Changing Pathways - Arctic Monitoring and Assessment ...

Expect changes and some surprises. Theseare the main conclusions from a review of thepathways by which contaminants are trans-ported to, from, and within the Arctic andhow these pathways might respond to shiftsin climate.

During the 1990s, wind and weather pat-terns in the Arctic were quite different fromthe previous three decades. It is too early tosay whether this is part of a natural, recurringchange in climate regimes or the result ofglobal warming. Nevertheless, the conditionsprovide some important indications abouthow pathways can change and potentiallyalter the load of contaminants to differentparts of the Arctic. Despite the uncertainty,one truth still stands. When it comes to con-

taminants, the Arctic is not remote or isolatedfrom the rest of the world. Human activities inindustrial and densely populated areas willcontinue to influence what was once thoughtto be a pristine environment.

This chapter summarizes current knowledgeon contaminant pathways and how they relateto climate change. It thereby provides furtherelaboration and discussion of some pointsraised in the chapters Persistent Organic Pollu-tants, Heavy Metals, and Radioactivity, espe-cially looking at time trends and future per-spectives. The chapter touches on the effects oflong-term climate change in the Arctic. Thistopic will be treated in more depth in the forth-coming Arctic Climate Impact Assessment(ACIA), due in 2004.

Changing Pathways

European Route E4,Stockholm

Saqqannguaq Road, Narsaq, Greenland F

IN

N

LA

RS

EN

PO

LF

OT

O

/

PE

LL

E

ER

IC

HS

SO

N

Climate change in the ArcticThe Arctic is subject to natural climate cycles.Some occur over time scales as short as a fewyears, while others may span decades, cen-turies, or even millennia. In addition to thisnatural variability, the Arctic will be affectedby global climate changes related to increasesin greenhouse gases.

The following is a short introduction toclimate change and climate variability in theArctic.

Global climate changewill warm the Arctic

Human activities, such as the burning of fossilfuels, release greenhouse gases to the atmos-phere. They affect the Earth’s energy balance,which in turn has the potential to influencetemperatures and weather patterns. Expertopinion, as expressed by the Intergovernmen-tal Panel on Climate Change, is that somechanges are already apparent. This conclusionis based on comparisons with past tempera-ture records and indirect signs of climate var-iability during the past 1000 years. In the pastcentury, the global mean air temperature hasincreased by 0.6°C. Based on computer mod-els of the effects of greenhouse gases on theglobal climate, the Earth’s air temperatureis expected to increase by an additional 1.4to 5.8°C over the next century. The range rep-resents uncertainty about future emissions aswell as an uncertainty about their effects.Climate models show that the warmingwill be especially pronounced in the Arctic.Excluding the more extreme predictions, the

annual mean air temperature may still in-crease by 5°C near the pole and by 2-3°Caround the margins of the Arctic Ocean.However, there are large regional variations,even including cooling in some areas.

The greatest warming will probably occurin winter. By the end of the 21st century, some

98Changing Pathways

Large cluster of roseroot on stony shore.Kangerterajiva, Green-land.

Arctic Climate Impact AssessmentClimate change and variability, and, morerecently, notable increases in ultraviolet radi-ation, have become important issues in theArctic over the past few decades. Under theauspices of the Arctic Council, a programhas been initiated to evaluate and synthesizeknowledge about climate variability, climatechange, and increased ultraviolet radiationand their consequences. This Arctic ClimateImpact Assessment (ACIA) will also examinepossible future impacts on the environmentand its living resources, for example on hu-man health, and on buildings, roads andother infrastructure.

Three major documents will be completedby 2004. They are a peer-reviewed scientificreport, a synthesis document summarizingresults, and a policy document providing re-commendations for coping with and adapt-ing to change. The writing of the first twodocuments is guided by an Assessment Steer-ing Committee with the lead authors, repre-sentatives from the Arctic Monitoring andAssessment Programme (AMAP), the Pro-gram for the Conservation of Arctic Floraand Fauna (CAFF), the International ArcticScience Committee (IASC), other interna-tional bodies, and persons representing theArctic indigenous peoples.

PO

LA

R

PH

OT

OS

/

H

EN

NI

NG

T

HI

NG

models predict that climate change caused bygreenhouse gases might produce an ArcticOcean that is free of sea ice in the summer.

It is not clear to what extent global climatechange has already affected the Arctic. How-ever, current models predict changes that areconsistent with observations made during the1990s.

Recent climate trendsfollow from Arctic Oscillation

It is well known that climate can oscillatebetween different climate regimes. El Niño/La Niña in the Pacific is one example outsidethe Arctic. In the Arctic, these climate regimesare characterized by a high or low ArcticOscillation Index, which captures differentregimes in atmospheric circulation (see boxbelow). Wind and weather patterns affect icedrift and the distribution of water masses inthe Arctic, which in turn can change the extentof ice cover. Changes in air circulation canthus influence the transport of contaminantsinto and within the Arctic in several ways.

Since the 1960s, there has been a change inthe overall pressure pattern in the Arctic. The1990s in particular have been characterized bylower than average atmospheric pressure overthe pole. Expressed in a different way, a low

99Changing Pathways

�Model projections ofchange in sea ice coverfor the Arctic Ocean.Annual mean sea iceextent is shown for theNorthern Hemisphere assimulated by two differentclimate models, which dif-fer in how they treat mix-ing of the water mass.

10

12

Mean sea ice extent, million km2

1900 1950 2000 2050 2100

2

4

6

8

0

The Arctic Oscillation and the North Atlantic OscillationA leading component of variation in the Arctic’s climate is governed by the Arctic Oscillation, which captures differentregimes in atmospheric circulation in the northern hemisphere. Atmospheric circulation patterns can be described by differ-ences in sea-level air pressure, or the barometer reading in layman’s terms. The Arctic Oscillation Index is a measure of seasurface air pressure patterns. Specifically, it captures winter pressure anomalies north of 20°North.

Strongly correlated with the Arctic Oscillation is the North Atlantic Oscillation, which is a measure of the surface airpressure difference between the Icelandic Low and the Azores High. This index is an indication of the main wind patternsover the North Atlantic.

2

1

1

0

0

–2

–1

–4

0

Winter Arctic Oscillation Index

Summer Arctic Oscillation Index

North AtlanticOscillation Index

Summer

Winter

1960 1970 1980 1990

1960194019201900 1970 1980 1990

1960 1970 1980 1990

2000

2000

2000

+

–

+

–

+

–

The Arctic Oscil-lation Index since1960 and theNorth AtlanticOscillation Indexsince 1900. Themaps represent acombination of theArctic OscillationIndex and atmos-pheric pressurefields during win-ter and summer.

Arctic Oscillation Index had been replaced bya high Arctic Oscillation Index. The cause forthis shift is not completely understood. It couldbe the result of natural climate cycles, whereshort- and long-term patterns have coincidedto produce a very high index in the 1990s.It could also be a sign of the Arctic respondingto global climate change. Regardless of whichexplanation turns out to be correct, thechanges observed in the early 1990s providean example of how the Arctic might respondto global warming, including examples of howclimate change may alter the transport of con-taminants.

Winds, precipitationand temperatureThe atmosphere provides an important path-way for contaminant transport. Winds carrycontaminants from source regions, while pre-cipitation promotes deposition to the land andthe sea. Temperature plays a role in determin-

ing the relative distribution of contaminantsbetween the air, land, and ocean. Changes inwind patterns, precipitation, or temperaturecan thus change the routes of entry of conta-minants into the Arctic and the locations atwhich contaminants are deposited to surfacesor re-emitted to the air.

Wind patterns govern pollution transport

The Arctic is characterized by relatively pre-dictable patterns of sea-level air pressure.Every winter, high-pressure areas form overthe continents, while low pressure cells domi-nate the northern Pacific (the Aleutian Low)and the northern Atlantic (the Icelandic Low).

These low- and high-pressure areas pro-duce wind patterns that pump airborne pollu-tants into the Arctic. The Icelandic Low pro-duces westerly winds over the eastern NorthAtlantic and southerly winds over the Nor-wegian Sea, which can carry pollution rapidlyfrom eastern North America and Europe intothe High Arctic. Similarly, the Aleutian Low

100Changing Pathways

Atmospheric pressurefields and wind patternsin winter and summerwith the Arctic Oscilla-tion Index strongly posi-tive (left) or stronglynegative (right).

1025

1014

1014

1018

1018

1014

1014

1014

1018

1018

1015

1005

1015

1015

1015

1015

1015

1010

1010

1006

1010

HH

H

H

H H

H

L

L L

L

LLL

Positive Arctic Oscillation Index(cyclonic)

Negative Arctic Oscillation Index(anticyclonic)

Positive Arctic Oscillation Index(cyclonic)

Negative Arctic Oscillation Index(anticyclonic)

Winter

Summer

Beaufort High

Siberian High

Aleutian Low

Icelandic Low

tends to steer air from Southeast Asia into theBering Sea, Alaska, and the Yukon Territory.Here, however, the mountains along the westcoast of North America obstruct the airflow,while intensive precipitation on their westernflanks provides a mechanism to deposit conta-minants to the surface.

In summer, the continental high-pressurecells disappear and the oceanic low-pressurecells are less intense. The result is muchweaker transportation of air and pollutantsinto the Arctic from southern areas duringsummer.

With a high Arctic Oscillation Index, as inthe 1990s, the Icelandic low deepens. More-over, it extends farther into the Arctic, acrossthe Barents Sea and into the Kara and LaptevSeas north of Russia. This increases windtransport eastward across the North Atlantic,across western and central Europe, and intothe Norwegian Sea. Also, deep storms withstrong winds become more frequent and ex-tend farther into the Arctic.

The result of this shift in winter wind pat-terns is that the Arctic becomes more stronglyconnected to industrial regions of North Amer-ica and Europe. The storms also carry rain orsnow, which can wash contaminants from theair and deposit them on the ground, on ice, orin the water.

The wind patterns in the Pacific appear tochange very little in a shift from a low to ahigh Arctic Oscillation Index.

Changes in wind patterns will affect all air-borne contaminants. For example, spraying ofpesticides in eastern North America and Eu-rope is more likely to show up as peaks inArctic air measurements during a high ArcticOscillation Index. Similarly, re-emissions ofpreviously deposited organic pollutants in thesoil and water of North America and Europewill enter these same pathways and thus betransported more readily to the north. How-ever, as we will see below, increased transportby air can be offset by other factors.

Precipitation transfers pollutantsfrom the air to slower ocean currents

Air transport of particle-associated metalssuch as lead, cadmium, and zinc will be af-fected by changing wind patterns. However,these pollutants are scavenged inefficientlywithin the Arctic and thus tend to stay in theair rather than deposit to the surface. Theactual load to land and sea surfaces in theArctic depends strongly on the amount andkind of precipitation. Changes in snow andrain patterns thus have a much greater poten-tial to alter loading than does a change inwind patterns.

Particulate metals wash out in high precipi-tation areas. If this occurs over the sea, metalscan then be carried by ocean currents. For ex-ample, lead from leaded gasoline, which is still

used in many Russian and Eastern Europeancars, rains out in the Nordic Seas and in thesouthern portion of the Eurasian Basin. Oceantransport is much slower than air transportand the pollution signal to various parts of theArctic can thus be delayed.

The changes in winds and temperature thatare associated with shifts in Arctic Oscillationare likely to affect precipitation. The networkto monitor changes is sparse, however, and itis thus difficult to assess trends. Over a longertime span, the past 40 years, there are indica-tions that precipitation has increased over Can-ada’s North by about 20 percent. More mois-ture is probably also moving into the Barents,Kara, and Laptev Seas, carried by the strongsoutherly winds in the Norwegian Sea duringautumn and winter. Models for long-term cli-mate change predict that the Arctic will be-come a wetter place, and a greater fraction of


Difference in winter pre-cipitation between lowand high North AtlanticOscillation Index.

Storm over the Norwe-gian Sea. Satellite image.

–1.5 –1.2 0.30 0.6– 0.6– 0.9 0.9 1.2 mm/day

Difference in winter precipitation

KO

NG

SB

ER

G

SA

TE

LL

IT

E

SE

RV

IC

ES

/

N

OA

A

the atmospheric particles that enter the Arcticare thus likely to deposit there.

Snow and fog are far more efficient thanrain in removing some contaminants from theair and depositing them to the surface. For met-als, both a change in the amount of precipita-tion or in the relative amounts of rain and snowcan thus have a large impact on transport.

In 1991, the Canadian air monitoring sta-tion at Alert recorded a marked dip in aerosolmetal concentrations. It was noted that thisdecrease coincided with the economic collapsethat followed the fall of the former SovietUnion, which significantly reduced emissionsof some heavy metals in Russia. However, theair concentrations could also have been af-fected by the shift toward a high Arctic Oscil-lation Index that occurred at this time. It is dif-ficult to determine the relative importance ofthe two explanations without data that bothcover a wide range of sites and span severalclimate change cycles. Nevertheless, the Alertexample illustrates that caution must be usedin assigning causes for contaminants trends inrelatively short time series.

Scavenging by rain and snow can also beimportant for particle-associated POPs, suchas some PCBs, and for POPs that to some

extent dissolve in water, such as HCHs andtoxaphene. High precipitation in the NordicSeas and southern Eurasian Basin would thusincrease the role of ocean currents and ice aspathways. In the Bering Sea, rainout has selec-tively removed beta-HCH from the air, andswitched the mode of delivery to the ArcticOcean from transport by winds to transportby ocean currents. Beta-HCH, a componentof the pesticide technical HCH, is especiallylikely to move from air to seawater.

Most of the Arctic has become warmer

Parts of the Arctic have become warmer in thepast 40 years. In spring, surface air tempera-tures in almost the entire High Arctic show asignificant warming. In the Eurasian part ofthe Arctic Ocean, there is a trend toward alonger period of the year when the sea ice ismelting. As an Arctic average, temperaturesover land have increased by up to 2°C perdecade during the winter and spring. However,there are significant regional variations. For ex-ample, on a yearly average basis, the westernGreenland-Baffin Bay area has been cooling.

Changes in air temperature can have a di-rect physical effect on the transport of somecontaminants. This is true for substanceswhose volatility, solubility, and adsorption tosolids are sensitive to temperature, which is thecase for most POPs. The previous AMAPassessment described how volatile contami-nants can reach the Arctic from their sourceregions in the south by a series of ‘hops’.Higher temperatures in the Arctic would leadto an increased potential for atmospherictransport. Previously deposited organic pollu-tants would also be volatilized once again andmove back into the atmosphere. On the otherhand, if the temperature difference betweenthe pole and equator decreases, as predictedby models, the global thermodynamic contrastthat favors the Arctic as a final reservoir wouldweaken. Higher temperatures could also speedup some of the chemical reactions that removepollutants from the atmosphere. Increases inultraviolet radiation, which are connected toozone depletion in the Arctic, also promotechemical reactions that destroy or change theform of contaminants.

Even more important than the effects on airchemistry might be that higher temperatureswill lead to more efficient degradation of con-taminants by aquatic microorganisms. Foralpha-HCH, a simple calculation shows that asignificant increase in temperature in the upperwater layers of the Arctic Ocean could sub-stantially reduce the environmental half-life ofthis substance. One model has tried to predicthow an increase in temperature would changethe health risk from hexachlorobenzene (HCB)to people in a temperate region. HCB poses ahealth risk partly because it biomagnifies inmarine food webs and can reach people from


1900

J F M A M J J A S O N D

1920 1940 1960 1980– 1.0

– 0.5

0

0

– 0.2

0.2

0.4

0.6

0.5

1.0

Temperature anomaly, °C

Temperature change per decade, °C

Average 1951-1980

55-85°N

Central Arctic Ocean

– 0.2- – 0.1

0.5 - 0.6

0.6 - 0.8

– 0.3- – 0.2

0.4 - 0.5

– 0.4- – 0.3

0.2 - 0.4

– 0.5- – 0.4

0.1 - 0.2

Temperature change,1961-1990,

°C per decade

North Pole

Temperature trends forthe Arctic showing theannual surface tempera-ture trends over theNorthern Hemisphereexpressed as rates ofchange for the period1961-90 (map), tempera-ture anomalies (55-85° N)for 1900-1995 evaluatedagainst the average for1951-1980, and (lowerpanel) the trend by monthin surface air tempera-ture of the central ArcticOcean for the period1979-1995 showing therecent warming to bemainly a winter-springphenomenon.

traditional foods. The model implied a reducedexposure with increasing temperatures. The rea-son is that higher temperatures would enhancedegradation and also force this pollutant fromwater into the air, reducing the water concen-tration and, therefore, reducing the amount ofHCB entering the bottom of the food web.

Changing water flows in rivers

Changes in temperature and precipitation willaffect runoff and flow in Arctic rivers. So far,changes in flow seem to be within normal year-to-year variability. With long-term climatechanges, models suggest that the flow in theYenisey, Lena, and Mackenzie Rivers is likelyto increase. In other rivers, such as the Ob, itmay decrease. For smaller rivers at high lati-tudes, the seasonal patterns of river flow arelikely to change. It is projected that earliersnowmelt in spring would change the timing,amplitude, and duration of spring flow.

There are also some changes in where riverwater goes once it has entered the ocean. Thisis discussed in the section New pathways inArctic Ocean surface waters on page 105.

Lakes, land, and glaciersIce can act as a physical barrier for contami-nants and also, at times, as a reservoir. Whathappens when higher temperatures melt ice inlakes, in the ground, and in glaciers?

Lakes are sensitive to changes

Arctic lakes are sensitive to climate change,as temperatures directly affect the timing offreeze up in the fall and ice melt in spring.This, in turn, affects the flow of water to,within, and from the lake. There are no studiesthat show effects of changes in the Arctic Os-cillation Index on Arctic lakes. In North Amer-ica, long-term change has been observed, how-ever. Over the past 100 years, there has been adelay of several days in freeze-up, while thespring break-up now comes almost a week ear-lier than it did a century ago. Changes in waterflow through lakes can have a large impact onthe transport of contaminants. Currently, Arc-tic lakes appear to retain only a small fractionof the contaminants they receive. The peak inrunoff from the snowmelt in their catchmentareas comes before the ice on the lake hasmelted or before the water in the lake hasbegun mixing from top to bottom, as it doeswhen the lake warms up in summer. The run-off, which contains recently deposited contam-inants, therefore traverses the lake just underthe ice or above most of the water column,flowing out as quickly as it flows in.

With the reduced ice cover and loss of per-mafrost that is expected with climate change,Arctic lakes will probably become more like

lakes farther south. Specifically, the water col-umn will mix earlier, increasing the likelihoodthat contaminants will be retained in the lake.Moreover, the warmer water, along with windmixing and more organic matter from the sur-rounding land, may influence primary produc-tion. A change in the amount or timing of pri-mary production may increase the opportunityfor contaminants to enter the food web directly.However, it could also lead to more sedimenta-tion, which, at least temporarily, removes con-taminants to the bottom sediments.

Permafrost changes may increasemercury cycling and natural radioactivity

In the Arctic, ice is a more or less permanentfeature on land. The soil is typically gripped inpermafrost, and only the relatively thin activelayer on top thaws in the summer. This layer,which supports all biological processes andany vegetation, can be limited to the top meteror less. In the 1990s, permafrost degradationoccurred in some parts of Alaska and Russia,but not in northeastern Canada. This matchesthe distribution of air temperature trendsobserved and predicted by climate models.

Permafrost melting will lead to more nutri-ents and sediments reaching lakes and rivers.The flow of organically bound carbon andmercury may also increase. Episodic, large-scale releases of organically bound mercurymay become a dominant feature accompany-ing permafrost degradation. Clearly, Arctic

lakes would be vulnerable, but increased inputof carbon is also projected for Arctic seas, sug-gesting an increased load of mercury, whichfollows the carbon, in the marine environment.Hudson Bay may be especially vulnerable dueto its large drainage basin and because perma-frost melting is likely in the area. Mercury con-centrations in snow have increased in this area,as have mercury fluxes to sediment.

Along the coasts, sea-level rise will promoteerosion, which could disturb contaminatedsites. It may also damage structures such aspipelines, thus releasing potentially contami-nating substances to the environment.


Aerial view of polygontundra, Lena Delta,Russia.

MA

GN

US

E

LA

ND

ER

Change in permafrost also has an implica-tion for radon that diffuses out of the ground.This radionuclide is not generally related toanthropogenic activities but comes from soilsand bedrock. Radon is trapped in frozenground in the Arctic, but with warmer temper-atures, more radon will diffuse out of soils,increasing the dose of this element and itsdecay products to people.

Glaciers could become sources of DDT

Glaciers have accumulated snow and ice overmillennia. They also act as reservoirs for someairborne contaminants. When the glaciersmelt, these contaminants can be re-emitted tothe air or be released in the meltwater. In theArctic, North American glaciers have been

shrinking since the 1960s. In the CanadianArchipelago, the glacial melt was exceptionallystrong in the 1990s, corresponding to the highArctic Oscillation Index. In the European Arc-tic, the trend is not as clear. Scandinavian glac-iers have grown during the 1990s, whereasmost Svalbard glaciers continue to shrink atthe same rate as they did throughout the 1900s.Russian glaciers may be retreating, but this isdifficult to establish because of limited data.Measurements from the Agassiz Ice Cap inCanada give a hint of the size of glaciers as apotential source for contaminants. For DDT,glacial melt may provide an important climate-modulated source. For HCHs and PCBs, thissource is small compared with the reservoir inthe Arctic Ocean.

Ocean transportThe Arctic Ocean is divided into distinct lay-ers. Below 800 meters is Arctic deep water,with a very long residence time. From 200 toabout 800 meters is the Atlantic Layer. At thevery top is the Arctic surface water, which isthe most important for contaminant transportwithin the Arctic Basin. Between the surfacewater and the Atlantic layer is the halocline,a transition zone of increasing salinity. The sig-nificance of ocean transport for contaminantsto, from, and within the Arctic has been in-creasingly recognized during the past fewyears. Currents are sluggish compared withwinds, and oceans therefore become importantlater in a contaminant’s history. However, theocean may have a much larger capacity tocarry contaminants than the air, allowing cur-rents eventually to catch up with and surpass

Glacier at Kangerlus-suatsiaq, West Green-land. The light greyzones at each side of theglacier show the formerextent.

Atlantic-Pacific front during high Arctic Oscillation Index, early 1990s

Normal placement of the Atlantic-Pacific front

0200400600800

1000

2000

3000

4000

Ice Polar mixed layer Pacific halocline Atlantic halocline

Arctic deep water

Atlantic layer

ca. 10 years

ca. 300 years

ca. 290 years

ca. 75 years

ca. 25 yearsca. 30 years

ca. 10 years

MakorovBasin

AmundsenBasin

NansenBasin

CanadaBasin

Bold figures denote residence times

AlphaRidge

NansenGakkelRidge

Lomo-nosovRidge

Pacific water

Atlantic water

Norwegian Seaand Greenland Sea

deep water

Bering Strait Fram StraitDepth, m 75°N 80°N 85°N 90°N 85°N 80°N

BeringStrait

FramStrait

The stratification ofthe Arctic Ocean.showing the polarmixed layer, thePacific and Atlanticdomains of influenceand the haloclines.The red lines showthe normal placementand the displacementof the Atlantic Pacificfront during the highArctic OscillationIndex of the early1990s.

PO

LA

R

PH

OT

OS

/

H

EN

NI

NG

T

HI

NG

atmospheric transport in importance. Some ofthe ocean pathways have already exhibitedchanges clearly related to the Arctic Oscillation.

New pathwaysin the Arctic Ocean surface waters

Surface ocean water pathways follow twobasic trajectories: the Transpolar Drift thatcrosses the Eurasian Basin and exits throughFram Strait, and the circulating BeaufortGyre on the North American side of theArctic Ocean (see figure on page 3). With ahigh Arctic Oscillation Index, water in theTranspolar Drift moves closer to NorthAmerica, while the Beaufort Gyre retreatsinto the Canadian Basin.

More important than changes in trajectoriesare the effects on the halocline. This is a transi-tion zone of increasing salinity, which serves asa barrier for transfer of heat and contaminantsfrom Arctic surface water to the Atlantic waterbelow. In the 1990s, the halocline in the Eur-asian Basin weakened. The most likely reasonwas that changes in wind patterns forcedfreshwater from the Russian rivers emptyinginto the Laptev and Kara Seas eastward, di-verting their flow toward the East SiberianShelf. The freshwater input to the Arctic Oceanis important for the development of stratifica-tion in the water column. A consequence ofthis diversion, therefore, would have been a

reduction in stratification in the EurasianBasin and increased stratification in the Can-adian Basin.

The diversion of the Russian river outflowaffects the transport of persistent organic pol-lutants both from the rivers and within the Arc-tic Ocean. Specifically, instead of entering theTranspolar Drift to exit the Arctic Ocean with-in about two years, the pollutants would enterthe Canadian Basin, which has a ten-year resi-dence time. Pollutants would thus stay in theArctic Ocean much longer, especially increas-ing the load in the Canadian Basin. Further-more, once in the Canadian Basin, pollutantsfrom Russian rivers might then exit via theCanadian Archipelago instead of via the westside of Fram Strait. The increased residencetime would lead to increased sedimentation,making it likely that more sediment-boundcontaminants would remain in the Arctic.

Driftwood from Siberiafound at Fleming Fjord,north of Ittoqqortoor-miit, East Greenland.

105

2000

600+

600+

200200

330

525

85

620430

330

?

Atlantic-Pacific Front Atlantic-Pacific

Front

Pre 1990Low Arctic Oscillation Index

River inflow, km3/year

Post 1990High Arctic Oscillation Index

River inflow, km3/year

1979

Freshwater runoff distribution

1990-1994

Low

Salinity

High

GreenlandAlaska

Russia

Changes in the distribu-tion of freshwater runoffin the Arctic Ocean be-tween low Arctic Oscil-lation Index, 1979, andhigh Arctic OscillationIndex, 1990-94 (uppermaps), and changes inthe amounts of river in-flow to the Arctic Oceanunder same conditions.

PO

LA

R

PH

OT

OS

/

H

EN

NI

NG

T

HI

NG

The Atlantic’s increased role leads todeclines in cadmium

For the Atlantic Layer, the Arctic Oscillationinfluences the flow of water into and out of theArctic Ocean. Communication with the Pacificis through Bering Strait, while communicationwith the Atlantic is through Fram Strait andthrough the Norwegian and Barents Sea. Im-portant contaminants in the Atlantic inflowinclude radionuclides from European repro-cessing plants, and any change in the flow ofAtlantic water may thus affect concentrationsand distribution of radionuclides in the Arctic

marine environment. In the past few decades,the North Atlantic Oscillation Index has in-creased, causing changes in distribution ofwater masses in the Nordic Seas. This hasbrought contaminants from the reprocessingplants closer to the Norwegian coast and intothe Barents Sea.

Traditionally, the Arctic Ocean has beenthought of as a quiet, steady-state system char-acterized by several relatively stable layers.During the 1990s, there were some spectacularchanges. The front between Atlantic and Pa-cific water was forced closer toward NorthAmerica, which increased the Atlantic’s area ofinfluence in surface water by some 20 percent.Water in the Atlantic layer is both warmer andsaltier than the Pacific water that it displaced.

The declining role of Pacific water in theArctic Ocean has implications for cadmium,a toxic metal that biomagnifies in the marinefood web. In the ocean, the distribution of thismetal is largely controlled by natural biogeo-chemical cycles, with the Pacific having highernatural concentrations than the Atlantic. Be-cause the Pacific inflow through the BeringStrait is a dominant source to the surfacewaters of the Arctic, reduced Bering inflowsince the 1940s has probably led to reductionin cadmium input. Furthermore, the encroach-ment of Atlantic water during the recent highArctic Oscillation Index will have reduced thedomain of Pacific water that is relatively en-riched with cadmium within the Arctic. Changesin upwelling or mixing are also likely to affectthe entry of cadmium into surface water fromdeeper layers.

Sea iceOne of the prominent features of the ArcticOcean is its ice cover. Changes in ice coverhave already occurred and the effects of thison persistent organic pollutants and mercurymay become increasingly important.


Main features of oceancirculation in the NorthAtlantic and the NordicSeas during high and lowNorth Atlantic Oscilla-tion Index.

Low Arctic Oscillation Index High Arctic Oscillation Index

Temperature

Warm

Cold

Greenland

Russia

Alaska

LabradorSea

IrmingerSea

GreenlandSea

IcelandSea

NorwegianSea

BarentsSea

High North Atlantic Oscillation Index

Low North Atlantic Oscillation Index

Atlantic waterArctic water

An exceptionally strongshift to high Arctic andNorth Atlantic Oscilla-tion Indices in about1989 increased the influ-ence of Atlantic water(red) in the Arctic basin.The Atlantic layer cur-rents are relatively fastand move water at a rateof 300-1600 kilometersper year along the mar-gins of the basin.

Less ice cover

During the 1990s, the scientific communityrecognized with some alarm that Arctic sea icehad retreated over the past three decades. Thechanges included a reduction in the area cov-ered by sea ice, an increase in the length of theice melt season, and a loss of multi-year ice.The rate of loss has been difficult to estimatebut is approximately 3 percent per decade.

Most of the ice has been clearing duringsummer over the shelves of the Eastern Arctic,north of Russia. Multi-year ice has decreasedeven more rapidly and been partly replaced byfirst-year ice. First-year ice melts more easily

than multi-year ice because it is thinner andsaltier. In the East Siberian and Beaufort Seas,there were unusually large areas of open waterin late summer at various times during the1990s. It appears that the marginal seas arebecoming only seasonally covered with ice,and that the extent of the permanent ice packis decreasing.

The loss of sea ice is consistent with whatcan be expected under a high Arctic Oscil-lation Index. Several factors are probablyinvolved including more heat being trans-ported to the pole by southerly winds. Evenmore important might be that winds causechanges in the distribution of ice. Ice-thicknessmeasurements made from submarines indicatethat the multi-year ice in the Central ArcticOcean has been getting thinner. Most of theinformation has been gathered in the interiorof the Arctic Ocean, and the decrease mightbe, at least in part, the product of a shift in thedistribution of multi-year ice toward NorthAmerica.

Sudden but temporary changes in ice coverhave occurred earlier in Arctic history. Over acentury ago, the whaling fleet experienced adramatic decrease in ice cover in the NorthAmerican Arctic. In the Barents Sea, about 15percent of sea ice cover was lost around 1920.

Increased exchange of POPsbetween sea and air

Some persistent organic pollutants have accu-mulated in the Arctic Ocean surface waters.The low temperatures of the Arctic, whichdecreases their volatility in air and increasestheir tendency to dissolve in water, acts as adriving force in moving them from the air tothe water. This pathway is especially importantfor compounds that prefer cold water, alpha-HCH being a prime example. Once these pollu-tants are in the water, they can become trappedunder the ice and retained in the water masses,some of which have long residence times.

�The change in ArcticOcean sea ice extentfrom 1979 to 1995showing the ice loss tobe predominantly a latewinter–summer phenom-enon.

��

��

Winter maximum Summer minimum

Total ice cover

Partial ice cover

Open water

Monthly averages

Monthly deviations

1995199019851980

J J JF M A A S O N DMMonth

8.0

Sea ice extent, million km2

Sea ice extent, change per decade, %

7.0

6.0

5.0

4.0

1.0

0.5

0

0

–0.5

–0.5

–1.0

–1.5

–2.0

–2.5

–3.0

–3.5

–4.0

0.5

–1.0

Winter maximum andsummer minimum Arcticsea ice cover as derivedfrom satellite imagery.


For alpha-HCH, discontinued use of thepesticide technical HCH has led to a drasticreduction in air concentrations. As a conse-quence, the ice-covered areas of the ArcticOcean became oversaturated relative to atmos-pheric levels. If the ice cover disappears, theseareas will become a source to the atmosphere.Other contaminants, such as PCBs and toxa-phene, are still loading into the Arctic Oceanfrom the air. The same loss of ice cover couldthus lead to increased loading of these twocontaminants into Arctic surface water.

Ice changes will affectmercury deposition in the Arctic

As described in the chapter Heavy Metals,mercury deposition in the Arctic increases dra-matically at polar sunrise due to an extraordi-nary set of circumstances. The phenomenon iscalled mercury depletion. Although the mech-anisms behind mercury depletion are not yetfully understood, results of investigations todate indicate that gaseous mercury in the airreacts with bromine compounds to form par-ticulate and reactive mercury. The brominecompounds are formed when bromine, emittedfrom seawater and sea salts, reacts with ozonein the presence of ultraviolet light (hence theconnection to the return of the sun). The reac-tive mercury that is produced is efficiently re-moved from the atmosphere and some of itremains in the snow. Some will eventually endup in meltwater and may thus enter aquaticecosystems. The sensitivity of the Arctic tomercury probably lies in the fact that meltwa-ter and runoff can drain into surfaces belowthe ice, where ice cover blocks the re-emissionof mercury to the air.

Change in sea ice cover can affect thisunique deposition mechanism if the availabil-ity of bromine is altered. Initially, it is likelythat climate change will contribute to increas-ing the amount of first-year ice around thepolar margins, leading to saltier ice and snow.It could thus enhance the emission of bromineand possibly extend the area of mercury deple-tion events.

With further climate change, parts of theArctic will become more temperate. Mercury

deposition would decrease, and at the sametime more mercury could escape back to theatmosphere after being deposited. The endresult would be less accumulation of mercuryin marine and aquatic environments. It isharder to predict whether levels in biotawould also change. Mercury biomagnifiesand its levels depend on the structure of thefood web. Changes in the food web structurecould, therefore, be much more importantthan changes in physical pathways. Mercurylevels may also be affected by changes in per-mafrost, and increase with an increased fluxof organic carbon to both freshwater andmarine environment. In summary, the com-plexity of mercury pathways combined withobvious sensitivities to climate change shouldalert us to the possibility of surprises in thefuture.

Shifting routes from drifting sea ice

The general patterns of ice drift have been rec-ognized since the beginning of the 1900s, andfollow the same trajectories as ocean surfacewater in the Transpolar Drift and the BeaufortGyre. Only recently have these ice trajectoriesbeen mapped in detail. The new data suggestthat there are two characteristic modes of icemotion, one during a low Arctic OscillationIndex and one during a high index. During alow index, which prevailed from the 1960s tothe 1980s, the Transpolar Drift moves ice di-rectly from the Laptev Sea across the EurasianBasin and into the Greenland Sea. By contrast,during a high index, the prevailing situationfor much of the 1990s, this ice transport routeis diverted or splits. Some goes to the Green-land Sea and some moves across the Lomono-sov Ridge and into the Canadian Basin. At thesame time, the Beaufort Gyre shrinks back intothe Beaufort Sea and becomes disconnectedfrom the rest of the Arctic Ocean. This meansthat it exports less ice to the East Siberian Seaand only imports a little ice from north of theCanadian Archipelago.

The changes in ice-drift patterns have impli-cations for the transport of sediment and anycontaminants trapped in the ice. Specifically,when ice moves away from the East Siberian


Ice drift patterns foryears with low and highArctic Oscillation Index.

Beaufort Gyre

Russia

Canada

Greenland

Beaufort Gyre

Transpolar Drift Transpolar Drift

Low Arctic Oscillation Index High Arctic Oscillation Index

and Laptev Seas, new thin ice can form closeto the coast, increasing the opportunity forsediments to be trapped. Moreover, less icemoves from the North American to the Eur-asian part of the Arctic Ocean.

Biological impactsClimate change will have impacts on plantsand animals in the Arctic and thus on the bio-logical pathways for contaminants. Althoughwe can infer the types of changes that arelikely to occur, we cannot predict their scopeand timing. The following are examples ofprocesses that should be examined further.

Changing plant cover will affect deposition

Vegetation provides surfaces onto which air-borne contaminants can deposit when airmasses pass over the land. Forests, for exam-ple, have a unique ability to take chemicalsfrom the air via foliage and thence to a long-term reservoir in the soil.

Warmer winters will promote growth ofwoody shrubs and stimulate a northwardmigration of the treeline. So far, there is noevidence of large changes on the Arctic tundra.However, if permafrost melts and the watertable changes, such changes could occur muchmore rapidly in the Arctic than in other regionsof the world.

Aquatic ecosystems are sensitive to changes

Not only temperature but also changes in lightand the flow of nutrients will affect freshwaterecosystems. For example, loss of permafrostwill increase inputs of nutrients from the sur-rounding soil. Spring algae blooms will prob-ably come earlier.

In the summer, increased water temperaturewill negatively affect fish species that are sensi-tive to temperature or have temperature thresh-olds during their spawning. Each species has tobe evaluated separately, but trout and graylingare known to be sensitive. Increased wintertemperatures will enhance microbial decompo-sition. Insects, phytoplankton, and zooplank-ton will also be affected, some positively andsome adversely.

Along the North American Arctic coast, theloss of estuarine ice may displace cisco, whichmight be replaced by anadromous fish fromthe Pacific Ocean.

For marine fish, it is well known thatchanges in climate or ocean currents can affectthe distribution of commercially importantstocks, such as Atlantic cod and herring. Watertemperatures are important, as are the distrib-ution of prey and predators and the currentsthat determine the movement of larvae from

spawning grounds to their nursery areas.Atlantic cod and its main food item capelinare likely to move northeastwards. Spring-spawning herring may return to the samemigration route they followed in the mid-1960s, when the water temperature aroundIceland was higher than today. Other moresoutherly species may become distributed

farther north toward and into the Arctic. Thiswill lead to the introduction of new speciesinto the Arctic marine ecosystem. The conse-quences are difficult to predict but may includechanges in the food web and thus in the loadof contaminants in biota. Another possibilityis changes in migratory routes and contami-nants along the route.

Changes in sea icecan alter marine ecosystems

In marine ecosystems, many contaminants arebiomagnified in food webs, particularly thosewith many trophic levels. Therefore, anychanges in food web structure can potentiallyhave a large impact on contaminant burdens intop predators. The changes can be initiated atthe bottom of the food web, for example ifchanges in light and nutrient cycles alter con-ditions for phytoplankton and zooplankton.Food-web changes can also be initiated at thetop, by altering predation patterns, for exam-ple among bears and seals.

The amount of sea ice influences both lightconditions and the distribution of nutrients inthe water. Change in stratification of thewater column is important in this respect anda decrease in mixing of water layers hasalready been noted in the Greenland Sea andin the Canadian Basin. The availability ofnutrients influences the algae that are respons-ible for primary production at the bottom ofthe food web. This is true for the phytoplank-ton in the water column and also for the algaethat grow on the bottom of the ice and sup-port a unique ice-associated food web. Some


Possible changes in thedistribution of fish spe-cies if the seawater tem-perature increases 1-2 °C.

Herring

Cod

Capelin

Mackerel

North Seaherring

Anchovy, sardine

Mackerel,blue tuna

N O R W E G I A N S E A

of the algal production falls to the bottom ofthe ocean, where it supports the benthic foodweb. The distribution of sea ice thus has amajor impact on the distribution of organicmatter between the water column and theseabed.

The Beaufort and Chukchi Seas, crossedduring the drift of the SHEBA (Surface HeatBudget of the Arctic Ocean) Project in 1997-98, provide a dramatic example of a large-scale bottom-up change in the marine foodweb. Compared with a study 20 years earlier,this new close look at life in the water revealeda marked decrease in large diatoms and largemicrofauna within the ice. The high ArcticOscillation Index of the 1990s had divertedriver water into this area, causing a strongstratification of the surface waters. The resultwas a decrease in the supply of nutrients frombelow, and a species composition that wasmore typical of freshwater ecosystems. Theloss of large diatoms could potentially producea shift toward smaller zooplankton grazers,perhaps then introducing an extra step at thebottom of the food web.

The Bering Sea provides another recentexample of how bottom-up changes can per-meate an entire ecosystem. In 1997-98, therewere massive blooms of small phytoplankton.Because they were smaller than the diatomsthat typically bloom in the Bering Sea, theywere grazed on by copepods instead ofeuphausiids. The short-tailed shearwater nor-mally feeds on the euphausiids, and the lack offood may have contributed to a large die-off of

this bird species. The change in quantity ofdifferent zooplankton probably also decreasedfood availability for fish, whales, seals, andwalrus, causing die-offs and long-distancedisplacements.

Loss of sea ice would lead to Arctic shelfseas looking more like temperate seas. Theimplications for food web structures are verydifficult to predict and we should be preparedfor surprises. One such warning sign was themassive blooms of jellyfish in the Bering Seaduring the 1990s. Large-scale changes pro-duced by the Arctic Oscillation have the poten-tial to alter the balance between upwelling anddownwelling along the coast, through changesin either the distribution of ice cover or inaverage wind speed and direction. Shifts in theArctic Oscillation thus have the capacity tocause large-scale shifts in shelf ecosystems. Inregions that are important for commercial fish-eries, such changes can have major impacts onthe regional economy.

Sea ice is also a crucial habitat for manyspecies at the top of the food web. Ringedseals need landfast ice for pupping, which inturn influences the migration of polar bearsthat feed on the ringed seals. A decrease insuitable habitat for ringed seals to pup couldlead to declines in their populations, with thepossible consequence that polar bears could beforced to find other food sources or starve.Ringed seals feed on Arctic cod. If changes inthe ice alter the balance between seals andpolar bears, they would likely affect the Arcticcod as well.

Walrus provide an excellent example tochallenge our predictive capability. Most wal-rus feed on bottom-dwelling organisms andare thus fairly low in the food web. Some wal-rus, however, are known to eat seals, and theirhigher position in the food web is reflected inhigher contaminant levels. Many walrus usedrifting ice for their haulouts because it pro-vides good access to nearby feeding areas,reducing the amount of energy required tofeed. If the summer ice edge retreats north ofthe relatively shallow areas where walrus canfeed, as happened in the summer of 1998 inthe Chukchi Sea, the walrus may be forcedeither to starve or to prey on seals. The latteradaptation would place walrus much higher inthe food web.

Less sea ice could, however, benefit otherspecies. Eiders, which also feed on the ben-thos, need open water in which they can dive.By benefiting some species and hindering oth-ers, the loss of sea ice is likely to cause majoralterations in the marine food web. This isparticularly true for changes caused in certainkey species. Arctic cod, for example, plays acentral role linking lower levels of the foodweb to seals, beluga, and many birds. Anychanges to Arctic cod abundance or distribu-tion could propagate both up and down thefood web.


Walrus grazing on mussels.Most walrus feed low in thefood web, for example bygrazing on mussels. How-ever, some individuals huntseals, thus receiving highercontaminant intakes. If cli-mate change were to causea shift in feeding habits, itwould thus have implica-tions for contaminant lev-els in this species.

M A G N U S E L A N D E R

Human activities will increase

Climate change will inevitably bring changes tohuman activities in the Arctic, with subsequenteffects on contaminant loads and pathways.

For people, food habits have a great impacton exposure to contaminants. Changes inhunting opportunities because of changed ani-mal distribution and availability or changedability to travel over ice or land will thus havean impact. If a hunted animal is suddenlyhigher in the food web, its contaminant loadcould increase, thus increasing exposure evenfor people whose food habits remain the same.

A warmer Arctic with less sea ice will alsoencourage shipping, tourism, and oil exploita-tion, all of which increase the risk for contami-nation of new areas. More severe stormswould further increase risks connected to ship-ping and other offshore activities. The expan-sion of commercial fisheries from the Arcticmarginal seas into the Arctic Ocean wouldalso likely affect food web structure and rela-tive abundance of many species. Although thenet effects of changes in human activities andbehavior in the Arctic are impossible to predictwith confidence, changes are certain to occur.They, in turn, will affect sources, pathways,and eventual fate of contaminants in the Arc-tic, including human exposure.

SummaryLong term-climate change and natural climatecycles affect the transport of contaminants toand within the Arctic. The 1990s provided anexample of how widespread change can rapidlypervade much of the Arctic including winds,weather patterns, ocean currents, and sea ice.It is, however, difficult to predict whether long-term climate change will lead to a generallydecreased or increased contaminant load.

Some pathway changes clearly lead to moreefficient transport, one example being increased

transport of airborne pollutants from easternNorth America and Eurasia. Another exampleis Atlantic water carrying more radionuclidesfrom European processing plants. Lead thathas been deposited in the ocean to the westof Europe would also follow this pathway.A third example is the longer residence timein the Arctic Ocean for contaminants that arecarried by ocean surface waters.

Long-term climate changes are likely toaffect pathways that are influenced by sea ice.Such pathways will be important for manypersistent organic pollutants that partiallydissolve in water, some of which are currentlytrapped under the ice. Mercury is likewisetrapped under ice. For mercury, changes in seaice cover may also influence newly discoveredphysical pathways that enhance the depositionof mercury to surfaces.

Changes in lake ice and permafrost willaffect lake hydrology, potentially increasingthe input of contaminants into freshwaterecosystems and possibly releasing contami-nants that have accumulated in soil or havebeen improperly disposed of in earlier times.

Many contaminants pose a problem in theArctic because they biomagnify in food webs.Changes in food web structure, therefore,have a great potential to alter contaminantlevels in top predators. However, the com-plexity of ecosystems and our incompleteunderstanding of the dependence of manyspecies on habitats like sea ice make it espe-cially difficult to predict change, and oneshould expect surprises.

A final conclusion is that the load of per-sistent organic pollutants, heavy metals, andradionuclides in the Arctic is dependent onmany factors that operate after the contami-nant has been released from its source. In thelong run, anthropogenic emissions that affectthe climate may become as important as theemissions of the contaminants themselves indetermining the extent to which these con-taminants reach and affect the Arctic.


Tourists visiting theNorth Pole on an ice-breaker cruise take thePolar Plunge.B

RY

AN

&

C

HE

RR

Y

AL

EX

AN

DE

R

Changing Pathways - Arctic Monitoring and Assessment ...

Documents

Transcript of Changing Pathways - Arctic Monitoring and Assessment ...