasonal time scales.
0 20 40 60 80 100 120 140 160 180

Oceanography | June 2013 179

 Roger Revelle associated with sea ice. For example,
Hurricane Sandy’s highly unusual west-
For almost half a century, ward turn into the mid-Atlantic coast
Roger Revelle was a leader in the occurred when the Arctic Oscillation
field of oceanography. Revelle was in a strongly negative (blocking)
trained as a geologist at Pomona phase. The absence of westerly winds
College and the University of indeed enabled the storm to track west-
California, Berkeley. In 1936, he ward. Was this related to the unprec-
received his PhD in oceanogra- edented retreat of sea ice in the autumn
phy from the Scripps Institution of Oceanography. As a young of 2012 (Figure 4)? The connection
naval officer, he helped persuade the Navy to create the Office between Arctic sea ice and Hurricane
of Naval Research (ONR) to support basic research in ocean- Sandy is tenuous because of uncertain-
ography and was the first head of ONR’s geophysics branch. ties in the chain of associations linking
Revelle served for 12 years as the Director of Scripps (1950–1961, sea ice with trajectories of individual
1963–1964), where he built up a fleet of research ships and initi- storms. It is fair to surmise, however, that
ated a decade of expeditions to the deep Pacific that challenged sea ice loss may have increased the odds
existing geological theory. that a late-season hurricane would take
Revelle’s early work on the carbon cycle suggested that the an unusual westward turn in middle lati-
sea could not absorb all the carbon dioxide released from burn- tudes. Associations between sea ice and
ing fossil fuels. He organized the first continual measurement individual events will likely be an active
of atmospheric carbon dioxide, an effort led by Charles Keeling, area of research in the coming years.
resulting in a long-term record that has been essential to current
research on global climate change. With Hans Suess, he published Conclusion
the seminal paper demonstrating the connection between increas- Melting of Arctic sea ice has conse-
ing atmospheric carbon dioxide and burning of fossil fuels. Revelle quences for life both in the Arctic and
kept the issue of increasing carbon-dioxide levels before the public in the mid-latitudes of the Northern
and spearheaded efforts to investigate the mechanisms and conse- Hemisphere. Will the rapid loss of sea ice
quences of climate change. continue into the future? Global climate
Revelle left Scripps for critical posts as Science Advisor to the models project such a continuation,
Department of the Interior (1961–1963) and as the first Director especially in the warm season, through
of the Center for Population Studies at Harvard (1964–1976). the remainder of the century. To date,
Revelle applied his knowledge of geophysics, ocean resources, and the actual sea ice retreat is ahead of the
population dynamics to the world’s most vexing problems: pov- pace of sea ice loss projected in nearly
erty, malnutrition, security, and education. all climate models (Stroeve et al., 2012).
In 1957, Revelle became a member of the National Academy Although sea ice may well increase in
of Sciences to which he devoted many hours of volunteer service. some years or even in multiyear peri-
He served as a member of the Ocean Studies Board, the Board ods because of natural variability in the
on Atmospheric Sciences and Climate, and many committees. climate system (Kay et al., 2011), cur-
He also chaired a number of influential Academy studies on rent projections indicate an essentially
subjects ranging from the environmental effects of radiation to ice-free Arctic Ocean in the summer by
understanding sea level change. sometime around the middle of this cen-
Photo credit: SIO Archives, UCSD tury and even sooner if the actual loss
continues to outpace model-projected
losses. Given the accelerating sea ice

180 Oceanography | Vol. 26, No. 2

loss in the past decade, the implica- Group I to the Fourth Assessment Report Serreze, M.C., and J.A. Francis. 2006. The
of the Intergovernmental Panel on Climate Arctic amplification debate. Climate
tions for middle-latitude as well as Change, 996 pp. Dynamics 76:241–264, http://dx.doi.org/​
Arctic residents have grown in signifi- Kaufman, D.S., D.P. Schneider, N.P. McKay, 10.1007/s10584-005-9017-y.
C.M. Ammann, R.S. Bradley, K.R. Briffa, Shakhova, N., I. Semiletov, A. Salyuk, V. Yusupov,
cance and urgency. With sea ice retreat
G.H. Miller, B.L. Otto-Bliesner, J.T. Overpeck, D. Kosmach and O. Gustafsson. 2010. Extensive
emerging as a trigger of changes in B.M. Vinther, and Arctic Lakes 2K Project methane venting to the atmosphere from
climate throughout much of the United Members. 2009. Recent warming reverses long- sediments of the East Siberian Arctic shelf.
term Arctic cooling. Science 325:1,236, http:// Science 327:1,246–1,250, http://dx.doi.org/​
States, the Arctic’s role as a bellwether dx.doi.org/10.1126/science.1173983. 10.1126/science.1182221.
of change is not just a concern about a Kay, J.E., M.M. Holland, and A. Jahn. 2011. Shimada, K., T. Kamoshida, M. Itoh, S. Nishino,
Inter-annual to multi-decadal Arctic sea ice E. Carmack, F. McLaughlin, S. Zimmerman,
remote and beautiful part of the globe. extent trends in a warming world. Geophysical and A. Proshutinsky. 2006. Pacific Ocean
There is increasing awareness that what Research Letters 38, L15708, http://dx.doi.org/​ inflow: Influence on catastrophic reduction of
10.1029/2011GL048008. sea ice cover in the Arctic Ocean. Geophysical
happens in the Arctic does not stay
Kinnard, C., C.M. Zdanowicz, D.A. Fisher, Research Letters 33, L08605, http://dx.doi.org/​
in the Arctic. E. Isaksson, A. de Vernal, and L.G. Thompson. 10.1029/2005GL025624.
2011. Reconstructed changes in Arctic sea ice Stroeve, J.C., M.C. Serreze, M.M. Holland, J.E. Kay,
over the past 1,450 years. Nature 479:509–512, J. Maslanik, and A.P. Barrett. 2011. The Arctic’s
Reference s http://dx.doi.org/10.1038/nature10581. rapidly shrinking sea ice cover: A research syn-
Alexeev, V.A.,V.V. Ivanov, R. Kwok, and Liu, J., J.A. Curry, H. Wang, M. Song, and thesis. Climatic Change 110:1,005–1,027, http://
L.H. Smedsrud. 2013. North Atlantic warm- R.M. Horton. 2012. Impact of declining Arctic dx.doi.org/10.1007/s10584-011-0101-1.
ing and declining volume of arctic sea ice. sea ice on winter snowfall. Proceedings of the Stroeve, J.C., V.M. Kattsov, A.P. Barrett,
The Cryosphere Discussions 7:245–265, http:// National Academy of Sciences of the United M.C. Serreze, A. Pavlova, M.M. Holland, and
dx.doi.org/10.5194/tcd-7-245-2013. States of America, http://dx.doi.org/10.1073/ W.N. Meier. 2012. Trends in Arctic sea ice
AMAP (Arctic Monitoring and Assessment pnas.1114910109. extent from CMIP5, CMIP3 and observations.
Programme). 2011. Snow, Water, Ice and Overland, J.E., and M. Wang. 2010. Large- Geophysical Research Letters 39, L16502, http://
Permafrost in the Arctic-SWIPA. Arctic scale atmospheric circulation changes are dx.doi.org/10.1029/2012GL052676.
Monitoring and Assessment Programme, associated with the recent loss of Arctic Thompson, D.W.J., and J.M. Wallace. 1998. The
Oslo, Norway, 668 pp. sea ice. Tellus 62A:1–9, http://dx.doi.org/​ Arctic oscillation signature in the wintertime
Beszczynska-Möller, A., R.A. Woodgate, 10.1111/j.1600-0870.2009.00421.x. geopotential height and temperature fields.
C. Lee, H. Melling, and M. Karcher. 2011. Perovich, D., W. Meier, M. Tschudi, S. Gerland, Geophysical Research Letters 25 (9):1,297–1,300,
A synthesis of exchanges through the main and J. Richter-Menge. 2013. Sea ice. Arctic http://dx.doi.org/10.1029/98GL00950.
oceanic gateways to the Arctic Ocean. Report Card: Update for 2012. Available online Zhang, J., Y.H. Spitz, M. Steele, C. Ashjian,
Oceanography 24(3):82–99, http://dx.doi.org/​ at http://www.arctic.noaa.gov/reportcard/ R. Campbell, L. Berline, and P. Matrai. 2010.
10.5670/oceanog.2011.59. sea_ice.html (accessed May 2, 2013). Modeling the impact of declining sea ice on the
Cohen, J.L., J.C. Furtado, M.A. Barlow, Perovich, D., and J. Richter-Menge. 2009. Loss of arctic marine planktonic ecosystem. Journal
V.A. Alexeev, and J.C. Cherry. 2012. sea ice in the Arctic. Annual Review of Marine of Geophysical Research 115, C10015, http://
Arctic warming, increasing snow cover Science 1:417–441, http://dx.doi.org/10.1146/ dx.doi.org/10.1029/2009JC005387.
and widespread wintertime boreal cool- annurev.marine.010908.163805.
ing. Environmental Research Letters 7, Polyakov, I.V., V.A. Alexeev, U.S. Bhatt,
014007, http://dx.doi.org/10.1088/​ E.I. Polyakova, and X. Zhang. 2010. North
1748-9326/7/1/014007. Atlantic warming: Patterns of long-term
Francis, J.A., and S.J. Vavrus. 2012. Evidence trends and multidecadal variability. Climate
linking Arctic amplification to extreme Dynamics 34:439–457, http://dx.doi.org/​
weather in mid-latitudes. Geophysical 10.1007/s00382-008-0522-3.
Research Letters 39, L06801, http:// Screen, J.A., and I. Simmonds. 2013. Exploring
dx.doi.org/10.1029/2012GL051000. links between Arctic amplification and
Grebmeier, J.M., J.E. Overland, S.E. Moore, mid-latitude weather. Geophysical Research
E.V. Farley, E.C. Carmack, L.W. Cooper, Letters 40:959–964, http://dx.doi.org/10.1002/
K.E. Frey, J.H. Helle, F.A. McLaughlin, grl.50174.
and S.L. McNutt. 2006. A major ecosystem Serreze, M.C., A.P. Barrett, and J.C. Stroeve.
shift observed in the northern Bering Sea. 2012. Recent changes in tropospheric water
Science 311:1,461–1,464, http://dx.doi.org/​ vapor over the Arctic as assessed from radio-
10.1126/science.1121365. sondes and atmospheric reanalyses. Journal
Honda, M., J. Inoue, and S. Yamane. 2009. of Geophysical Research 117, D10104, http://
Influence of low Arctic sea-ice minima on dx.doi.org/10.1029/2011JD017421.
anomalously cold Eurasian winters. Geophysical Serreze, M.C., A.P. Barrett, J.C. Stroeve,
Research Letters 36, L08707, http://dx.doi.org/​ D.N. Kindig, and M.M. Holland. 2009. The
10.1029/2008GL037079. emergence of surface-based Arctic amplifica-
(IPCCP) Intergovernmental Panel on Climate tion. The Cryosphere 3:11–19, http://dx.doi.org/​
Change. 2007. Climate Change 2007: The 10.5194/tc-3-11-2009.
Physical Science Basis. Contribution of Working

Oceanography | June 2013 181