A
M
A
P
AMAP Assessment 2006:
A
s
s
e
Acidifying Pol utants, Arctic Haze,
s
s
m
e
and Acidification in the Arctic
n
t
2
0
0
6
:
A
c
i
d
i
f
y
i
n
g
P
o
l
u
t
a
n
t
s
,
A
r
c
t
i
c
H
a
z
e
,
a
n
d
A
c
i
d
i
fi
c
a
t
i
o
n
i
n
t
h
e
A
r
c
t
i
c
Arctic Monitoring and Assessment Programme (AMAP)
ISBN 82-7971-046-9
I s l a n d s
n
Pacific Ocean
i a
t
C
u
e
l
Pribilof Islands
A
I
Okhotsk Sea
T
Gulf of
Bering Sea
r
Alaska
i
ve
E
C
Prince William Sound
Anchorage
Rn
Bering
kou
Strait
Y
CHUKOTKA
A
R
Fairbanks U S A
Alaska
S
Kolyma
Chukchi
S
A
Range
Sea
Wrangel Island
I
Y U K O N
Indigirka
B
rooks
Barrow
A
B
East
t
h
P
Mackenzie River
e
Prudhoe Bay
Siberian
T
a
E
b
a
a
ce
Sea
s
Beaufort Sea
c
R
a
iv
N O R T H W E S T
e
R
New
R
e
Great
Canada
g
iv
r
T E R R I T O R I E S
e
C
d
Siberian
Tiksi
E
N
r
Slave
a Basin
i
Lena
Slave R.
( N W T )
n
R
Islands
Lake
a
I
d
v
Amundsen Gulf
C A N A D A
i
e
a
le
n
Laptev
A
R
R
Victoria Island
B
Sea
a
ende
Bathurst
Island
A r c t i c s
M Makarov
i
O c e a n
n
Basin
Kh
N U N A V U T
-
atang
E
a
N
R.
Churchill
Severnaya
Taymir
Alpha
B a s i n Zemlya
Peninsula
D
Amundsen
L
a
e
n
v
onosov Ridge
Basin
o
Norilsk
T
c
n
Yenisey
a
Ellesmere Island
s
I
Lom
Hudson Bay
t
e
s
Canadian Arctic Archipelago
r
.
Agassiz Ice Cap
S
R U S S I A
o
akkel Ridge
u
Alert
Foxe Basin
nd
r a s i a n
Franz
S
A
u
Josef
Baffin
Nansen
E
Land
Island
it
Baffin
Nansen-G
Basin
Novaya
tra
Bay
Station Nord
Zemlya Kara
S
Fram
E
n
Sea
Vorkuta
R
Q U E B E C
Salekhard
Strait
so Iqaluit
d
Svalbard
Us
u
a
Ob
Ny-Ålesund
H
Lomonosovfonna
Pechora
G R E E N L A N D
Davis Strait
Barents
berland Sound
Sea
W
C
U
um
Sea
P
R
C
Spitsbergen
ec
Bjørnøya
ho
A
r
L A B R A D O R
Greenland
a
L
Nuuk
S
Sea
Murmansk
Labrador
T
Nikel
Kola
Sea
Jan Mayen
Peninsula
Tromsø
Lapland
White
Arkhangelsk
Denmark
Oulanka Sea
Severnaya Dvina
I
Strait
Abisko
Karelia
C
Reykjavik ICELAND
Norwegian
Sea
FINLAND
St. Petersburg
Helsinki
Moscow
FAROE
Tórshavn ISLANDS
SWEDEN
NORWAY
Stockholm
Oslo
Atlantic Ocean
Chernobyl
Skager ak
North DENMARK
Sea
-2000
-500
0
50
100
200
300
500
1000
1500
2000
3000 4000 m
AMAP Assessment 2006:
Acidifying Pollutants, Arctic Haze,
and Acidification in the Arctic
Arctic Monitoring and Assessment Programme (AMAP), Oslo, 2006
ii
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
ISBN: 82-7971-046-9
© Arctic Monitoring and Assessment Programme, 2006
Published by
Arctic Monitoring and Assessment Programme (AMAP), P.O. Box 8100 Dep, N-0032 Oslo, Norway (www.amap.no)
Citation
Whole report: AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic. Arctic Monitoring and
Assessment Programme (AMAP), Oslo, Norway. xii + 112pp.
Subsection: e.g.: Bishop, K., M. Forsius, Ø. Kaste, H. Laudon, T. Moiseenko and B.L. Skjelkvåle, 2006. Chapter 6.3. Episodic acidifica-
tion. In: AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic, pp. 78 - 81. Arctic Monitoring and
Assessment Programme (AMAP), Oslo, Norway.
Ordering
AMAP Secretariat, P.O. Box 8100 Dep, N-0032 Oslo, Norway
This report is also published as electronic documents, available from the AMAP website at www.amap.no
Production
Scientific, technical, and linguistic editing
Carolyn Symon (carolyn.symon@btinternet.com) and Simon Wilson (AMAP Secretariat)
Lay-out and technical production management
Satu Turtiainen, Finnish Environment Institute, P.O.Box 140, 00251 Helsinki, Finland
Design and production of computer graphics
Satu Turtiainen, Erika Varkonyi, Petri Porvari and Marjut Nyman, Finnish Environment Institute
Cover Photo
Dan Aamlid
Printing and binding
Vammalan Kirjapaino Oy, Vammala, Finland
AMAP Working Group:
John Calder (Chair, USA), Yuri Tsaturov (Vice-chair, Russia), Per Døvle (Vice-chair, Norway), Russel Shearer (Canada), Morten Olsen
(Denmark), Outi Mähönen (Finland), Helgi Jensson (Iceland), Gunnar Futsæter (Norway), Cynthia de Wit (Sweden), Jan-Idar Solbakken
(Permanent Participants of the Indigenous Peoples Organizations).
AMAP Secretariat:
Lars-Otto Reiersen, Simon Wilson, Yuri Sychev, Inger Utne.
Indigenous peoples organizations, AMAP observing countries, and international organizations:
Aleut International Association (AIA), Arctic Athabaskan Council (AAC), Gwitch'in Council International (GCI), Inuit Circumpolar
Conference (ICC), Russian Association of Indigenous Peoples of the North (RAIPON), Saami Council.
France, Germany, Netherlands, Poland, United Kingdom.
Advisory Committee on Protection of the Sea (ACOPS), Association of World Reindeer Herders (AWRH), Circumpolar Conservation
Union (CCU), European Environment Agency (EEA), International Arctic Science Committee (IASC), International Arctic Social Sciences
Association (IASSA), International Atomic Energy Agency (IAEA), International Council for the Exploration of the Sea (ICES), Interna-
tional Federation of Red Cross and Red Crescent Societies (IFFCRCS), International Union for Circumpolar Health (IUCH), International
Union for the Conservation of Nature (IUCN), International Union of Radioecology (IUR), Nordic Council of Ministers (NCM), Nordic
Council of Parliamentarians (NCP), North Atlantic Marine Mammal Commission (NAMMCO), Northern Forum (NF), OECD Nuclear
Energy Agency (OECD/NEA), OSPAR Commission (OSPAR), Standing Committee of Arctic Parliamentarians (SCAP), United Nations
Economic Commission for Europe (UN ECE), United Nations Environment Programme (UNEP), World Health Organization (WHO),
World Meteorological Organization (WMO), World Wide Fund for Nature (WWF).
AMAP data centers:
International Council for the Exploration of the Sea (ICES), Norwegian Institute for Air Research (NILU), Norwegian Radiation Protection
Authority (NRPA), University of Alaska Fairbanks (UAF).
iii
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
4.3.1. Direct effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi
4.3.2. Indirect effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Executive Summary to the AMAP Arctic Pollution 2006
4.3.3. Surface albedo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Ministerial Report. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ix
4.4. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Chapter 1 · Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Chapter 5 · Effects on Terrestrial Ecosystems . . . . . . . . 41
5.1. Effects on soils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
5.1.1. Acidity status of soils on the Kola Peninsula . . . . . . . . . 41
Chapter 2 · Sources of Acidifying Pollutants
5.1.1.1. Natural factors affecting soil acidity. . . . . . . . . . 42
and Arctic Haze Precursors . . . . . . . . . . . . . . . . 2
5.1.1.2. Sulfur dioxide emissions and soil acidity . . . . . 43
5.1.1.3. The role of overburden and
2.1. Sources within the Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
bedrock chemistry . . . . . . . . . . . . . . . . . . . . . . . . . 44
2.1.1. Stationary sources: industry and energy. . . . . . . . . . . . . . 2
5.1.1.4. Connections between soil condition and
2.1.2. Local air pollution in Russian cities . . . . . . . . . . . . . . . . . . 4
ecosystem quality . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.1.3. Oil and gas activities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
5.1.1.5. Temporal trends in soil acidity . . . . . . . . . . . . . . 48
2.1.4. Shipping activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
5.1.2. Acidification and the acidity status of soils
2.1.5. Natural sources within the Arctic: wildfires . . . . . . . . . . 7
in the Norilsk area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.2. Sources outside the Arctic and atmospheric transport
5.1.3. Effects on soil micro-organisms . . . . . . . . . . . . . . . . . . . . 48
to the Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
5.2. Effects on vegetation in the European Arctic . . . . . . . . . . . . . 50
2.3. Emissions estimates used in modeling . . . . . . . . . . . . . . . . . . . 9
5.2.1. Lichen-dominated and mountain
birch (tundra) ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . 50
5.2.2. Coniferous forest ecosystems . . . . . . . . . . . . . . . . . . . . . . 53
Chapter 3 · Concentrations and Deposition of
5.2.3. Reindeer grazing, climate change, nitrogen
Acidifying Pollutants . . . . . . . . . . . . . . . . . . . . . . 11
deposition, and other factors. . . . . . . . . . . . . . . . . . . . . . . 55
5.2.4. Needs and recommendations for future
3.1. Atmospheric and transport processes for
research and monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . 56
air pollutants in the Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.3. Effects on fauna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.1.1. Sulfur . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.3.1. Effects on birds and mammals . . . . . . . . . . . . . . . . . . . . . 57
3.1.2. Nitrogen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
5.3.1.1. Mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
3.2. Distribution of monitoring stations . . . . . . . . . . . . . . . . . . . . . 12
5.3.1.2. Birds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.3. Concentrations, distribution, and
5.3.1.3. Concluding comments on birds and
trends in air and precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . 15
mammals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3.1. Air. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.3.2. Effects on invertebrates. . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
3.3.2. Precipitation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
5.3.2.1. Size, individual performance,
3.3.2.1. General pattern. . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
and population structure . . . . . . . . . . . . . . . . . . . 59
3.3.2.2. Russian Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.3.2.2. Changes in population densities. . . . . . . . . . . . . 60
3.4. Episodes and exposure to sulfur and nitrogen . . . . . . . . . . . . 21
5.3.2.3. Changes in species richness, diversity,
3.5. Concentrations in seasonal snow cover . . . . . . . . . . . . . . . . . . 22
and community structure . . . . . . . . . . . . . . . . . . 61
3.5.1. General pattern . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.3.2.4. Concluding comments on invertebrates . . . . . . 62
3.5.2. Russian Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.4. Critical loads of acidity and their exceedance . . . . . . . . . . . . . 62
3.6. Pollution history from ice cores and lake sediments. . . . . . . 24
3.7. Modeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.7.1. Validation of the system for temporal trend analysis . . 26
Chapter 6 · Effects on Freshwater Ecosystems . . . . . . . 64
3.7.2. Trend analysis based on measurements
at Station Nord and DEHM model results . . . . . . . . . . . 28
6.1. Evidence from water quality monitoring . . . . . . . . . . . . . . . . . 64
3.7.3. Effects of natural climate variations on
6.1.1. Current status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
long-range transport to the Arctic . . . . . . . . . . . . . . . . . . 29
6.1.1.1. Northern Fennoscandia and
3.7.4. Scenarios . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
the Kola Peninsula . . . . . . . . . . . . . . . . . . . . . . . . . 68
6.1.1.2. Iceland . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
6.1.1.3. Svalbard and Bear Island . . . . . . . . . . . . . . . . . . . 68
Chapter 4 · Arctic Haze. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.1.1.4. New critical loads and exceedance calculations
for the Euro-Arctic Barents region . . . . . . . . . . . 69
4.1. The arctic haze phenomenon . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.1.1.5. Current status in Arctic Canada . . . . . . . . . . . . . 70
4.2. Trends in arctic haze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.1.1.6. Naturally acidic lakes in Arctic Canada. . . . . . . 70
4.2.1. Chemical composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
6.1.1.7. Alaska . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.2.2. Optical properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
6.1.1.8. Northern Russia, Siberia . . . . . . . . . . . . . . . . . . . . 71
4.3. Effects of aerosol on the climate system in the Arctic. . . . . . 38
6.1.2. Temporal trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
iv
6.1.2.1. Lakes in Finland, Norway, and Sweden. . . . . . . 71
6.5.2.1. Processes in air . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
6.1.2.2. Lakes on the Kola Peninsula . . . . . . . . . . . . . . . . 73
6.5.2.2. Processes in terrestrial areas . . . . . . . . . . . . . . . . 89
6.1.2.3. Swedish repeated lake survey . . . . . . . . . . . . . . . 74
6.5.2.3. Processes in surface waters . . . . . . . . . . . . . . . . . 89
6.1.2.4. Concluding comments on trends . . . . . . . . . . . . 74
6.2. Effects of acidification on arctic biota . . . . . . . . . . . . . . . . . . . . 74
6.2.1. Current status. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Chapter 7 · Air Pollution and Health Impacts
6.2.1.1. Phytoplankton and periphyton . . . . . . . . . . . . . . 75
in the Arctic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
6.2.1.2. Macroinvertebrates . . . . . . . . . . . . . . . . . . . . . . . . 75
6.2.1.3. Fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6.2.2. Temporal trends. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.1. Major air pollutants of health concern . . . . . . . . . . . . . . . . . . . 91
6.2.2.1. Invertebrates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.2. Key epidemiological findings . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.2.2.2. Fish . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
7.3. The arctic perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
6.3. Episodic acidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
7.4. The shifting panorama . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
6.3.1. Acidic episodes in the Kola region . . . . . . . . . . . . . . . . . . 78
6.3.2. Acidic episodes in the Dalelva catchment
in eastern Finnmark, Norway . . . . . . . . . . . . . . . . . . . . . . 78
Chapter 8 · Conclusions and Recommendations . . . . . 94
6.3.3. Acidic episodes in northern Sweden . . . . . . . . . . . . . . . . 79
6.3.4. Concluding comments on episodic acidification. . . . . . 80
8.1. Sources of acidifying pollutants and
6.4. Evidence from paleolimnological studies . . . . . . . . . . . . . . . . 81
arctic haze precursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.4.1. Millennial trends in lake acidification. . . . . . . . . . . . . . . 81
8.2. Trends in air concentrations and deposition . . . . . . . . . . . . . . 94
6.4.1.1. Fennoscandia and the Kola Peninsula . . . . . . . . 81
8.2.1. Air and precipitation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
6.4.1.2. Concluding comments on
8.2.2. Model projections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
millennial-scale acidification . . . . . . . . . . . . . . . . 82
8.3. Arctic haze . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.4.2. Recent acidification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
8.4. Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.4.2.1. Fennoscandia and the Kola Peninsula . . . . . . . . 82
8.4.1. Human health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.4.2.2. Siberia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.4.2. Terrestrial ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
6.4.2.3. Svalbard. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
8.4.3. Freshwater ecosystems . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
6.4.2.4. Concluding comments on
8.5. Links between acidification, arctic haze,
recent acidification. . . . . . . . . . . . . . . . . . . . . . . . . 84
and other environmental issues . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.5. Interaction between acidification and
8.6. Gaps in knowledge and recommendations
other environmental issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
concerning monitoring and research needs. . . . . . . . . . . . . . . 97
6.5.1. Interactions concerning climate change and
8.6.1. Geographical gaps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
UV radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8.6.2. Data availability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97
6.5.1.1. Anticipated changes in climate . . . . . . . . . . . . . . 85
8.6.3. Trends in air and precipitation . . . . . . . . . . . . . . . . . . . . . 97
6.5.1.2. Anticipated changes in hydrology and
8.6.4. Pathways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
water quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
8.6.5. Effects monitoring and research. . . . . . . . . . . . . . . . . . . . 98
6.5.1.3. Recovery from acidification in
8.6.6. Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
surface waters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
6.5.1.4. Impacts of DOC changes on
UV radiation in lakes. . . . . . . . . . . . . . . . . . . . . . . 88
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
6.5.2. Interactions concerning heavy metals/
contaminants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
v
Preface
This report details the results of the 2006 AMAP assessment
bility of the scientific experts involved in the preparation of the
of Acidifying Pollutants, Arctic Haze, and Acidification in the
assessment. The lead country for the AMAP Acidification and
Arctic. It builds upon the previous AMAP acidification assess-
Arctic Haze Assessment under AMAP phase II was Finland.
ment presented in `AMAP Assessment Report: Arctic Pollution
The assessment is based on work conducted by a large number of
Issues'* that was published in 1998.
scientists and experts from the Arctic countries (Canada, Den-
mark/Greenland/Faroe Islands, Finland, Iceland, Norway, Rus-
The Arctic Monitoring and Assessment Programme (AMAP) is
sia, Sweden, and the United States), together with contributions
a group working under the Arctic Council.
from indigenous peoples' organizations, from other organiza-
tions, and from experts in other countries.
The Arctic Council Ministers have requested AMAP:
· to produce integrated assessment reports on the status and
AMAP would like to express its appreciation to all of these ex-
trends of the conditions of the Arctic ecosystems;
perts, who have contributed their time, effort, and data; and
· to identify possible causes for the changing conditions;
especially to the lead experts who coordinated the production of
· to detect emerging problems, their possible causes, and the
this report, and to referees who provided valuable comments and
potential risk to Arctic ecosystems including indigenous
helped ensure the quality of the report. A list of the main con-
peoples and other Arctic residents; and
tributors is included in the acknowledgements on pages vi - vii of
· to recommend actions required to reduce risks to Arctic
this report. The list is not comprehensive. Specifically, it does not
ecosystems.
include the many national institutes, laboratories and organiza-
tions, and their staff, which have been involved in the various
This report provides the accessible scientific basis and validation
countries. Apologies, and no lesser thanks, are given to any in-
for the statements and recommendations made in the AMAP
dividuals unintentionally omitted from the list. Special thanks
Overview report, `Arctic Pollution 2006: Acidification and Arc-
are due to the lead authors responsible for the preparation of the
tic Haze'** that was delivered to Arctic Council Ministers at
various chapters of this report.
their meeting in Salekhard, Russia in October 2006. It includes
extensive background data and references to the scientific lit-
The support of the Arctic countries is vital to the success of
erature, and details the sources for figures reproduced in the
AMAP. AMAP work is essentially based on ongoing activities
Overview report. It also includes conclusions and recommenda-
within the Arctic countries, and the countries also provide the
tions of a scientific nature, such as proposals for filling gaps in
necessary support for most of the experts involved in the prepa-
knowledge, and recommendations relevant to future monitoring
ration of the assessments. In particular, AMAP would like to
and research work. Some of these are taken up in the Overview
express its appreciation to Finland for undertaking the lead role
report, although that report focuses more on recommendations
in supporting the Acidification and Arctic Haze assessment.
that specifically focus on actions aimed at improving the Arctic
Special thanks are also offered to the Nordic Council of Ministers
environment.
for their financial support to the work of AMAP, and to sponsors
of other bilateral and multilateral projects that have delivered
To allow readers of this report to see how AMAP interprets and
data for use in this assessment.
develops its scientifically-based assessment product in terms of
more action-oriented conclusions and recommendations, the `Ex-
The AMAP Working Group that was established to oversee this
ecutive Summary of the Arctic Pollution 2006 overview report'
work, and the AMAP Acidification and Arctic Haze assessment
is reproduced in this report on pages ix to xi.
group are pleased to present its assessment.
The AMAP assessment is not a formal environmental risk as-
sessment. Rather, it constitutes a compilation of current knowl-
John Calder
edge about the Arctic region, an evaluation of this information
AMAP Working Group Chair
in relation to agreed criteria of environmental quality, and a
statement of the prevailing conditions in the area. The assess-
Martin Forsius
ment presented in this report was prepared in a systematic and
AMAP Acidification and Arctic Haze assessment lead
uniform manner to provide a comparable knowledge base that
(Finland)
builds on earlier work and can be extended through continuing
work in the future.
Lars-Otto Reiersen
AMAP Executive Secretary
The AMAP scientific assessments are prepared under the direc-
tion of the AMAP Working Group. The product is the responsi-
Oslo, July 2006
* AMAP, 1998. AMAP Assessment Report: Arctic Pollution Issues. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway.
xii+859 p.
** AMAP, 2006. Arctic Pollution 2006: Acidification and Arctic Haze. Arctic Monitoring and Assessment Programme (AMAP), Oslo, Norway.
vi
Acknowledgements
The AMAP Working Group would like to thank the following persons for their work in preparing the AMAP Assessment 2006:
Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic.
Assessment Lead:
Martin Forsius, Finnish Environment Institute, Helsinki, Finland
Scientific Secretary:
Marjut Nyman, Finnish Environment Institute, Helsinki, Finland
Scientific Authors:
Chapter 1 · Introduction:
* Martin Forsius
Chapter 2 · Sources of Acidifying Pollutants and Arctic Haze Precursors:
* Lars R. Hole, Norwegian Institute for Air Research, Tromsø, Norway
Jesper Christensen, National Environmental Research Institute, Roskilde, Denmark
Martin Forsius
Marjut Nyman
Andreas Stohl, Norwegian Institute for Air Research, Norway
Simon Wilson, AMAP Secretariat, Oslo, Norway
Chapter 3 · Concentrations and Deposition of Acidifying Air Pollutants:
* Lars R. Hole
Jesper Christensen
Veronika A. Ginzburg, Russian Federal Service for Hydrometeorology and Environmental Monitoring and Russian Academy of
Sciences, Moscow, Russia
Vladimir Makarov, Melnikov Permafrost Institute, Yakutsk, Russia
Natalia A. Pershina, Main Geophysical Observatory Voeikov, St. Petersburg, Russia
Alla I. Polischuk, Main Geophysical Observatory Voeikov, St. Petersburg, Russia
Tuija Ruoho-Airola, Finnish Meteorological Institute, Helsinki, Finland
Peotr Ph. Svistov, Main Geophysical Observatory Voeikov, St. Petersburg, Russia
Vitaly N. Vasilenko, Institute of Global Climate and Ecology, Moscow, Russia
Chapter 4 · Arctic Haze:
* Patricia Quinn, National Oceanic and Atmospheric Administration (NOAA), PMEL, Seattle, USA
Betsy Andrews, NOAA, CMDL, Boulder, USA
Ellsworth Dutton, NOAA, CMDL, Boulder, USA
Tuija Ruoho-Airola
Glenn Shaw, University of Alaska, Fairbanks, USA
Chapter 5 · Effects on Terrestrial Ecosystems:
* John Derome, Finnish Forest Research Institute, Rovaniemi, Finland
* Sirkku Manninen, University of Helsinki, Helsinki, Finland
Julian Aherne, University of Trent, Ontario, Canada
Paavo Hellstedt, University of Helsinki, Helsinki, Finland
Jean-Paul Hettelingh, Netherlands Environmental Assessment Agency, Bilthoven, The Netherlands
Kevin Hicks, Stockholm Environment Institute at York, University of York, Heslington, UK
Satu Huttunen, University of Oulu, Oulu, Finland
Juha Kämäri, Finnish Environment Institute, Helsinki, Finland
Galina Kashulina, Kola Science Centre, Apatity, Russia
Mikhail Kozlov, University of Turku, Turku, Finland
Annamari Markkola, University of Oulu, Oulu, Finland
Maximilian Posch, Netherlands Environmental Assessment Agency, Bilthoven, The Netherlands
Anna-Liisa Ruotsalainen, University of Oulu, Oulu, Finland
Reijo Salminen, Geological Survey of Finland, Espoo, Finland
Elena Zvereva, University of Turku, Turku, Finland
Chapter 6 · Effects on Freshwater Ecosystems:
6.1. Evidence from water quality monitoring
* Brit Lisa Skjelkvåle, Norwegian Institute for Water Research, Oslo, Norway
Julian Aherne
Martin Forsius
Natalia A. Gashkina, Water Problems Institute of Russian Academy of Sciences, Moscow, Russia
Jean-Paul Hettelingh
Dean Jeffries, Environment Canada, National Water Research Institute, Ontario, Canada
Jaakko Mannio, Finnish Environment Institute, Helsinki, Finland
Tatyana Moiseenko, Russian Academy of Sciences, Moscow, Russia
Maximilian Posch
John Stoddard, U.S. Environmental Protection Agency, Corvallis, USA
*
Lead
Authors
vii
Jussi Vuorenmaa, Finnish Environment Institute, Helsinki, Finland
Anders Wilander, Swedish University of Agricultural Sciences SLU, Uppsala, Sweden
6.2. Effects of acidification on Arctic biota
* John L. Stoddard
Tarja Bergman, Lapland Regional Environment Centre, Rovaniemi, Finland
Laura Forsström, University of Helsinki, Helsinki, Finland
Atte Korhola, University of Helsinki, Helsinki, Finland
Antti Lappalainen, Finnish Game and Fisheries Research Institute, Helsinki, Finland
Marjut Nyman
Ann Kristin Schartau, Norwegian Institute for Nature Research, Trondheim, Norway
Jouni Tammi, University of Helsinki, Helsinki, Finland
Valery Yakovlev, Institute of the North Industrial Ecology Problems, Kazan, Russia
6.3. Episodic acidification
* Kevin Bishop, Swedish University of Agricultural Sciences, Uppsala, Sweden
Martin Forsius
Øyvind Kaste, Norwegian Institute for Water Research, Grimstad, Norway
Hjalmar Laudon, Swedish University of Agricultural Sciences SLU, Umeå, Sweden
Tatyana Moiseenko
Brit Lisa Skjelkvåle
6.4. Evidence from paleolimnological studies
*Atte Korhola
Marjut Nyman
6.5. Interaction between acidification and other environmental issues
* Martin Forsius
Marjut Nyman
Chapter 7 · Air Pollution and Health Impacts in the Arctic:
* Jon Øyvind Odland, Institute of Community Medicine, University of Tromsø, Norway
Chapter 8 · Conclusions and Recommendations:
* Martin Forsius
John Derome
Lars R. Hole
Sirkku Manninen
Marjut Nyman
Patricia Quinn
Brit Lisa Skjelkvåle
John L. Stoddard
Contributors:
Dan Aamlid, Norwegian Forests Research Institute, Ås, Norway
Piotr Glowacki, Institute of Geophysics, Polish Academy of Sciences, Warszawa, Poland
Trygve Heshagen, Norwegian Institute for Nature Research, Trondheim, Norway
Johan C.I. Kuylenstierna, Stockholm Environment Institute at York, University of York, Heslington, UK
John Munthe, IVL Swedish Environmental Research Institute Ltd, Gothenburg, Sweden
Tadeusz Nied wied , Faculty of Earth Sciences, University of Silesia, Sosnowiec, Poland
Anne Owen, Stockholm Environment Institute, Stockholm, Sweden
Hans Tømmervik, Norwegian Institute for Nature Research, Tromsø, Norway
Jan Weckström, University of Helsinki, Helsinki, Finland
AMAP would like to acknowledge the provision of a range of data products. Data from Canada (Alert and Snare Rapids) were pro-
vided by the National Air Pollution Surveillance (NAPS) Network, a co-operative program of the federal, provincial, territorial, and
municipal government monitoring agencies. Data from Hornsund, Svalbard, were provided by the Polish Academy of Science. M.T.
Pavlova and T.A. Sokolova from the Voeikov Main Geophysical Observatory, St. Petersburg, Russia prepared data on the Russian Arc-
tic. Agriculture and Agri-food Canada provided (online) soil maps and associated data. AMAP would also like to acknowledge EMEP
and the International Cooperative Programmes (ICPs) of the Working Group on Effects under the UN ECE Convention on Long-range
Transboundary Air Pollution for data provision and collaboration.
Reviewers:
External reviewers:
Guy Fenech, United Kingdom (all chapters)
Peringe Grennfelt, IVL Swedish Environmental Research Institute Ltd, Sweden (all chapters)
National reviewers:
Olle Westling, IVL Swedish Environmental Research Institute Ltd, Sweden
Yngve Brodin, Swedish Environmental Protection Agency, Sweden
Jonas Rodhe, Swedish Environmental Protection Agency, Sweden
*
Lead
Authors
viii
ix
Executive Summary to the Arctic Pollution 2006 Ministerial Report
The first AMAP assessment Arctic Pollution Issues: A State
pollution within the Arctic. Epidemiological studies indi-
of the Arctic Environment Report documented direct evi-
cate that differences in health status of populations in areas
dence of acidification effects on the Kola Peninsula and in
of the Arctic with some of the highest levels of acidifying
limited areas of northern Norway and Finland, and around
air pollutants, the Norwegian and Russian border popula-
Norilsk in the Taymir region of Russia, mainly related to
tions, are more associated with socio-economic conditions
emissions from smelters in or close to these arctic areas.
than environmental pollution.
Acidification effects were also seen in some sensitive low-
deposition areas of the European Arctic receiving pollut-
ants from long-range transport. Data for areas of the North
Trends
American Arctic and eastern Siberia that, due to their geol-
ogy, are potentially vulnerable to acidification were gen-
Some air and precipitation monitoring stations have now
erally lacking. So although the assessment did not find
generated time series datasets that are long enough to show
evidence of acidification effects in these areas, it concluded
whether concentrations are increasing, decreasing, or stay-
that improved information on possible acidification effects
ing the same over time. Sulfate concentrations measured in
in these regions of the Arctic was desirable.
air at monitoring stations in the High Arctic (Alert, Canada;
The present assessment builds on information in the
and Ny-Alesund, Svalbard) and at several monitoring sta-
first assessment and fills several gaps in knowledge. In par-
tions in subarctic areas of Fennoscandia and northwestern
ticular it examines information on trends over the ten-year
Russia show decreasing trends since the 1990s. In contrast,
period since the first assessment was completed. It also
levels of nitrate aerosol are increasing during the haze sea-
addresses the need for more information on local sources
son at Alert (Canada), and possibly also at Barrow (Alaska)
of acidifying pollutants within the Arctic that were previ-
but longer data series are needed to confirm this trend.
ously unknown or insufficiently quantified; the need for
The increasing trends in nitrate are particularly apparent
more information on contaminant levels and trends in
in recent years indicating a decoupling between the trends
some areas; the need to integrate physical and biological
in sulfur and nitrogen. These observations are supported
models with information on environmental measurements
by modeling results.
of sources and pathways; and the need for more informa-
Although further improvement in the acidification
tion on the combined effects of climate change and con-
status of the terrestrial and freshwater ecosystems of the
taminant pathways on acidification in the Arctic and arctic
Arctic can be expected during the period until 2020, this is
haze, including improvements of models for assessments.
dependent on the implementation of existing international
This assessment also considers links to hemispheric pol-
agreements to reduce emissions of acidifying substances.
lution issues.
The Gothenburg Protocol to the UN ECE LRTAP Conven-
tion is the most important agreement in this connection.
However, model projections based on full implementation
Arctic Acidification
of the Gothenburg Protocol indicate that the decreasing
trends in deposition observed between 1990 and 2000 are
Arctic acidification is a subregional issue, and is only of
likely to level off. Measurement data indicate that down-
major concern in areas with both sensitive geology and
ward trends in concentrations may already be leveling off
levels of acid deposition elevated to a point that exceeds
at some sites.
the system's acid neutralizing capacity. Arctic haze is a
visible manifestation of long-range transported air pollu-
It is therefore recommended that:
tion. Arctic haze is largely composed of sulfate aerosol and
· All arctic countries are encouraged to ratify the UN
particulate organic matter, which builds up in the arctic
ECE LRTAP protocol to Abate Acidification, Eutroph-
atmosphere during wintertime and appears in springtime
ication, and Ground-level Ozone (the `Gothenburg
over large regions of the Arctic, both in North America and
Protocol') and to support its implementation.*
Eurasia as haze layers with reduced visibility.
· Arctic countries look into the need to strengthen the
Sulfur is the most important acidifying substance in the
provisions of the existing international agreements,
Arctic, with nitrogen of secondary importance. Significant
and consider the need for new instruments to reduce
anthropogenic sources of sulfur emissions, and to a lesser
emissions of acidifying substances.
extent nitrogen emissions, exist within the arctic region. In
addition, long-range transported air pollutants contribute
Significant reductions in emissions from the non-ferrous
to acidification and arctic haze in the Arctic. Emissions
metal smelters on the Kola Peninsula, and to a lesser ex-
from natural sources within the Arctic (volcanoes, marine
tent the Norilsk smelters, in the Russian Arctic have been
algae, and forest fires) are very difficult to quantify and
achieved over the past ten years. Chemical monitoring
almost impossible to project.
data show that lakes in the Euro-Arctic Barents region are
Studies to date have been unable to show any signifi-
showing clear signs of a regional-scale recovery from acidi-
cant health effects that are directly associated with emis-
fication. Lakes close to the sources on the Kola Peninsula
sions from the smelters that are the main sources of sulfur
are showing the clearest signs of recovery.
* The Protocol entered into force on 17 May 2005. As of July 2006, Denmark, Finland, Norway, Sweden and the United States have both signed
and ratified, accepted, or approved the Protocol, Canada has signed but not yet ratified the Protocol, and Iceland and the Russian Federation
have neither signed nor ratified the Protocol.
x
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
However, non-ferrous metal production remains the
noscandia has several background air monitoring stations
dominant source of emissions of acidifying gases to the
for acidification parameters, most areas of the Arctic have
atmosphere within the Arctic. Other significant anthro-
few, if any, background air monitoring stations.
pogenic sources of sulfur emissions within or close to the
Remote stations that are not affected by local or region-
Arctic include energy production plants and mining in-
al air pollutants are useful for studying trends in the levels
dustries. Sources of nitrogen emissions within the Arctic
of pollutants transported into the Arctic from long-range
include transportation, in particular shipping, and oil and
sources. Under AMAP, a network of arctic air monitoring
gas activities. Detailed information on all of these sources
stations has been established to assess trends in a range
is generally lacking.
of pollutants, including acidifying substances, persistent
organic pollutants, and metals such as mercury; however
It is therefore recommended that:
in recent years the overall coverage of this network has
· Information on emissions from arctic point sources in
been reduced such that coverage is limited, particularly in
Russia, in particular information on emissions from
Russia and the United States.
the non-ferrous metal smelters on the Kola Peninsula
and at Norilsk should continue to be made available.
It is therefore recommended that:
Information on emissions in other arctic areas should
· A critical review of the existing arctic air monitoring
be improved.
network be conducted to identify the optimal number
· The impacts of acidification from arctic shipping and
and location of long-term background monitoring
oil and gas activities, including future scenarios for
stations for air and precipitation chemistry.
emissions associated with these sources should be
· To the extent possible, this network should be inte-
assessed.
grated with other monitoring and research planning,
with the aim of developing a network of `multi-pur-
pose` background air monitoring stations in the Arc-
Links between Acidification, Arctic
tic.
Haze, and other Environmental Issues
The causes and the effects of acidifying air pollutants
Episodic events
and arctic haze are closely linked to other environmental
problems. It is not clear how climate change will influence
Short-term events of high atmospheric concentrations of
future acidification and arctic haze pollution in the Arctic.
sulfur dioxide are responsible for direct damage to veg-
The effects of haze aerosols on the arctic climate are com-
etation at varying distances from the smelters. At many
plicated by feedbacks between aerosols, clouds, radiation,
sites a large proportion of the annual acid deposition is
snow and ice cover, and vertical and horizontal transport
accumulated in just a few days.
processes. Whether the pollutant aerosols cause an overall
Similarly, pollutants deposited onto the snow pack ac-
warming or an overall cooling is not yet known.
cumulate throughout the polar winter and are released
The amount of haze precursors (haze-inducing sub-
rapidly into rivers and lakes with snowmelt in spring.
stances) reaching Alaska and the Canadian Arctic appears
These pulses of very acidic water can cause short periods
to have increased since the late 1990s. The frequency, se-
of very toxic conditions. Freshwater biota can be critically
verity, and duration of boreal forest fires appear to be in-
affected during acidic episodes and therefore assessments
creasing and the pollution plumes from these summer fires
need to address both average conditions and conditions
can extend over vast areas. In intense fire years, boreal
that may occur during episodic events.
forest fires may be the dominant source of black carbon
(soot) for the Arctic. The importance of Asian sources to
It is therefore recommended that:
acidification and arctic haze pollution in the Arctic is not
· Further studies, with high temporal resolution, be
yet clear.
conducted on the ecological impact of pulses or epi-
sodic events.
It is therefore recommended that:
· Future AMAP assessments view acidification and
arctic haze in the wider context of air pollution and
Effects on terrestrial and
climate change. The issues addressed in this more
freshwater ecosystems
integrated type of assessment should include hemi-
spheric transport of air pollutants, emissions from
In the European Arctic there are clear direct effects of sulfur
forest fires, particulate matter, and climate change
dioxide emissions on trees, dwarf shrubs, and epiphytic
effects.
lichens. The present deposition of acidifying compounds
resulting from long-range transport of anthropogenic emis-
sions at lower latitudes does not appear to be a threat to
Gaps in Knowledge Monitoring,
terrestrial ecosystems in most of the Arctic. In terms of their
Research, and Modeling
effects on plants, it is difficult to differentiate between the
effects of acidifying air pollutants and elevated heavy met-
Atmospheric monitoring
al levels in soils. Habitat destruction and possible changes
in food availability are strongly reducing biodiversity in
Acidification is not known to have serious impacts in the
the immediate vicinity of the smelters.
Arctic outside the Kola/Fennoscandia region and the
Taymir region in the vicinity of Norilsk. However, knowl-
It is therefore recommended that:
edge of acidification status in the Arctic is far from com-
·
Future studies be conducted on terrestrial ecosystems
plete, particularly in relation to future effects. While Fen-
to address the combined effects of acidifying sub-
xi
Executive Summary to the Arctic Pollution 2006 Ministerial Report
stances and heavy metals and other relevant factors
· Studies be conducted to identify and provide esti-
in an integrated manner.
mates of sources of black carbon to the Arctic.
· Data sets gathered during aircraft and ground-based
Available terrestrial and freshwater monitoring data pro-
surveys, in particular, long-term data sets, be inte-
vide irregular and incomplete coverage of the Arctic, even
grated for use in three-dimensional arctic climate
in acid-sensitive regions. Similarly, assessments of bio-
models designed to evaluate climate forcing by arctic
logical effects of acidification in arctic surface waters are
haze.
largely based on sparse and isolated data.
It is therefore recommended that:
Cooperation on monitoring
· Coordinated monitoring and research be carried out
to provide more chemical and biological data on ef-
Close cooperation between AMAP and other international
fects and trends in terrestrial and freshwater ecosys-
organizations involved with monitoring and modeling
tems in the most impacted areas of the Arctic.
deposition and effects of acidifying pollutants within the
European Arctic, such as programs under the UN ECE
LRTAP Convention, have proven mutually beneficial. The
Modeling
new EANET (Acid Deposition Monitoring Network in
East Asia) initiative represents an opportunity to develop
Modeling is one of the most important tools available for
similar cooperation in relation to monitoring in the Far
gaining insight into the possible pollution status of the
East of Asia.
extensive areas of the Arctic where the observational net-
works are absent or poorly developed. Models also allow
It is therefore recommended that:
investigation of scenarios for future trends, and for link-
· AMAP continues to develop its cooperation with
ages between contaminant pathways and, for example,
relevant international organizations, in particular to
climate change.
obtain more precise data on emissions from southeast
Asia and to investigate the possible impact of these
It is therefore recommended that:
emissions on the Arctic.
· Existing air transport and deposition models be im-
· Resources be made available to ensure that relevant
proved and further validated using measurements of
existing and future national data on acidification pa-
sulfur compounds, nitrogen compounds, and black
rameters, in particular from arctic monitoring sta-
carbon in the Arctic, including measurements con-
tions, are reported to the AMAP database at NILU
ducted during field campaigns.
according to agreed procedures.
1
Chapter 1
Introduction
Martin Forsius
As a consequence of the intense scientific and environmen-
Within the Arctic itself there are a few but very sig-
tal policy work on acidification, the atmospheric emissions
nificant sources of air pollutants. Production of copper,
of acidifying pollutants, the atmospheric processes, and
nickel, and other non-ferrous metals from sulfur-bearing
the environmental impacts of these pollutants on different
ores creates the largest emissions of acidifying compounds
ecosystems are now well understood. The acidifying com-
(mainly sulfur) and heavy metals. Most of these smelter
pounds sulfur dioxide (SO2), nitrogen oxides (NOX), and
emissions come from the Nikel, Zapolyarnyy, and Monche-
ammonia (NH3) have different sources. Sulfur dioxide
gorsk complexes on the Kola Peninsula and from Norilsk
emissions are mainly associated with point sources such
on the Taymir Peninsula in northwestern Siberia. There are
as power plants, smelters, pulp and paper mills, and oil
also emissions from several large cities, notably Murmansk
and gas processing. As well as these point sources, NOX
with around 400 000 inhabitants. Consequently, the regions
emissions are also derived from diffuse sources such as
surrounding the large smelter complexes in northern Rus-
vehicles. Ammonia emissions are almost entirely derived
sia, as well as the northeastern areas in the neighboring
from agricultural sources and so are more difficult to
countries of Norway and Finland, are the areas where most
quantify. A full understanding of the acidification prob-
acidification and other air pollution impact studies have
lem requires information on the emissions and processes
been undertaken over the last few decades.
of neutralizing compounds such as base cations. Neutral-
At the First Ministerial Conference on the Arctic Envi-
izing compounds are derived from anthropogenic and
ronmental Protection Strategy (AEPS), held in Rovaniemi,
natural sources.
Finland in June 1991, Ministers of the Arctic States estab-
Rapidly increasing scientific evidence on acidifica-
lished the Arctic Monitoring and Assessment Programme
tion during the 1970s and 1980s was the starting point
(AMAP) to `monitor the levels of anthropogenic pollutants
for international negotiations on controlling emissions of
in relevant compartments of the Arctic environment'. Min-
compounds that undergo long-range transport. The Con-
isters further identified persistent organic pollutants, heavy
vention on Long-Range Transboundary Air Pollution (UN
metals, and radioactivity as the key pollutants that should
ECE CLRTAP) and the air pollution work and directives
be a priority during the first phase of AMAP (19911997).
of the European Commission have been key international
The Ministerial Conferences in Nuuk, Greenland (1993)
activities in this respect. Large international emissions
and Inuvik, Canada (1996) extended this list to include:
databases, pollutant transport models, environmental
acidification and arctic haze, and petroleum hydrocarbon
impacts monitoring and assessment programs, and inte-
pollution, in a subregional context; and environmental con-
grated assessment models have been created within these
sequences of, and biological effects due to global climate
frameworks and have served as technical infrastructures
change and stratospheric ozone layer depletion, relevant
for negotiations (e.g., Sliggers and Kakebeeke, 2004; UN-
to the Arctic.
ECE, 2004b). Work on the acidification issue has been one
An assessment of acidification and arctic haze was car-
of the great environmental `success stories' and SO2 emis-
ried out within the AMAP framework and reported in the
sions were reduced by around 67% in Europe between 1980
extensive and fully-referenced AMAP Assessment Report:
and 2000; with many countries having reductions of almost
Arctic Pollution Issues (AMAP, 1998). This concluded that
90% (EMEP, 2004).
there was direct evidence of acidification effects on the
Since the 1970s, the focus on air pollution has wid-
Kola Peninsula and in a limited area of the Norwegian
ened considerably and is now moving toward issues such
part of eastern Finnmark. Direct damage to forests, fish,
as eutrophication, small particulates and health effects,
and invertebrates was documented near the Kola smelters.
ground-level ozone, heavy metals, and the interactions
There were no indications of any acidification impacts in
between these issues and with climate change. Rapid in-
the North American Arctic although large areas were con-
dustrial growth and increasing energy use in several world
sidered potentially vulnerable to acidification.
regions (including heavily populated countries such as
This assessment updates information in the previous
China and India) have added a new dimension. There is
assessment. The main aim has been to include new knowl-
also increasing evidence of the importance of hemispheric
edge of processes, sources, and pathways, to fill gaps in
transport of pollutants, such as ozone, mercury, and per-
knowledge on contaminant levels and trends and to correct
sistent organic pollutants.
possible errors in the previous assessment. Interactions
In the 1970s it was recognized that even remote parts
with other issues (mainly climate change and heavy met-
of the Arctic are influenced by air pollution, mainly due to
als) and connections to hemispheric pollution issues are
anthropogenic activities at lower latitudes. Arctic haze was
also considered.
first reported in the 1950s by pilots flying in the Canadian
The structure of this report follows the pollution path-
and Alaskan Arctic, but it was not until the mid-1970s that
ways from emission sources (chapter 2), through trans-
its anthropogenic origin was established (e.g., Rhan et al.,
port and deposition processes (chapter 3), and impacts and
1977). Arctic haze is a varying mixture of sulfate, particu-
processes of aerosols (chapter 4), to the present status and
late organic matter, nitrogen compounds, dust and black
trends in the chemical and biological responses to acidify-
carbon, as well as trace elements such as heavy metals and
ing pollutants in terrestrial and aquatic ecosystems (chap-
other contaminants. Arctic haze aerosol particles thus pro-
ters 5 and 6). Chapter 7 reviews the human health impacts
vide a transport pathway for contaminants to the Arctic.
of acidifying air pollutants in the Arctic, while chapter 8
The impact of arctic haze on climate forcing is receiving
summarizes the main conclusions of this assessment and
increasing attention.
presents recommendations for further work.
2
Chapter 2
Sources of Acidifying Pollutants and Arctic Haze Precursors
Lars R. Hole, Jesper Christensen, Martin Forsius, Marjut Nyman, Andreas Stohl, and Simon Wilson
2.1. Sources within the Arctic
ade, these smelters remain the largest sources within the
Arctic and so this chapter also focuses on these sources.
Within the Arctic, the major anthropogenic emissions of
Non-ferrous metal production is one of the greatest
nitrogen oxides (NOX) and sulfur dioxide (SO2) are asso-
sources of environmental pollution in Russia, second only
ciated with sources located in a limited number of areas
to energy production in accounting for `harmful emissions'
where industrial enterprises and/or population centers
(SO2, dust, NOX, etc.) to the atmosphere. Sulfur dioxide,
occur. With the exception of oil and gas related activities,
originating from the roasting and smelting of sulfur-con-
these emission sources are almost exclusively located in
taining minerals, makes up the bulk (ca. 80%) of the emis-
the northern territories of the Russian Federation. Accord-
sions from non-ferrous metal processing plants. Around
ing to the Russian Federal Statistical Committee (Goskom-
75% of the SO2 emissions are associated with the produc-
stat), emissions of SO2 in arctic regions of Russia accounted
tion of nickel in the nickel-cobalt sector, with most of the
for 33% of the total SO2 emissions from all territories of the
rest from the copper production sector. Emissions from
Russian Federation.
aluminum plants are mainly `dust' (particulates) and car-
Despite generally very low population densities, large
bon dioxide (CO2).
towns and cities do exist within the (sub-)Arctic, notably
Based on information (19982001) from reports of com-
Murmansk with its population of 400 000. There are also
panies involved in non-ferrous metal production in Russia,
natural sources located within the Arctic; these include
ten companies are responsible for around 85% of the total
volcanoes that emit SO2 in the volcanically active areas of,
`harmful emissions' to the atmosphere. The non-ferrous
for example, Kamchatka, Alaska, and Iceland. However
metal smelters at Norilsk, Zapolyarnyy, and Monchegorsk
volcanic emissions are not addressed in this assessment.
(all operated by MMC Norilsk Nickel) are all within or
Natural biogenic emissions associated with forest fires are
close to the Arctic and together account for 68% of the
another source of natural emissions and these are increas-
total `harmful emissions' from non-ferrous metal produc-
ingly prevalent in arctic areas as a consequence of climate
tion in Russia. The Norilsk Nickel consortium is the larg-
change (ACIA, 2004). These emissions are described fur-
est producer in the Russian non-ferrous metal sector and
ther in section 2.1.5.
one of the world's largest producers of nickel, palladium,
Energy production and transport in and around arctic
platinum, cobalt, and copper.
urban population centers are responsible for reduced air
Time series of SO2 emissions from the non-ferrous met-
quality at the local scale, including pollution by NOX, SO2,
al smelter at Nikel (on the Kola Peninsula) between 1980
and fine particulates, which can have negative health im-
and 2002 (Figure 2.1) illustrate the substantial reductions
plications (see chapter 7). However, industrial sources are
that have occurred over this period. Decreased emissions
responsible for the major emissions of concern within the
during the early 1990s are largely due to economic de-
Arctic. Although few in number, some of these industrial
cline following the break-up of the former Soviet Union,
sources, in particular those associated with non-ferrous
although by 1995 emissions at Nikel had returned to their
metal smelting operations, are significant, both at a re-
1990 levels (ca. 180 kt SO2). The main emissions reductions
gional scale and in terms of global ranking of individual
have been achieved since 1996.
sources.
In recent years, the MMC Norilsk Nickel plants have al-
The first AMAP assessment (AMAP, 1998) concluded
so considerably reduced their emissions of SO2. Abatement
that NOX makes a negligible contribution to the acidifi-
measures include the (by-)production of sulfur (ca. 260 kt
cation effects observed within the Arctic, and that arctic
in 2000) and sulfuric acid (ca. 30 kt in 2000) from captured
sources are insignificant compared with the amount of
SO2, although this is not a profitable activity due to high
NOX carried into the Arctic via long-range transport. The
transportation costs and distance from markets. Emissions
present assessment confirms this finding, but due to the
from the Pechenganickel (Zapolyarnyy/Nikel) and Seve-
differences achieved in reductions of SO2 relative to NOX,
ronickel (Monchegorsk) combines have been reduced from
and the potential for increased development of the Arc-
around 255 kt in 1991 to around 150 kt in 2000, and from
tic that may enhance NOX emissions within the region,
around 180 kt in 1992 to around 45 kt in 2000, respectively,
this assessment addresses NOX in more detail than the
as a result of introducing emission control technologies and
first AMAP assessment. There is also increasing evidence
the closure of the Severonickel smelting plant.
documenting intercontinental transport of air pollutants,
In 1999, MMC Norilsk Nickel announced extensive re-
including NOX, SO2, fine particles, and black carbon.
construction of its production facilities, including the in-
troduction of technologies to reduce emissions of harmful
substances. Measures include modernizing the processing
2.1.1. Stationary sources: industry and energy
of copper-nickel ores using new flotation reagents in order
to reduce sulfur minerals prior to roasting and smelting,
The first AMAP assessment clearly identified the non-fer-
and reconstruction of facilities to produce and utilize tech-
rous metal smelters at Nikel, Monchegorsk, and Zapol-
nical-grade sulfur on a profitable basis.
yarnyy on the Kola Peninsula and at Norilsk in northern
Part of the reconstruction of facilities on the Kola Penin-
Siberia, as the largest (anthropogenic) sources of acidifying
sula is being funded through environmental cooperation
air pollutants within the Arctic (AMAP, 1998). Although
agreements between Russia, Norway, and Finland, and
emissions reductions have been achieved over the last dec-
emission reductions from these sources are being imple-
3
SO concentrations in air at Svanvik, g/m3
SO emissions from Nikel, kt
2
2
35
350
Svanvik
Svanvik
30
300
Nikel
Nikel
R U S S I A
25
250
N O R W A Y
20
200
F I N L A N D
S W E D E N
15
150
10
100
5
50
0
0
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
Figure 2.1. Annual SO2 concentration in air at ground level at the Svanvik monitoring station and SO2
emissions from the non-ferrous metal smelters at Nikel (after Hagen et al., 2005).
mented in accordance with agreements under the LRTAP
Archangelsk Oblast, Nenets Autonomous Okrug and Komi
Convention to reduce SO2 emissions relative to 1980 levels
Republic were 240 (ca. 85% from the Kola smelters), 77,
by 40% by 2005 and 50% by 2010.
89, 3.8, and 65 kt, respectively in 2002. Additional indus-
In addition to the smelter complexes, other industries
try related sources are the pulp and paper industry in the
also emit SO2 to the atmosphere including the energy sec-
Republic of Karelia and Archangelsk Oblast and the oil
tor. An inventory of the 130 largest coal-fired power sta-
and gas industry in the Nenets Autonomos Okrug and
tions in the Russian Federation (VTI/IVL, 2004) lists nine
Komi Republic.
plants within the Arctic (see Figure 2.2). These had com-
With the possible exception of oil and gas related sourc-
bined SO2 emissions of around 80 kt in 2002 (85% from the
es (see section 2.1.3), there are few significant point sources
plants at Severodvinsk, Vorkuta, and Apatity), which is less
of SO2 within the Arctic outside the territories of the Rus-
than 5% of the combined SO2 emissions from the Kola and
sian Federation. The few sources that do exist include min-
Norilsk smelters in the same period.
ing activities and small-scale power generation or waste
Total SO2 emissions from the five regions of north-
incineration plants located in population centers.
western Russia Murmansk Oblast, Republic of Karelia;
Stationary combustion
Iron and steel production
Non-ferrous metal production
AMAP boundary
SO emission, kt
2
250 Nikel
SO emission, kt
2
Norilsk
2000
0
250
1000
Zapolyarnyy
0
1992
2002
Apatity
0
250 Monchegorsk
Severodvinsk
0
Inta
1992
2002
Vorkuta
Regional SO emission in 2002, kt
2
0
5
40
80
100
250
50 kt
Figure 2.2. Sulfur dioxide emissions from metallurgical industry sources and major coal-fired power plants in Russia (there are no significant point
sources north of 60º N in other Arctic countries; pale coloured symbols are sources south of 60º N). Also shown are total emissions of SO2 in 2002
from the five regions of northwestern Russia. Subplots show trends in emissions from Russian non-ferrous metal smelters. Orange columns on
the subplots, and on the map (for the four power plants with the largest SO2 emissions), represent emissions in 2002 (note: the scale on the Norilsk
subplot is different from all other data shown). Sources: smelter emissions: Ministry of Natural Resources official statistics; coal-fired power plant
emissions: VTI/IVL, 2004.
4
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
2.1.2. Local air pollution in Russian cities
2.1.3. Oil and gas activities
Urban air monitoring data from Russian cities can help
The Arctic is estimated to contain at least 25% of the world's
assess the influence of the main pollution sources (Figure
undiscovered petroleum resources (Ahlbrandt, 2002). It
2.3). However, it is important to remember that such data
has been suggested that by 2010, annual volumes of 150
represent average and maximum concentrations in ambi-
million tonnes of oil may be shipped by sea as a result
ent urban air, and that as well as industrial emissions these
of the development of oil production and transportation
concentrations also reflect other urban activities including
infrastructure in the Barents region alone (Frantzen and
vehicle emissions.
Bambulyak, 2003).
The highest NOX concentrations among cities in the
As energy consumption and air pollution are directly
Russian Arctic region were observed at Norilsk and Mur-
linked, any increase in oil and gas related activities is likely
mansk up to 1998, but since then high concentrations have
to result in increased pollutant emissions. According to
also occurred in Salekhard and Nikel. Concentrations of
WWF (2005) oil and gas development will pose a major
SO2 in Russian cities have been consistently highest in
threat to the Barents Sea region.
Nikel and Norilsk.
From the exploration phase to final closure of the
Between 1990 and 2003 there was a clear increase in
production field, oil and gas production involves emis-
NOX concentrations in Russian cities. This is particularly
sions and discharges to air and water. Emissions to air
obvious at Nikel, Norilsk, and Salekhard after 1997. Av-
include exhaust gases containing CO2, NOX, sulfur oxides
erage and maximum NOX concentrations are now two
(SOX), methane (CH4) and non-methane volatile organic
to three times higher in Nikel than they were between
compounds (nmVOC) from various types of combustion
1990 and 1997, and one and a half to two times higher
equipment and sources including gas turbines, engines
in Monchegorsk, Norilsk, and Salekhard. Maximum NOX
and boilers; gas flaring; and oil and gas burning in connec-
concentrations in Norilsk increased until 1999, but have
tion with well testing and well maintenance work. Other
decreased significantly since then. These increases also
sources of hydrocarbon gases (CH4 and nmVOC) include
reflect the big increase in the number of private vehicles
gas ventilation, minor leaks, and diffuse emissions and
in the Russian Arctic well as in the rest of Russia over the
boiling down of hydrocarbon gases (largely nmVOC) from
last ten years.
the storage and loading of crude oil. Power generation,
There was a general decrease in SO2 concentrations be-
using natural gas and diesel oil as fuel, is the predominant
tween 1990 and 2003 in Nikel, Monchegorsk, and Norilsk.
reason for CO2 and NOX emissions, followed by gas flar-
However, interannual variability is very high (in Norilsk
ing. The main cause of SOX emissions is the combustion of
annual average concentrations can vary by a factor of two).
sulfur containing hydrocarbons.
This variability makes it impossible to detect clear trends
The Arctic Offshore Oil and Gas Guidelines prepared
in SO2 concentration.
by the Arctic Council Working Group on Protection of
the Arctic Marine Environment (PAME, 2002) provide an
overview of potential impacts of offshore development
in the Arctic and how these impacts should be taken into
NO concentration in air, g/m3
X
consideration in offshore oil and gas activities. The guide-
60
lines cover air emissions, which mainly originate from
50
the combustion of fuels for power generation, and direct
40
emissions from the production, treatment, storage, and
transportation of oil and gas. According to PAME, "the
30
activities will entail considerable inputs of gases into the
20
atmosphere from power generation, flaring, well testing,
10
leakage of volatile petroleum components, supply activi-
ties and shuttle transportation. These air emissions may
0
have effects on the climate and they may cause acidification
on nearby land and contribute to emissions of any number
of hazardous substances." Methane, which is a powerful
SO concentration in air, g/m3
2
Norilsk
greenhouse gas, is also released to the atmosphere by gas
100
250
drilling, from leaky pipelines, and by venting and flaring
80
200
activities on oil and gas rigs (AMAP, 1998).
The ongoing AMAP assessment of petroleum hydro-
60
150
carbons in the Arctic will address the environmental im-
40
100
pact of oil and gas developments in the Arctic, including
the associated effects on human health and social and eco-
20
50
nomic consequences. It will address emissions from oil and
gas activities, however comprehensive coverage of all air
0
0
emissions from the many activities related to oil and gas
1990
1992
1994
1996
1998
2000
2002
production may warrant a separate assessment at some
point in the future.
Salekhard
At present, the main areas of oil and gas exploitation in
Norilsk
Nikel
the Arctic include the North Slope of Alaska, the offshore
Monchegorsk
areas in the Norwegian and Barents Sea, and the Yamalo-
Murmansk
Nenets and Komi areas of Russia. Extensive petroleum
resources also exist in the Canadian Arctic, which may be
Figure 2.3. Trends in NOx and SO2 concentrations in air in various cities
in the Russian Arctic.
developed if the economic situation is favorable. Oil and
Chapter 2 · Sources of Acidifying Pollutants and Arctic Haze Precursors
5
gas exploration activities are being actively pursued in
types is required in the vicinity of planned/prospective
other parts of the Arctic, including the offshore areas west
land-fast constructions or facilities related to the oil and gas
of Greenland, around the Faroe Islands, and in the Kara,
industry. The DNV report concluded that oil-field specific
Chukchi, and East Siberian Seas.
investigations should be made of the effects on the local
Although detailed information on emissions associated
environment of increased emissions to air. Also, that it is
with oil and gas activities in the Arctic is lacking, available
difficult to estimate the effect of increased air pollution
information can be used as an example. The petroleum
from oil and gas activities on the arctic haze phenomenon,
sector makes a considerable contribution to the national
but that this may be significant since the haze is known
total emissions for Norway. For example, a large part of
to consist of anthropogenically produced impurities (see
Norway's VOC emissions are from petroleum-related
chapter 4).
emissions in the North Sea. Similarly, an estimated 20% of
While the impact of oil and gas activities on climate
Norwegian NOX emissions in 2000 were derived from oil
in the arctic areas is a less well-covered issue, a warming
and gas resource development (this figure excludes related
climate has several consequences for oil and gas produc-
shipping activities). For SOX, total emissions by the Norwe-
tion in the Arctic. Climate change is projected to lead to less
gian oil and gas industry in 2003 were around 600 tonnes;
sea ice. The use of oil and gas resources in remote arctic
the low emissions figures reflect the low sulfur content of
locations will increase as the ice-free season increases and
the fuel gas used in gas turbines and engines. As a result,
ice cover decreases. Increased transportation activities will
diesel combustion was the main source of SOX emissions
increase the risk of oil spills. Emissions to the atmosphere
from the oil industry. Annual Norwegian SO2 emissions
are also likely to increase. Transport in general uses more
were 23 000 tonnes in 2003; thus, the Norwegian oil indus-
energy than industrial, agricultural, commercial, and in-
try accounts for approximately 2.5% of total emissions in
stitutional sectors combined. In addition to enhancing oil
Norway (OLF, 2003).
and gas activities, climate warming will increase risks in
A study into the possible impact of hydrocarbon ac-
winter operations such as the construction of ice roads
tivities in the Lofoten-Barents Sea area involved develop-
(Thurston, 2003).
ment scenarios with three levels of activity. Air emissions
In many cases, international agreements commit sig-
in each scenario included emissions from offshore and in-
natories to limiting their air emissions to certain levels.
shore facilities and from transportation. The areas affected
Greenhouse gas emissions may be increased by oil and gas
by emissions from hydrocarbon production in the Barents
activities, Norway can meet its commitments to reducing
Sea-Lofoten area are Finnmark, Troms, and northernmost
greenhouse gas emissions through the Kyoto mechanisms
Nordland (Guerreiro et al., 2003). In each scenario, the total
using national reductions or international emissions trad-
increase in air emissions relative to national emissions in
ing. Other gases (NOX, SO2, and VOC) causing regional
2001 is 3 to 8% for CO2 and 0.7 to 2.8% for NOX. Transport
environmental problems, such as acid rain, eutrophication,
contributes 50% for NOX (DNV, 2003). However, the pro-
and ground-level ozone, are regulated by the LRTAP Con-
jected increase in deposition (e.g., for nitrogen) relative to
vention. Of the countries with prospective oil and gas fields
that at present seems higher than the relative increase in
on the continental shelf, the Gothenburg Protocol has been
emissions; up to 25% at sea and 5 to 15% on land owing
signed and ratified by Norway and the USA and signed by
to the currently very low deposition values (around 400
Canada. Russia has not signed the protocol.
mg/m2 per year).
The extensive impact assessment of year-round petro-
leum activities in the Lofoten-Barents Sea by Norwegian
2.1.4. Shipping activities
authorities indicates that there will be emissions to air,
which although mostly insignificant in terms of impact,
Emissions from shipping are an important component of
may in some cases exceed critical limits.
total global emissions of SO2 and NOX and are of concern in
According to DNV (2003) the normal level of activity
the Arctic because they can influence nearby landmasses.
in the oil industry does not cause serious effects in the
According to the EDGAR database (Emission Database for
marine environment. However, in some cases, there may
Global Atmospheric Research), international marine trans-
be significant CO2 emissions to air. For example, increases
port contributed around 7300 kt of SO2 and around 9600
of 15 to 20% in total CO2 emissions may occur due to in-
kt of nitrogen dioxide (NO2) to global emissions in 2000;
creased energy use by the discharge of produced waters.
corresponding to around 5 and 7.5% of total global emis-
Transportation would cause most of the nitrogen emis-
sions, respectively. For both SO2 and NOX, the eight arctic
sions. The contribution from the oil industry to acidifica-
countries are responsible for around 25% of the estimated
tion is considered to be low and emissions are thought not
global emissions due to international marine transport.
to exceed critical limits (DNV, 2003). None of the scenarios
Within the Arctic, emissions to the air from shipping
led to exceedance of critical loads for terrestrial ecosystems.
in 1998 were estimated at 30 kt of SO2 and 105 kt of NOX
However, critical loads for northern ecosystems are not
(Table 2.1). These estimates are based on a statistical emis-
well established and more careful mapping of ecosystem
sions modeling approach using available data on ship-
Table 2.1. Emissions to air in the Arctic and fuel consumption by arctic shipping in 1998 (kt) (Norwegian Maritime Directorate, 2000).
Particulate
NOX
CO
nmVOC
SO2
CO2
matter
CH4
N2O
Fuel
Wet cargo
6.4
0.6
0.2
3.2
276
0.44
0.03
0.01
87
Dry cargo
40.1
4.9
1.6
10.6
2101
1.38
0.20
0.05
663
Fishing vessels
47.6
5.6
1.8
13.9
2412
1.83
0.23
0.06
761
Icebreakers/Tugs 11.3
1.4
0.5
2.4
610
0.30
0.06
0.02
192
Total
105.4
12.5
4.1
30.0
5399
3.95
0.51
0.14
1703
nmVOC: non-methane volatile organic compounds.
6
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Ports-of-call
Navigation route
Northern Sea Route
Chukchi Sea
Northwest Passage
Oil/gas regions
Beaufort Sea
Arctic fishing grounds
Arctic Islands
Western
Siberia
Barents Sea
Timan-
Pechora
Norwegian
Sea
Figure 2.4. Shipping routes, oil and gas regions, and fishing grounds in the Arctic.
ping volume, estimated fuel consumption, and relevant
Union. Recently, however, traffic volumes have increased
emission factors (DNV, 1999). In addition to commercial
and are projected to continue to increase over the next 10
shipping (cargo transport and cruise shipping), these es-
to 15 years, with part of this increase associated with the
timates also include emissions from fishing vessels. The
transport of oil from northern Russia to markets in Europe
DNV study noted inadequacies in the available shipping
and Asia (Brigham and Ellis, 2004). Transport of petroleum
statistics, however. For example, the estimates do not fully
products from northern Russia, together with increasing
address emissions from shipping on inland waterways,
cruise traffic is also resulting in increased shipping activ-
which is a very significant shipping activity in the Arctic
ity in the Barents Sea and the Norwegian Sea. There are
and in northern Russia in particular. Also, the estimates do
also extensive maritime activities in Canadian arctic wa-
not include emissions from military shipping. Although
ters (i.e., the Beaufort Sea, Northwest Passage, and Hud-
military activity in the Arctic (such as operations by the
son Bay), including deliveries of general cargo and fuel
Russian Northern Fleet) have been reduced considerably
through the Mackenzie River system, and the supply of
since the end of the Cold War, emissions associated with
cargo and petroleum products to eastern arctic communi-
naval vessels and other military installations may be a
ties. Projected reductions in sea-ice thickness and extent
significant additional component.
associated with climate change, and a resulting increase
The most extensive shipping activities in the Arctic
in the length of the season during which the NSR and
take place in Russia, with the Northern Sea Route (NSR)
Northwest Passage are open for navigation have raised
carrying the largest volume of traffic of any arctic seaway.
the possibility that shipping activities in arctic waters may
The NSR connects the Barents Sea and the Bering Strait
increase substantially in future (ACIA, 2004). According
and serves ports at the mouths of the major Russian riv-
to a recent study (Lee Behrens, 2000) the most significant
ers (Figure 2.4) providing a connection with the extensive
emissions from merchant shipping in 2000 were associated
ship and barge transport along the major inland water-
with the transport of dry cargo.
ways of northern Russia. The NSR was officially opened
The other major source of emissions was fishing ves-
for international transit trading in 1991; although there
sels, reflecting the large size of the fleets, high fuel con-
has been little commercial utilization by non-Russian ves-
sumption by (in particular larger) fishing vessels, and
sels to date. The NSR is restricted by sea ice and open
the long periods spent at sea. Because fishing is a largely
for as little as two and a half months of the year at some
seasonal activity, at least for the smaller fishing vessels
points. Annual traffic along the NSR peaked at around
(<30 m) that make up 80% of the fishing fleets, emissions
6600 kt in 1987 but by 2000 had decreased to around 1500
occur primarily during the summer. Large-scale commer-
to 2000 kt, reflecting declining economic activity in this
cial fishing activities are concentrated in the Barents and
part of Russia following the break-up of the former Soviet
Bering Seas.
7
Chapter 2 · Sources of Acidifying Pollutants and Arctic Haze Precursors
Although estimates of emissions from shipping in the
2.2. Sources outside the Arctic and
Arctic are subject to considerable uncertainty due to the
atmospheric transport to the Arctic
current lack of detailed information on shipping statis-
tics, several observations are possible. Taken together,
emissions from shipping in the Arctic are currently small
Because of its remoteness, the arctic troposphere was long
compared with global emissions from marine traffic (<1%
believed to be extremely clean. Arctic haze was described
for SO2 and approximately 1% for NOX). Sulfur dioxide
for the first time in the 1950s and in the 1970s it became
emissions from arctic shipping are also minor compared to
obvious that anthropogenic sources outside the Arctic
emissions from other sources within the Arctic (e.g., <5%
strongly contributed to pollution of the arctic atmosphere
of the emissions from Russian smelters), however this is
(see chapter 4). The haze phenomenon, accompanied by
not the case for NOX with shipping emissions more than
high levels of gaseous air pollutants (e.g., hydrocarbons,
double the reported emissions from Russian non-ferrous
Solberg et al., 1996), has been observed regularly since then,
metal production.
especially in the lower troposphere, and is due to the spe-
The environmental impact of NOX and SO2 emissions
cial meteorological situation in the Arctic in winter and
from shipping in the Arctic will depend upon their con-
early spring (Shaw, 1995).
tribution relative to existing background concentrations in
The extreme dryness of the arctic troposphere mini-
the areas concerned. The seasonal nature of some inputs
mizes wet deposition, thus leading to a very long lifetime
must also be considered. Increased shipping traffic density
for aerosols in the Arctic in winter. Surfaces of constant
may contribute significantly to the load of acidifying gas-
potential temperature form closed domes over the Arc-
ses in some areas (e.g., close to ports), however available
tic, with minimum values in the arctic boundary layer
evidence suggests that shipping is not currently a major
(Klonecki et al., 2003). This isolates the lower troposphere
contributor to overall levels of acidifying pollution in the
from the rest of the atmosphere by a transport barrier,
Arctic. An increase in arctic shipping traffic is expected
the so-called `Arctic Front'. Meteorologists realized that in
during the next 10 to 15 years, as a result of increasing oil
order to facilitate isentropic transport, a pollution source
transport (particularly in the Barents Sea) and a longer
region must have the same low potential temperatures as
navigation season due to reduced ice cover. However,
the arctic haze layers (Carlson, 1981; Iversen, 1984; Barrie,
emissions associated with this increased shipping activ-
1986). This rules out most of the world's pollution source
ity will probably be partly compensated by improved
regions because they are too warm, and leaves northern
technology, and possibly also by reduced levels of fishing
Eurasia as the main source region for the arctic haze (Rahn,
activity. Thus, dramatic changes in sulfur and nitrogen
1981; Barrie, 1986). There, the Arctic Front can be located
emissions from shipping are not expected in this period.
as far south as 40º N on average in January (see Figure
In the longer term, a more significant increase in shipping,
4.1). Furthermore, northern Eurasia is on a preferred path-
for example due to ice-free conditions in the NSR, could
way into the polar dome that involves diabatic cooling of
lead to more substantial emissions of acidifying pollutants
air traveling over snow-covered land. This transport is
in the region.
highly episodic and often related to large-scale blocking
events (Raatz and Shaw, 1984; Iversen and Joranger, 1985).
In contrast, air masses leaving North America's densely
2.1.5. Natural sources within
populated east coast are heated diabatically (Klonecki et
the Arctic: wildfires
al., 2003) because of the frequent warm conveyor belts over
the downwind North Atlantic Ocean (Stohl, 2001; Stohl et
Wildfires are a large episodic source of black carbon and
al., 2002; Eckhardt et al., 2003). Southeast Asia is located at
other pollutants (Lavoué et al., 2000). It was realized only
even higher potential temperatures than North America
recently that boreal wildfire emissions affect the atmos-
and so was also rejected as a source of arctic haze, although
phere at a hemispheric scale (Wotawa et al., 2001). Pol-
an Asian desert origin was suggested for the elevated haze
lution plumes originating from boreal fires have been
layers (Rahn et al., 1977) that are also quite frequent (Lei-
observed over downwind continents (Forster et al., 2001)
terer et al., 1997).
and can circle the entire northern hemisphere (Damoah et
Pollution transport to the Arctic varies considerably
al., 2004). An aircraft campaign frequently found aerosol
(see section 3.7.3 on the effects of natural climate varia-
plumes from Alaskan and maybe also Siberian forest fires
tions on long-range transport to the Arctic). During the
over the Alaskan Arctic (Shipham et al., 1992) and a PhD
`positive' phase of the North Atlantic Oscillation (NAO),
thesis suggests a link between boreal forest fires and black
transport from all three northern hemisphere continents
carbon observations in Greenland and at other arctic sites
(Europe, North America, and Asia in order of significance)
(Lavoué, 2000). The increase in areas burned over recent
into the Arctic is enhanced, resulting in higher arctic pol-
decades (Lavoué et al., 2000; Kasischke et al., 2005), which
lution levels (Eckhardt et al., 2003; Duncan and Bey, 2004).
is probably due to a warming in the boreal region, is a
Given the long-term variability in the NAO, its status must
matter of concern. There are speculations that black carbon
be considered when studying arctic pollution trends (Mac-
deposits from boreal forest fires could enhance the melting
donald et al., 2005).
of arctic glaciers and sea ice (Kim et al., 2005). Boreal fires
Recently, new issues have attracted scientific interest.
are a summer-time phenomenon that occur when removal
Anthropogenic emissions in southern and eastern Asia
mechanisms (wet and dry deposition) are relatively effi-
have grown rapidly over the past few decades, especially
cient and the arctic troposphere is generally much cleaner
black carbon emissions. These are now much larger than
than in winter. But precipitation (and hence wet removal)
in Europe, North America, and Russia combined. Black
is suppressed by the high particle numbers in the vicinity
carbon is important because it absorbs solar radiation and
of the fires (Andreae et al., 2004), possibly allowing much
leads to a possibly strong albedo reduction if deposited
of the black carbon to reach the Arctic where it is then
onto snow or ice (Hansen and Nazarenko, 2004). In a re-
deposited.
cent model study, Koch and Hansen (2005) suggested that
8
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Days air north of 70º N
southern Asia is now the dominant source of black carbon
16
in the arctic upper troposphere and is comparable to the
European source near the arctic surface. However, if the
14
thermodynamic argument holds that in winter it is vir-
12
tually impossible for a southern Asian air mass to reach
the arctic lower troposphere (Carlson, 1981; Iversen, 1984;
10
Bowling and Shaw, 1992), then it is not clear how Asian
black carbon, given black carbon's short atmospheric life-
8
time of 6 2 days (Park et al., 2005), can intrude into the
6
polar dome in the model of Koch and Hansen (2005).
There is also controversy regarding transport from the
4
stratosphere to the arctic troposphere. It is known that tro-
2
popause folds, which are frequent at the Polar Front, also
2000
2001
2002
2003
2004
2005
occur at the Arctic Front (Rao and Kirkwood, 2005). Several
0 - 0.1 km
3 - 5 km
studies suggest a strong influence of transport from the
0.5 - 1.5 km
5 - 8 km
stratosphere on the ozone concentrations in the arctic free
troposphere (Dibb et al., 2003; Allen et al., 2003). Other stud-
Figure 2.5. Time series of monthly mean arctic age of air, averaged over
the central Arctic (north of 70º N) at different altitudes.
ies suggest a much lower frequency of stratosphere-tropo-
sphere-transport events reaching the lower troposphere in
the Arctic than at mid-latitudes (James et al., 2003; Sprenger
and Wernli, 2003), and their influence on surface ozone at
three days in the upper troposphere. In the most isolated
Alert seems small (Dibb et al., 1994). In fact, because po-
regions of the Arctic, air is exposed to continuous darkness
tential temperatures in the lower stratosphere are higher
for, on average, 10 to 14 days in December (Figure 2.6).
than in pollution source regions at mid-latitudes, even
Transport from the stratosphere to the lower troposphere
more diabatic cooling is required for stratospheric air to
is much slower in the Arctic than at mid-latitudes. In the
penetrate the polar dome.
central Arctic, for instance, the probability that air near the
In a recent paper Stohl (2006) used the Lagrangian par-
surface was transported from the stratosphere within 10
ticle dispersion model FLEXPART to construct a global
days is only about 1% in winter and 0.3% in summer. Air
data set of 1.4 million continuous trajectories. At the model
pollution can be transported into the Arctic along three
start, particles were distributed homogeneously in the at-
pathways: low-level transport followed by ascent in the
mosphere and were then transported for 5.5 years using
Arctic, low-level transport alone, and uplift outside the
both resolved winds from ECMWF (European Centre for
Arctic, followed by descent in the Arctic. Only the last
Medium-range Weather Forecasts) analyses and param-
pathway is frequent for pollution originating from North
eterized turbulent and convective transport. Based on this
America and Asia, whereas European pollution can follow
data set, a climatology of transport in and to the Arctic was
all three pathways in winter, and pathways one and three
developed. It was found that the time air resides continu-
in summer.
ously north of 70º N its `arctic age' is highest near the
Sensitivities of arctic air masses to emissions of air
surface in the North American sector of the Arctic. North
pollutants were calculated for transport times of up to 30
of 70º N and near the surface, the mean arctic age of air is
days before the air masses reached the Arctic. They were
about a week in winter and two weeks in summer (Figure
highest over Siberia and Europe in winter and over the
2.5). Arctic age decreases rapidly with altitude to about
oceans in summer. Using an inventory for anthropogenic
Winter
Summer
80°N
80°N
70°N
70°N
0
1
2
3
4
5
6
7
8
9
10
0
2
4
6
8
10
12
14
16
18
days
Figure 2.6. Mean arctic age of air in the lowest 100 m of the atmosphere in January (winter) and July (summer).
9
Chapter 2 · Sources of Acidifying Pollutants and Arctic Haze Precursors
black carbon emissions, it was found that black carbon
Black carbon, parts per trillion mass
potential source contributions from Asia to the Arctic are
120
much lower than those from central Europe and Rus-
Winter
sia (Figure 2.7), despite much higher emissions in Asia.
100
The Eurasian origin of black carbon in the Arctic is also
confirmed by the Danish Eulerian Hemispheric Model
80
(DEHM) (Figure 2.8). For time scales of five and ten days,
south Asian black carbon potential source contributions
60
near the arctic surface are only 1.6 and 10% of the corre-
sponding European values. Using an inventory for black
40
carbon emissions from boreal and temperate forest fires,
black carbon potential source contributions to the Arctic
20
(particularly from fires in Siberia) were much larger than
anthropogenic black carbon potential source contributions
0
in summer (see also section 2.1.5). Boreal forest fires may
even dominate the annual total black carbon budget in the
120
Arctic in years of strong burning. Measurements of black
Summer
carbon (or light absorption by aerosols) at the surface and
100
aloft from aircraft combined with modeling studies such
as that of Stohl (2006) are required to further define sources
80
of black carbon to the Arctic.
60
40
2.3. Emissions estimates used in modeling
20
Several emissions areas in the northern hemisphere con-
tribute to air pollution in the Arctic. Emissions scenarios for
0
the model studies in this assessment were based on global
0
5
10
15
20
25
30
Days back in time
emissions on a 1º x 1º grid from the EDGAR database for
Anthropogenic emissions
Forest fires
SO2, NOX, and VOC for 1990 and 1995. Emissions of am-
European
European
monia (NH
American
American
3) and NOX from lightning and soil emissions
Asian
Asian
from the global GEIA (which provides gridded data, aggre-
gated according to various categories at global or regional
Figure 2.7. Black carbon potential source contributions from conti-
scales, from several inventories) were also used. Modified
nents as a function of transport time to the Arctic, for the subset that
gridded EDGAR emissions were obtained for 2000, 2010,
also reaches a minimum arctic altitude below 1000 m, for winter and
summer.
2020, and 2030 based on IIASA emissions scenarios (Frank
Dentener, European Commission Joint Research Centre,
height, km
pers. comm.). For 2010 and 2020 there were two types of
emissions scenarios: The `Current LEgislation' (CLE) sce-
nario, which reflects the current perspectives of individual
12.5
countries on future economic development and takes into
account the effects of presently agreed emission control
legislation in the individual countries. The other scenario
10.0
is the `Maximum technically Feasible Reduction' (MFR)
scenario. This projects the scope for emission reductions
offered by full implementation of presently available emis-
7.5
sion control technologies, while maintaining the projected
levels of anthropogenic activities. Global emissions are re-
distributed to the model grid, which is an extension of the
5.0
old 150 km resolution EMEP grid. The global emissions
scenarios are combined with the EMEP expert emissions of
SO2, NOX, NH3, and VOC for 1985, 1990, 1995 to 2003, and
2.5
expert emissions estimates for 2010 and 2020 (see http://
webdab.emep.int). Specific information on emissions from
the Russian non-ferrous metal smelters is also used (see
0.0
section 2.1.1).
0
10
20
30
40
50
60
0
25
50
75
100
Figure 2.9 shows emissions of SOx-S and NOX-N for
Black carbon in air, ng/m3
Source contribution, %
2000. Figure 2.10 projects emissions of SOx-S for 2020 for
Europe
the CLE and MRF emissions scenarios. For SOx-S, a com-
South America
parison of the two figures indicates that the CLE scenario
Russia
East Asia
results in only small changes in emissions, while the MRF
North America
scenario results in relatively large emissions reductions.
South Asia
Africa
This is also evident in Figure 2.11, which projects total
Figure 2.8. Concentration and origin profile of black carbon north of
emissions of both SOx-S (95 % of which is SO2) and NOx-N
the Arctic Circle as calculated by the Danish Eulerian Hemispheric
between 1990 and 2020.
Model (see section 3.7). Average for 1991 to 2001.
10
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
SO -S emissions in 2000 (total 52320 kt)
NO -N emissions in 2000 (total 21919 kt)
x
x
SO -S
NO -N
x
x
2000
2000
Figure 2.9. Estimated emissions of SOx-S and NOx-N in 2000.
1
5
10
50
100
500
1000
kt/grid cell/yr
Projected SO -S emissions in 2020 (CLE; total 51268 kt)
Projected SO -S emissions in 2020 (MFR; total 20199 kt)
x
x
SO -S
SO -S
x
x
CLE 2020
MFR 2020
Figure 2.10. Projected SOx-S emissions in 2020 for the CLE and MFR emissions scenarios.
1
5
10
50
100
500
1000
kt/grid cell/yr
SO -S emissions, Mt
NO -N emissions, Mt
x
x
70
25
CLE
MFR
60
20
50
40
15
30
10
20
5
10
0
0
1990
2000
2010
2020
1990
2000
2010
2020
Figure 2.11. Projected total emissions of SOx-S and NOx-N between 1990 and 2020 for the area shown in Figure 2.10.
11
Chapter 3
Concentrations and Deposition of Acidifying Pollutants
Lars R. Hole, Jesper Christensen, Veronika A. Ginzburg, Vladimir Makarov, Natalia A. Pershina, Alla I. Polischuk,
Tuija Ruoho-Airola, Peotr Ph. Svistov, and Vitaly N. Vasilenko
The first AMAP assessment on acidification in the Arctic
3.1. Atmospheric and transport processes
(Kämäri, 1998) highlighted that although there are few
for air pollutants in the Arctic
sources of acidifying air pollutants within the Arctic, high-
latitude ecosystems are particularly sensitive to pollution
and that some effects of acidification are evident even in
The atmosphere is an effective oxidizing medium, where
low deposition areas. Adverse effects of acidifying pollut-
the emitted sulfur and nitrogen compounds readily under-
ants were noted in fish populations in acidified lakes and
go reactions to oxidize further to sulfate and nitrate. Both
in forests and natural vegetation, including lichens. The
gas phase and liquid phase reactions in the atmosphere are
assessment described the few but significant point sources
important routes to the acidic sulfate and nitrate products.
of acidifying air pollutants in Norilsk (on the Taymir Pe-
The oxidized sulfur and nitrogen compounds can exist in
ninsula in northern Siberia) and on the Kola Peninsula.
gaseous form, bound in particles, or dissolved in cloud
These were the areas of greatest concern, in addition to the
droplets and rainwater. The atmospheric chemistry associ-
northern regions of Norway and Finland. The assessment
ated with the acid deposition process, especially the part
focused on sulfur as this was considered to dominate the
starting with nitrogen oxides, is complex.
acidification issue.
Emitted oxides and their reaction products are tran-
The assessment concluded that there had been no
sported with the airflows and simultaneously removed
trends in atmospheric concentrations of acidifying com-
from the atmosphere by dry and wet deposition processes.
pounds in either Canada or Alaska since the early 1980s,
Dry deposition rates of gaseous and particulate compounds
but that there were decreasing trends on Svalbard. There
are controlled by physical and chemical characteristics of
were no background data for Russia. It was considered that
the compounds, the properties of the ground surface, and
around 75% of the deposition could occur in the form of
the state of the atmospheric boundary layer (Sehmel, 1980;
dry deposition, but there were insufficient observations to
Wesely and Hicks, 2000). Wet deposition is highly affected
confirm this. Model results for sulfur dioxide (SO2) and sul-
by the amount of precipitation, but also by the type of
fate (SO4) compared well with observations for times series
aqueous phase (e.g., liquid water, snow, or ice). Wet depo-
at Station Nord (Greenland) and for long-term averages
sition removes the acid components from the atmosphere
at several EMEP stations. The assessment concluded by
by several processes, which occur both within and below
recommending that air chemistry stations be established in
clouds. Although atmospheric lifetimes of SO2, nitrogen
Alaska and eastern Russia, and that further resources were
oxides (NOX), and their oxidation products are of the order
required to support model verification (inter-comparison),
of a few days (Schwarz, 1979; Levine and Schwarz, 1982;
particularly in relation to dry deposition.
Logan, 1983), in the High Arctic during winter the atmosp-
heric half-life of sulfate has been reported to reach as much
as two weeks or more (Barrie, 1986). Transport distances
range from hundreds to thousands of kilometers (Seinfeld
and Pandis, 1998). Thus, besides the actual emissions, ma-
ny additional factors affect the observed concentrations
hv, O
OH
OCS
SO2
SO2-
4
and trends of the compounds involved in the acid deposi-
tion process, including their relative concentrations in the
atmosphere, the reversible nature of some reactions, and
Tropopause
meteorological conditions.
H O
2
2
S(+4)
S(+6)
3.1.1. Sulfur
O
OCS
3
Figure 3.1 shows schematically the atmospheric cycle for
OH
sulfur compounds. H2S, DMS, CS2, and OCS are sulfidies.
S(-2)
S(+4)
S(+6)
The reactions are controlled by many factors, both chemical
and physical, for example the concentrations of oxidizing
CS
SO
SO2-
2
2
4
OH, NO
agents (elemental oxygen, O; OH radicals; and ozone, O
3
OH
3)
DMS
or photons (hv), concentration ratios of different air pol-
H S
Other
Other
lutants, and the acidity of liquid water associated with
2
S(+4)
S(+4)
particles and humidity (Seinfeld and Pandis, 1998). The
atmospheric chemistry of the sulfur cycle is dominated
Emission
Deposition
by OH radical reactions in the gas phase which lead to the
production of gaseous sulfuric acid (H2SO), and by gaseous
and aqueous phase reactions between SO2 and hydrogen
Surface
peroxide (H2O2) and O3. The lack of sunlight in the Arctic
for large parts of the year (see section 2.2 on arctic age of
air) limits the production of the OH radical and hydrogen
Figure 3.1. Atmospheric sulfur cycle (Berresheim et al., 1995).
peroxide (Barrie, 1986). The seasonality of SO2 oxidation to
12
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
sulfate (SO 2-
4 ) is important in prolonging the presence of
O
1
3
sulfate aerosols in the Arctic into April and May (Barrie and
O( D)
N O
OH
2
NO
NO2
HNO3
hv
Hoff, 1984; Barrie, 1996). S(-2), S(+4) and S(+6) in Figure 3.1
illustrate different oxidation levels of sulfur.
hv
N2
Stratosphere
Troposphere
Transport
Transport
3.1.2. Nitrogen
Lightning
O3
OH
Figure 3.2 shows schematically the atmospheric cycle
N
N
Combustion
NO
NO
2O
2
HNO
2
3
hv
for nitrogen compounds. Important nitrogen-containing
compounds are molecular nitrogen (N
OH
2), nitrous oxide
H O
2
(N2O), nitric oxide (NO), nitrogen dioxide (NO2), nitric
Denitrification
acid (HNO
-
H O
2
+
-
3), inorganic and organic nitrates (NO3 ), am-
NH
NH
NO
3
4
3
monium (NH +
4 ), and ammonia (NH3). The same oxidizing
Fixation
agents as for sulfur control the reactions.
Deposition
Deposition
The previous AMAP assessment (AMAP, 1998) repor-
ted that, as air masses move from mid-latitudes to the
Arctic, the set of chemical species available to drive nitro-
Surface
gen chemistry changes. The nitrogen chemical cycle has a
considerable `dark' component, with night-time reactions
between NO2 and O3 to form NO3 and N2O5 (nitrogen pen-
Figure 3.2. Atmospheric nitrogen cycle (Seinfeld and Pandis, 1998).
toxide) (Seinfeld and Pandis, 1998). The reaction of N
The important anthropogenic nitrogen oxide emissions are illustrated
2O5
with water produces gas phase nitric acid (HNO
as combustion, which oxidizes molecular nitrogen (N2) to nitric oxide
3) and par-
(NO), and further oxidation by ozone (O
ticulate nitrate. The formation of the peroxy acetyl nitrate
3) to NO2.
(PAN) is especially important in the Arctic, since the alkyl
nitrate chemical removal mechanism is also dependent on
Air Research). In addition, a few national stations report
light and temperature. PAN is an atmospheric reservoir for
data to AMAP. Some stations have reported data since
nitrogen in the arctic winter, and these alkyl nitrates may
the mid-1970s. Figure 3.3 and Table 3.1 identify the sta-
contain as much as 75 to 80% of the airborne oxidized ni-
tions for which data have been used in this assessment.
trogen (NOy) (Bottenheim et al., 1993; Singh et al., 1992).
Most stations are located in the European Arctic. Very few
measurements from North America were available for this
assessment (data were available from Snare Rapids, Alert,
3.2. Distribution of monitoring stations
and Barrow). The only station in northern Greenland (Sta-
tion Nord) was closed in 2002.
The EMEP network was established to support the
The AMAP atmospheric monitoring network consists of
1979 Geneva Convention on Long-range Transboundary
a number of stations distributed throughout the Arctic.
Air Pollution (LRTAP). The LRTAP Convention provides
Most of these are EMEP (Co-operative Programme for
a broad framework for cooperative action on controlling
Monitoring and Evaluation of the Long-range Transmis-
and reducing the impact of transboundary air pollution
sion of Air pollutants in Europe) stations that also report
and establishes a process for negotiating measures to con-
to the AMAP database at NILU (Norwegian Institute for
trol emissions of air pollutants through legally binding
Palatka
Air monitoring
Precipitation monitoring
Air and precipitation monitoring
Ust-Moma
Arctic haze monitoring
Barrow
Deputatskiy
Russian precipitation network station
Zhigansk
Snare Rapids
Kyusyur
Tiksi
Polyamiy
Norilsk
Turukhansk
Dikson
Alert
Urengoy
Nord
Zeppelin
Ny-Ålesund
Hornsund
Naryan-Mar
Nikel Murmansk
Svanvik
Krasnoshelie
Arkhangelsk
Karasjok
Padum
Pinega
Jergul
Mud'yug
Abisko
Zarechensk
Janiskoski
Figure 3.3. Locations of background monitoring sta-
Oulanka
Reykjavik
Irafoss
Tustervann
tions for air and precipitation chemistry, and arctic haze
Bredkäl
and the Russian precipitation chemistry stations from
which data were used in this assessment.
13
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
Table 3.1. AMAP stations included in this assessment.
Operating
Height above
Station name
Latitude
Longitude
Mediaa
period
sea level (m)
Alert
82º 28' N
62º 30' W
A
1980-
210
Nord
81º 36' N
16º 40' W
A + P
1990-2002
20
Oulanka
66º 19' N
29º 24' E
A + P
1990-
310
Irafoss
64º 5' N
21º 1' W
A + P
1980-
61
Reykjavik
64º 8' N
21º 54' W
P
1994-
61
Jergul (with Karasjok)
69º 27' N
24º 36' E
A + P
1977-1997
255
Tustervann
65º 50' N
13º 55' E
A + P
1977-
439
Zeppelin (Ny-Ålesund)
78º 54' N
11º 53' E
A
1989-
474
Svanvik
69º 27' N
30º 2' E
A + P
1986-
30
Karasjok (with Jergul)
69º 28' N
25º 13' E
A + P
1997-
333
Ny-Ålesund
78º 55' N
11º 55' E
P
1982-
1
Janiskoski
68º 56' N
28º 51' E
P
1990-
118
Pinega
64º 42' N
43º 24' E
P
1979-
28
Bredkäl
63º 51' N
15º 20' E
A + P
1979-
404
Abisko
68º 21' N
18º 49' E
P
1992-
390
Nikel
69º 24' N
30º 12' E
A
Snare Rapids
63º 31' N
116º 00' W
P
Barrow
71º 19' N
156º 40' W
Hornsund
77º 00' N
15º 33' E
P
1988-
7
a A: air; P: precipitation
protocols. Within this process, the main objective of the
sub-divided into three sectors on the basis of climate: the
EMEP program is to provide signatories and subsidiary
Atlantic, Siberian, and Pacific sectors.
bodies to the LRTAP Convention with qualified scientific
Monitoring data from the Zarechensk, Padun, Kras-
information to support the development and further evalu-
noshelie, Naryan-Mar, Urengoy, Turukhansk, Zhigansk,
ation of the international protocols on emission reductions
Deputatskiy, Ust-Moma and Palatka sites were used to
negotiated within the convention. The EMEP program is
represent background levels. Except for the two EMEP-sta-
undertaken in collaboration with a broad network of scien-
tions reported here (Janiskoski and Pinega), there are no
tists and national experts that contribute to the systematic
monitoring sites in the Russian Arctic representing regional
collection, analysis, and reporting of emission data, meas-
or background conditions for air quality. Urban monitoring
urement data, and integrated assessment results. Since
sites in the industrial areas of five Russian cities Nikel,
2003, EMEP has had stations in 35 countries, varying from
Monchegorsk, Murmansk, Salekhard and Norilsk were
one to 25 stations in each country.
used to estimate pollutant levels in air at the nearest in-
The Acid Deposition Monitoring Network in East Asia
dustrial center.
EANET was established in 1998. Twelve countries
Pollutant levels and acidity of snow cover in the Rus-
participate in the network, which stretches from the Rus-
sian Arctic are monitored at 99 stations.
sianMongolian border to Indonesia. Regular monitoring
activity under EANET began in 2001. The EANET moni-
toring program includes, among others, measurements
of SO2, NO2, and O3 in air; acidity, electrical conductivity,
sulfate, and nitrate in precipitation; soils and vegetation,
and freshwater systems. Measurements are made at urban,
regional, and remote sites. Data from Mondy, the Russian
Table 3.2. Russian precipitation stations.
remote monitoring site situated near the RussianMon-
Station name
Start
Latitude
Longitude
golian border, can be used to estimate long-range atmos-
Atlantic sector
pheric transport from East Asia to the Arctic.
Zarechensk
1990
66.7º N
31.4º E
In 2002, 24 stations reported data relevant to acidifica-
Padun
1991
68.6º N
31.8º E
Murmansk
1991
69.0º N
33.1º E
tion and eutrophication to the AMAP database at NILU.
Krasnoshelie
1990
67.4º N
37.1º E
Most of these stations are located in Fennoscandia (see
Mud'yug
1958
64.9º N
40.3º E
Table 3.1).
Arkhangelsk
1991
64.6º N
40.5º E
The Russian national monitoring network for precipita-
Naryan-Mar
1962
67.7º N
53.0º E
tion chemistry has 130 monitoring stations. Precipitation
Siberian sector
Urengoy
1989
66.0º N
78.4º E
samples are analyzed at twelve regional analytical labora-
Dikson
1980
73.5º N
80.4º E
tories for between eight and eleven compounds, includ-
Turukhansk
1962
65.8º N
87.9º E
ing SO 2-
-
-
+
4 , Cl-, NO3 , HCO3 , NH4 , Na+, K+, Ca2+, Mg2+, the
Norilsk
1991
69.3º N
88.3º E
sum of ions, and pH. In addition to these stations, there
Polyarniy
1990
66.7º N
112.4º E
are 105 monitoring sites where pH only is analyzed. The
Zhigansk
1991
66.8º N
123.4º E
monitoring stations are unevenly distributed, with less
Kyusyur
1990
70.6º N
128.0º E
Tiksi
1990
71.6º N
128.9º E
than 40% situated in the vast Siberian region. Data have
Pacific sector
been recorded for up to 45 years at some stations. This as-
Deputatskiy
1990
69.3º N
139.7º E
sessment uses data from 18 monitoring stations within the
Ust-Moma
1990
66.5º N
143.2º E
Russian Arctic (see Table 3.2). The Russian Arctic may be
Palatka
1962
60.1º N
150.9º E
14
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
SO in air, g/m3
4
Alert (A)
Nord (N)
1.0
1.0
0.5
0.5
0
0
A
Zeppelin (Z) (Ny-Ålesund)
Karasjok/Jergul (K)
N
1.0
1.0
Z
J
K Sv
0.5
0.5
I
T
O
B
0
0
summer
Svanvik (Sv)
Janiskoski (J)
1.0
1.0
winter
0.5
0.5
0
0
Oulanka (O)
Tustervann (T)
1.0
1.0
0.5
0.5
0
0
Irafoss (I)
Bredkäl (B)
1.0
1.0
0.5
0.5
Figure 3.4. Trends in summer
and winter total sulfate con-
0
0
centrations in air within the
1980
1985
1990
1995
2000
2005
1980
1985
1990
1995
2000
2005
Arctic.
NO in air, g/m3
3
Alert (A)
Nord (N)
0.04
0.04
0.02
0.02
0
0
Zeppelin (Z) (Ny-Ålesund)
Karasjok/Jergul (K)
0.20
0.04
0.10
0.02
0
0
0.20
0.20
Janiskoski (J)
Oulanka (O)
0.10
0.10
0
0
Tustervann (T)
Bredkäl (B)
0.20
0.20
0.10
0.10
Figure 3.5. Trends in summer
0
0
and winter nitrate concentra-
1980
1985
1990
1995
2000
2005
1980
1985
1990
1995
2000
2005
tions in air within the Arctic.
15
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
3.3. Concentrations, distribution, and
trends in air and precipitation
There are few good time series for the main atmospheric
tistically significant, reductions have occurred since 1980.
compounds in the High Arctic. There are also few stations
Despite a flat or downward trend in SO2 concentrations in
that monitor both air and precipitation. Precipitation and
air at most sites, concentrations at Svanvik are still rela-
air chemistry data from the AMAP monitoring stations are
tively high and periodic peaks in concentration (`episodes')
presented in this assessment as average winter (Decem-
well above the recommended health limit (an hourly value
ber, January, February) and summer (June, July, August)
of 350 g/m3) are still common (Hagen et al., 2005). It is
values. In cases where a monthly value was missing, the
obvious that these high values can largely be attributed to
whole season was omitted from the analysis. Precipitation
the non-ferrous metal smelters in Nikel (Figure 3.6).
data were weighted according to precipitation amount.
Existence of a monotonic increasing or decreasing trend in
the time series after 1990 was tested using the non-para-
3.3.2. Precipitation
metric Mann-Kendall test at significance levels of p<0.1,
3.3.2.1. General pattern
p<0.05, and p<0.001 as a two-tailed test (Gilbert, 1987). The
estimate for the slope of a linear trend was calculated using
A decreasing trend in annual average sulfate concentra-
the non-parametric Sen's method (Sen, 1968). The Mann-
tions is also evident in the precipitation data (Figure 3.7).
Kendall test is suitable for cases where the trend may be
The data also show seasonal differences. At Ny-Ålesund
assumed to be monotonic, and thus with no seasonal cycle
(Svalbard) there are higher concentrations in winter, while
or cycle of any other type present in the data. Missing
at Zeppelin (approximately 1 km from Ny-Ålesund) the
values are allowed in the Mann-Kendall test and the data
seasonal difference in SO4 concentrations in air is smaller.
need not conform to any particular distribution. Sen's slope
estimator is the median of the slopes calculated from all
Sulfur dioxide
Barents Sea
pairs of values in the data series. The Sen's method is not
0
5
10 km
0
100
200
300 g/m3
greatly affected by data outliers and can be used when data
are missing (Salmi et al., 2002).
Kirkenes
NORWAY
3.3.1. Air
Figures 3.4 and 3.5 show time series for seasonal SO4 and
FINLAND
NO3 concentrations in air for the stations with the best air
quality time series in the Arctic. Table 3.3 summarizes the
Svanvik
trend statistics for these datasets as well as for SO2 and
NH
Nikel
Zapolyarnyy
4 in air. Significant downward trends are evident for
SO4 and SO2 in air at many stations, both in summer and
winter. Interestingly, it is only in summer that a significant
reduction in sulfate levels in air is seen in Svanvik (but not
RUSSIA
Nikel) despite emissions in the area having been signifi-
cantly reduced. For NO3 and NH4 there is no clear pattern,
although NH4 is decreasing at some stations.
An annual air quality study is undertaken in the border
areas between Norway and Russia (Hagen et al., 2005). Fig-
ure 2.1 showed the annual average SO2 concentrations in
air at Svanvik together with the total annual SO2 emissions
Figure 3.6. Mean concentrations of SO
from the non-ferrous metal smelters at Nikel. Although the
2 in air at Svanvik and Nikel
for different wind sectors for October 2004 to March 2005 (Hagen et
decreasing trend in air concentration after 1990 is not sta-
al., 2005).
Table 3.3. Trend statistics for acidifying pollutants in arctic air, 19902003.
Summer
Winter
SO2
SO4
NO3
NH4
SO2
SO4
NO3
NH4
Alert
NA
*-
NA
***-
**-
Oulanka
**-
**-
**-
Irafoss
NA
NA
NA
NA
NA
NA
Janiskoski
NA
*+
NA
Pinega
**+
Nikel
NA
NA
NA
NA
NA
NA
Tustervann
**-
*-
*-
Zeppelin (Ny-Ålesund)
*-
**-
**-
*-
Svanvik
*-
Karasjok/Jergul
**-
**-
**-
**-
Bredkäl
**-
*-
**-
**-
Nord
NA
**-
NA
NA: not available; significance level: * P<0.1, ** P<0.05, *** P<0.001; +/- indicate positive and negative trends respectively; an empty cell
indicates no significant trend.
16
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
SO in precipitation, mg/L
4
1.0
4.0
Zeppelin (Z) (Ny-Ålesund)
Karasjok/Jergul (K)
0.5
2.0
SR
0
0
1.5
1.5
Svanvik (Sv)
Janiskoski (J)
Z
J
K
1.0
1.0
Sv
Ab
P
T BO
0.5
0.5
0
0
summer
0.5
1.5
Abisko (Ab)
Oulanka (O)
winter
1.0
0.5
0
0
1.5
1.5
Tustervann (T)
Pinega (P)
1.0
1.0
0.5
0.5
0
0
1.5
0.5
Bredkäl (B)
Snare Rapids (SR)
1.0
0.5
Figure 3.7. Trends in weighted
summer and winter sulfate con-
0
0
centrations in precipitation wit-
1980
1985
1990
1995
2000
2005
1980
1985
1990
1995
2000
2005
hin the Arctic.
NO in precipitation, mg/L
3
0.5
0.5
Zeppelin (Z) (Ny-Ålesund)
Karasjok/Jergul (K)
0
0
0.5
1.0
Svanvik (Sv)
Janiskoski (J)
0.5
0
0
0.5
0.5
Abisko (Ab)
Oulanka (O)
0
0
0.5
0.5
Tustervann (T)
Pinega (P)
0
0
0.5
0.5
Bredkäl (B)
Snare Rapids (SR)
Figure 3.8. Trends in weighted
summer and winter nitrate con-
0
0
centrations in precipitation wit-
1980
1985
1990
1995
2000
2005
1980
1985
1990
1995
2000
2005
hin the Arctic.
17
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
3.3.2.2. Russian Arctic
It is also interesting to note that the significant increase
in pH at most stations is not necessarily coupled to a sig-
A significant amount of information on acidifying pollut-
nificant decrease in SO4 concentrations (Table 3.4). There
ants in precipitation is now available for the Russian Arctic.
is a notable marine sulfur contribution at several sites, in
The Russian Arctic is usually subdivided on the basis of
particular at Ny-Ålesund (not shown).
climatic conditions into the Atlantic, Siberian, and Pacific
There is no consistent pattern for NO3 in precipitation
sectors. Long-term data show that these regions also differ
(Figure 3.8) and for the summer data the only significant
in terms of precipitation chemistry and acidity. Trend sta-
increasing trend is at Ny-Ålesund, although the data for
tistics for a number of Russian stations are shown in Table
this site are sparse. More observations for nitrate both in
3.5. Trends are compared for the period since 1990, although
air and precipitation are required to better understand the
some measurements started as early as 1958 (Table 3.2).
development of nitrate pollution in the Arctic.
Table 3.4. Trend statistics for acidifying pollutants in arctic precipitation, 19902003.
Summer
Winter
Precip.
pH
SO4
SO4*
NO3
NH4
Precip.
pH
SO4
SO4*
NO3
NH4
Oulanka
*-
**-
**-
**-
**-
*-
*-
Irafoss
*+
NA
NA
*+
*-
NA
NA
Janiskoski
**-
**-
**-
**+
Pinega
*+
*-
*-
**-
**-
Tustervann
**+
***-
***-
***+
*-
***-
Karasjok/Jergul
**+
***+
**-
*-
*-
Svanvik
**+
**+
*-
*-
Zeppelin (Ny-Ålesund)
**+
*+
Bredkäl
***+
**-
*-
**+
**-
**-
**+
Abisko
*-
Snare Rapids
**+
**-
Hornsund
NA
NA
NA
*-
*+
NA
NA
NA
NA
SO4*: non-marine sulfate; NA: not available; significance level: * P<0.1, ** P<0.05, *** P<0.001; +/- indicate positive and negative trends
respectively; an empty cell indicates no significant trend.
Table 3.5. Trend statistics for Russian precipitation stations. Only significant trends are shown.
Winter
Summer
Mann-Kendall
p-value
Sen-slope
Mann-Kendall
p-value
Sen-slope
statistic
statistic
Atlantic sector
Zarechensk
Precipitation
-1.831
0.067
-1.678
pH
2.337
0.019
0.050
NH4-N
-1.736
0.083
-0.010
Padun
NH4-N
-1.960
0.050
-0.019
-1.870
0.062
-0.010
Murmansk
SO4-S
-2.741
0.006
-0.252
-2.303
0.021
-1.538
NO3-N
-2.141
0.032
-0.030
-1.916
0.055
-0.993
NH4-N
-3.558
0.000
-0.088
-2.901
0.004
-0.347
Krasnoshelie
-
Mud'yug
SO4-S
-1.831
0.067
-0.230
Arkhangelsk
Precipitation
1.952
0.051
0.968
SO4-S
-2.873
0.004
-0.198
NO3-N
-1.650
0.099
-0.022
NH4-N
-2.440
0.015
-0.035
Naryan-Mar
pH
2.146
0.032
0.038
2.198
0.028
0.043
SO4-S
-2.326
0.020
-0.043
NH4-N
2.623
0.009
0.043
Siberian sector
Urengoy
Precipitation
-2.562
0.010
-2.588
-1.830
0.067
-4.003
pH
1.837
0.066
0.071
-1.877
0.061
-0.059
SO4-S
-1.783
0.075
-0.032
Dikson
SO4-S
-2.728
0.006
-1.958
-2.524
0.012
-0.248
NO3-N
1.788
0.074
0.012
NH4-N
-2.180
0.029
-0.052
Turukhansk
Precipitation
-2.728
0.006
-1.958
Norilsk
-
Zhigansk
-
pH
2.658
0.008
0.040
Tiksi
pH
1.930
0.054
0.909
SO4-S
1.930
0.054
0.082
Pacific sector
Deputatskiy
-
Ust-Moma
SO4-S
2.573
0.010
0.068
Palatka
-
18
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Atlantic sector
The highest monthly concentration in precipitation was
The average monthly SO4-S concentration in precipita-
16.6 mg/L.
tion in the Atlantic sector of the Russian Arctic for 1990 to
The distribution of average monthly pH values in pre-
2004 was 1.8 ± 1.6 mg/L. Concentrations did not exceed 2.0
cipitation is shown in Figure 3.9. The most frequent pH
mg/L in more than 70% of samples. The highest monthly
values fell in the range 4.6 to 6.7 and their distribution
concentration in the industrialized regions (Murmansk,
mostly follows a normal distribution curve. The aver-
Arkhangelsk, Mud'yug) was 23.3 mg/L. In background
age pH value was close to 5.6. Values of 3.6 occurred in
areas (Zarechensk, Padun, Krasnoshelyie, and Naryan-
two samples only, while values above 7.4 occurred in 27
Mar) concentrations ranged from 0.3 to 0.8 mg/L.
samples. Acidity was highest during the spring and sum-
The average monthly NO3-N concentration in the At-
mer. Precipitation became less acidic from west to east.
lantic sector for 1990 to 2004 was 0.31 ± 0.3 mg/L. In 62%
In contrast to the Atlantic sector as a whole, for which
of samples concentrations did not exceed 0.3 mg/L. The
precipitation is mostly neutral, precipitation at the Mur-
highest monthly concentration was 3.61 mg/L. Concentra-
mansk station was mostly acidic (ranging from 3.2 to 6.5).
tions at background stations (Zarechensk, Padun, Kras-
Over 80% of precipitation events at Murmansk had a pH
noshelyie) were between 0.1 to 0.7 mg/L in more than
of less than 5.4.
85% of samples. The minimum nitrogen concentration cor-
The total amount of monthly precipitation in the At-
responded to the maximum amount of precipitation. The
lantic region varied from 25 to 78 mm, with the greatest
average monthly concentration of NH4-N in precipitation
quantities falling during spring and autumn. An inverse
was 1 ± 0.71 mg/L. Concentrations of NH4-N in precipita-
correlation was observed between pollutant levels and the
tion did not exceed 1.4 mg/L in more than 86% of samples.
total amount of precipitation per month.
Frequency
SO -S in precipitation, mg/L
4
200
5.0
Atlantic sector
Padum
Murmansk
4.0
150
3.0
100
2.0
50
1.0
0
0
pH
Padum
Murmansk
5.4
800
5.2
Siberian sector
5.0
600
4.8
4.6
400
4.4
4.2
200
4.0
1992
1994
1996
1998
2000
2002
2004
0
pH
4.8
M
160
4.7
Pacific sector
P
4.6
120
4.5
80
4.4
40
4.3
0
4.2
3
4
5
6
7
8
9
1.5
2.0
2.5
3.0
3.5
4.0
pH
SO -S, mg/L
4
Figure 3.10. Annual average sulfate sulfur concentrations and pH in
Figure 3.9. Frequency distribution for monthly pH values in precipi-
precipitation within the Russian Arctic at a heavily impacted site (Mur-
tation within the Russian Arctic for the Atlantic, Siberian, and Pacific
mansk) and a background site (Padun) since 1991, and the correlation
sectors for 1990 to 2004.
between sulfur and pH in samples from Murmansk.
19
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
Table 3.6. Concentrations of sulfur (mg/L), nitrogen (mg/L) and pH at background monitoring stations in the Russian Arctic, 1990-2004.
Sulfur (SO4)
Nitrogen (NO3)
Nitrogen (NH4)
N
N(NH4)/
S/ N
pH
average
min
max
average
min
max
average
min
max
N(NO3)
average
min
max
Winter (D, J, F)
Atlantic
0.43 0.020
1.40
0.16
0.01
0.68
0.35
0.03
1.21
0.51
2.19
0.84
5.70
4.35
6.60
Siberian
0.60 0.120
2.49
0.13
0.01
0.61
0.55
0.02
3.20
0.68
4.23
0.88
6.10
4.90
7.10
Pacific
0.68 0.230
2.23
0.14
0.01
0.67
0.56
0.04
2.62
0.70
4.00
0.97
6.50
5.10
7.50
Summer (J, J, A)
Atlantic
0.41 0.040
1.67
0.07
0.00
0.42
0.28
0.04
1.25
0.35
4.00
1.17
5.63
4.64
6.59
Siberian
0.40 0.130
1.49
0.12
0.01
0.34
0.35
0.01
1.70
0.47
2.92
0.85
6.20
5.20
7.20
Pacific
0.84 0.200
4.00
0.10
0.02
0.33
0.37
0.06
1.62
0.47
3.70
1.79
6.40
5.60
7.10
Year (12 months)
Atlantic
1.10
0.02
1.78
0.40
0.00
0.59
0.80
0.00
1.36
1.20
2.00
0.92
5.50
4.20
7.10
Siberian
0.70
0.12
2.65
0.20
0.01
0.68
0.30
0.01
3.20
0.50
1.50
1.40
6.00
4.70
7.30
Pacific
0.80
0.20
3.88
0.25
0.01
1.73
0.80
0.03
2.62
1.05
3.20
0.76
6.40
5.00
7.70
Since 1990 there has been a clear tendency for decreas-
SO -S in precipitation, mg/L
4
ing SO
Norilsk
Turukhansk
4 concentrations in precipitation together with an
100
1.0
increase in pH, especially near the main sources of an-
Turukhansk
thropogenic pollution. However, this was not the case for
Norilsk
80
0.8
precipitation in the Murmansk area which continued to
be acidic (Figure 3.10). Acidity of precipitation within the
60
0.6
Atlantic sector shows a strong correlation with sulfur con-
centration. There are no clear trends in NO3 for this period.
40
0.4
Trends in background areas of the Atlantic sector reflect
changes in neighboring industrialized regions but with
20
0.2
lower absolute values and less variability.
0
0.0
Siberian sector
1990
1992
1994
1996
1998
2000
2002
2004
Concentrations of SO4-S in the Siberian sector of the Rus-
sian Arctic ranged from zero to 143.06 mg/L in Norilsk
60
2.0
(which had the highest concentration throughout Russia).
The average level for the region (excluding Norilsk) did not
exceed 0.89 mg/L and there was a maximum concentration
50
1.6
of 16.85 mg/L. Concentrations were less than 0.5 mg/L
in 67% of samples. Background concentrations (Urengoy,
40
1.2
Turukhansk, and Zhigansk, Table 3.6) ranged from 0.2 to
1.2 mg/L with the minimum values in warm or spring
months. Concentrations below 1 mg/L were observed in
30
0.8
85% of samples only in Turukhansk. In contrast, concentra-
tions at Norilsk were between 30 and 60 mg/L in 87% of
20
0.4
samples. Average monthly concentrations at Norilsk were
J
F
M
A
M
J
J
A
S
O
N
D
50 to 60 times higher than at Turukhansk. Precipitation at
Norilsk was most polluted in May and September (Figure
3.11). Seasonal variation at the background station in Tu-
rukhansk followed the same pattern as at Norilsk.
N T
Figure 3.11. Annual and seasonal variations
Concentrations of NO
in sulfate sulfur concentrations in an indus-
3-N in precipitation within the
Siberian sector were less than 2.3 mg/L in 94% of sam-
trial area (Norilsk) and a background area
(Turukhansk) of the Siberian sector of the
ples. The maximum (about 6.5 mg/L) was observed at the
Russian Arctic.
Dikson and Kyusyur stations. The background level at Tu-
rukhansk did not exceed 0.2 mg/L in 73% of samples (SD
0.22 mg/L). The concentration of NH4-N in precipitation
at this station was higher than NO3-N in 90% of samples.
Background areas were characterized by an average NH4-
generally alkaline. The average pH is 6.7. pH values were
N level of 0.3 mg/L (SD 0.25 mg/L). The highest concentra-
between 6.2 and 7 in 80% of samples. Values of 3.9 were
tions (over 5 mg/L) mostly occurred at Kyusyur and the
observed twice and were over 7.7 in 20 samples. Precipita-
average value in this area was less 0.5 mg/L.
tion at Turukhansk had an average pH of 5.9.
There have been no significant trends in sulfur (Figure
3.11) or nitrogen levels in precipitation at Norilsk over the
Pacific sector
last 15 years. Although concentrations decreased in the
The Pacific sector of the Russian Arctic was characterized
mid-1990s they have now resumed the levels seen in the
by three stations: Deputatskaya, Ust-Moma, and Palatka.
early 1990s.
The average SO4-S concentration in precipitation was 1.4 ±
Precipitation over the Asian part of the Russian Arctic
1.7 mg/L. The highest concentration was 18.27 mg/L and
is generally less acidic than in the European part (Figure
the lowest was around the detection limit. Concentrations
3.9). Atmospheric precipitation in the Siberian sector is
did not exceed 1.7 mg/L in 85% of samples.
20
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Atlantic sector
The average NO3-N concentration in precipitation
SO -S, mg/L
Siberian sector
4
was 0.2 ± 0.2 mg/L. Concentrations did not exceed 0.14
Pacific sector
1.2
mg/L in 50% of samples. The highest concentration was
1.5 mg/L and the lowest was below the detection limit.
1.0
The average NH4-N concentration was around 0.64 ± 0.8
0.8
mg/L. Concentrations did not exceed 0.7 mg/L in 75% of
samples. The highest value was 8.01 mg/L and the lowest
0.6
was below 0.01 mg/L.
0.4
Precipitation in the Pacific sector has continued to
decrease in acidity over the last few years. The pH was
0.2
between 6.7 and 7.2 in 60% of samples and in eight sam-
ples the monthly pH was 7.6. The lowest value (5.0) was
0
observed twice. The annual precipitation was 400 to 500
mm, with 5 to 65 mm per month.
Total N, mg/L
Atlantic sector
1.4
A comparison of background pollutant levels in pre-
Siberian sector
Pacific sector
cipitation across the Russian Arctic is presented in Figure
1.2
3.12 and Table 3.6.
1.0
The acidity of precipitation within the Russian Arctic
did not increase over the study period (1990 to 2004). Acid-
0.8
ic precipitation was only observed on the Kola Peninsula,
0.6
where 60% of samples had pH values of 5.0 and episodic
precipitation events may have had values as low as 3.2.
0.4
Precipitation acidity within the Atlantic and Siberian sec-
0.2
tors decreased over the study period, while levels in the
Pacific sector remained the same.
0
Annual
Winter
Summer
Figure 3.12. Annual and seasonal variations in average background
levels of sulfate sulfur and total nitrogen in precipitation across the
Atlantic, Siberian, and Pacific sectors of the Russian Arctic.
Oulanka
Zeppelin (Ny-Ålesund)
Mean frequency, days
Mean frequency, days
30
30
Summer
Summer
1990-1993
1994-1997
1998-2001
15
15
0
0
45
45
Winter
Winter
30
30
15
15
0
0
N
NE
E
SE
S
SW
W
NW
N
NE
E
SE
S
SW
W
NW
Direction of air masses
Figure 3.13. Frequency and variation in the direction of air masses arriving at Oulanka and Zeppelin over 4-year periods during the 1990s.
21
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
3.4. Episodes and exposure to
Oulanka
sulfur and nitrogen
Mean SO exposure, g/m3/hr
2
0.4
Summer
1990-1993
1994-1997
1998-2001
Atmospheric transport and deposition data are usually
0.3
presented as mean values owing to the enormous amount
of data collected. However, a characteristic feature of ob-
0.2
servations in atmospheric concentration, particularly at
remote sites, is the occurrence of extreme peaks in concen-
tration. The 2002 AMAP assessment on the influence of
0.1
global change on contaminant pathways to, within, and
from the Arctic (Macdonald et al., 2003) stated that events
and short-term variations are important for the delivery
0
of contaminants to the Arctic. The assessment also stated
1.00
Winter
that the frequency of extreme events is likely to increase
with climate change.
0.75
A large proportion of the annual acid deposition is often
accumulated in just a few days. For example, a study in the
1990s found that the five worst days at Finnish background
0.50
stations could bring 20 to 30% of the annual bulk sulfate
load (Ruoho-Airola and Salmi, 2001). Days on which the
deposition exceeded the annual median value for the sta-
0.25
tion by a factor of 10 were counted as episode days. The
number of episode days in central and northern Finland
0
was significantly more frequent. The episodes were not just
due to high levels of precipitation since the mean sulfate
Zeppelin (Ny-Ålesund)
concentration in episodic rain was higher than the annual
Mean SO exposure, g/m3/hr
2
0.4
mean concentration. The episodes were mostly imported:
All year
the air mass arriving at the background stations had passed
over high emission areas outside Finland. High episodic-
0.3
ity has been detected in large areas of northern Europe
(Smith and Hunt, 1978) and eastern North America (Brook,
1995).
0.2
Sectoral exposure to atmospheric sulfur and nitrogen
compounds in the 1990s was examined at a Finnish arc-
0.1
tic station (Oulanka) and a Norwegian station (Zeppelin)
located in Spitsbergen (see Figure 3.13). Variations in the
atmospheric circulation over Fennoscandia have an im-
0
portant effect on the composition of the atmosphere. Daily
N
NE
E
SE
S
SW
W
NW
Direction of air masses
concentrations of gas phase and particulate sulfur and ni-
trogen compounds together with estimates of air transport
Figure 3.14. Mean exposure to sulfur dioxide at Oulanka and Zep-
pelin in relation to the direction of air masses arriving at the stations
routes (2-dimensional 925hPa trajectories obtained from
in the 1990s.
the EMEP MSC-W; http://www.emep.int/index_assess-
ment.html) enabled a sectoral examination of pollution
exposure. Sectoral exposure means the sum of the daily
the highest concentrations occurred in air masses from the
loads arriving from a specific sector over the period as-
north.
sessed. The sectoral distributions roughly illustrate the
In summer, exposure to SO2 at Oulanka is highest
relative importance of different transport directions to the
with the arrival of air masses from the northern sectors
dry deposition of sulfur and nitrogen (Ruoho-Airola et
which is where the high emissions on the Kola Peninsula
al., 2004).
are located. In winter, exposure is dominated by transport
Transport from the southwest and west dominated the
from the southern and southwestern sectors, and since
airflow at Oulanka, whereas transport was most frequent
the end of the 1990s from eastern sectors as well. At Zep-
from the north at Zeppelin. Figure 3.13 shows four-year
pelin exposure is dominated by transport from the north
mean values for the different transport sectors for each
and northeast throughout the year. Figure 3.14 shows the
station. The results are supported by an analysis of at-
sectoral distribution for SO2 exposure during the 1990s as
mospheric circulation during different states of the Arctic
four-year mean values.
Oscillation (Macdonald et al., 2003). In the 1990s, during a
Atmospheric sulfate is highest at Oulanka in air masses
positive phase in the AO index, southwesterly winds pre-
arriving from between the southeast and southwest. At
dominated at Oulanka, whereas at Spitsbergen the winds
Zeppelin the highest sulfate concentrations occur in air
were more from the north and east, particularly in winter.
masses arriving from the northeast and east. Exposure to
However, the frequency of westerly airflow decreased at
sulfate closely follows the pattern for SO2 with the excep-
Oulanka during the 1990s (Moberg et al., 2005).
tion that at Oulanka the winter distribution is also evident
At Oulanka, the SO2 concentrations in summer were
during summer.
highest in air masses arriving from the north and northeast,
The patterns of concentration and exposure for total
in winter concentrations were highest in air masses from
atmospheric ammonium (i.e., the sum of gaseous ammonia
sectors between the northeast and southwest. At Zeppelin,
and particulate ammonium) and total atmospheric nitrate
22
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Mean NH concentration, g/m3
Mean NO concentration, g/m3
4
3
0.20
0.20
1990-1993
1994-1997
1998-2001
0.15
0.15
0.10
0.10
0.05
0.05
0
0
Mean NH exposure, g/m3/hr
Mean NO exposure, g/m3/hr
4
3
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0
0
N
NE
E
SE
S
SW
W
NW
N
NE
E
SE
S
SW
W
NW
Figure 3.15. Mean concentration and exposure to nitrogen compounds at Zeppelin in relation to the direction of air masses arriving at the station
in the 1990s.
(i.e., the sum of gaseous nitric acid and particulate nitrate)
(Alaska, Finland, Sweden, Svalbard), a significant propor-
at Oulanka were similar to those for sulfate. At Zeppelin
tion (>50%) of SO4 was non-marine in origin. A few samples
the difference in concentration for the various air masses
from northern Europe showed a weak trend of decreasing
was small (Figure 3.15), thus exposure was highest from
pH with increasing SO4.
the northern sector as air masses were most frequent from
the north (Figure 3.13). However, recent research indicates
that the main source area for nitrate deposited on Svalbard
3.5.2. Russian Arctic
is Western Europe (Julin, 2003).
The pollution load and acidity of snow cover in the Russian
Arctic is monitored at 99 stations. Atmospheric deposition
loads of sulfur and nitrogen compounds per unit area are
3.5. Concentrations in seasonal snow cover
presented in Table 3.7. Average data for the European and
3.5.1. General pattern
Asian parts of Russia are also given for comparison (Be-
likova et al., 1984). Estimated atmospheric deposition loads
A recent study of snow samples from the end of winter
for the continental Russian Arctic (see Table 3.8) (Reviews
1996/1997 provides additional information about the acid
of the state of environment in Russian Federation, 1995)
deposition pattern in the Arctic (de Caritat et al., 2005). The
include estimates of the proportions of the sulfur and ni-
study was based on a total of 21 snow samples from 17 arc-
trogen derived from sources within the Russian Arctic and
tic locations, including sites in Norway, Sweden, Finland,
from transboundary sources. Deposition on the open sea
Svalbard, Russia, Alaska, Canada, Greenland, and Iceland.
areas of the Arctic Ocean is presented in Table 3.9 and the
Major element concentrations in the melted snow indicate
distribution of measured pH values in snow cover within
that the composition of most samples was consistent with
the Russian Arctic is shown in Figure 3.16. Not all snow
diluted seawater. Deviations indicate additional SO4 and Cl
monitoring sites reported data each year owing to difficul-
relative to seawater, suggesting an anthropogenic contribu-
ties with the transport and storage of samples under arctic
tion (Alaska, Finland, Sweden, Svalbard). The samples with
conditions.
the highest Na and Cl content (Canada, Russia) also have
Several conclusions can be drawn from the data pre-
higher Na:SO4 and Cl:SO4 ratios than seawater, suggesting
sented in Tables 3.7, 3.8, and 3.9 and Figure 3.16. It is clear
a slight contamination from (probably local) de-icing activi-
that the average level of atmospheric deposition in back-
ties. Local soil or rock dust inputs to the snow are indicated
ground areas of the Russian Arctic (i.e., outside the areas
by `excess' calcium (Alaska, Svalbard, Greenland, Sweden).
affected by the major industrial centers) is much lower than
No overall relationship was found between pH (range:
in the European and Asian parts of Russia to the south.
4.66.1) and total or non-marine SO4 (SO4*), but the data are
Also, that atmospheric deposition fluxes for both sulfur
too limited (with only one season sampled) to draw firm
and nitrogen decrease from west to east across the Russian
conclusions on long-range SO2 transport. In a few samples
Arctic. The highest levels of sulfur deposition occur within
23
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
Table 3.7. Average atmospheric deposition loads for sulfur and nitrogen in the Russian Arctic.
Atmospheric loads, kg/km2/yr
S
NOX-N
NH4-N
Total N
Kola Peninsula
Eastern part of peninsula (minimum values)
300
60
130
190
Entire region
452
66
137
203
50 km zone around smelters
800-3000
60-90
100-140
160-230
Arkhangelsk region (northern part)
180
100
142
242
Yamal Peninsula (entire region)
130
150
160
310
Taymir Peninsula
100
70
90
160
Entire region
50 km zone around Norilsk smelter complex
450-4000
70-120
80-120
150-220
Arctic Ocean coastal area and Yakutsk region
75
35
65
100
Arctic Ocean islands
Zemlya Frantsa Iosifa
140
42
125
167
Severnaya Zemlya and islands
90
50
100
150
Novosibirsk islands
40
10
65
75
European territory of Russia
810
180
450
630
Asian territory of Russia
350
90
180
270
Table 3.8. Atmospheric deposition loads for sulfur and nitrogen emitted from sources within the Russian Arctic and from transboundary sources.
Total deposition
Sources within
Transboundary
the Russian Arctic
sources
t/yr
%
t/yr
%
t/yr
%
Kola Peninsula
Sulfur
65500
100
46100
70.5
19400
29.5
Nitrogen (NOX)
9600
100
8000
8.3
8800
91.7
Total nitrogen
30000
100
Arkhangelsk region and Komi (Arctic part)
Sulfur
48400
100
5890
12.2
42510
87.8
Nitrogen (NOX)
13100
100
840
6.4
12260
93.6
Total nitrogen
46500
100
Tyumen region (Arctic part)
Sulfur
58600
100
1600
2.8
57000
97.2
Nitrogen (NOX)
13900
100
6600
47.5
7300
52.5
Total nitrogen
72600
100
Krasnoyarsk region (Arctic part)
Sulfur
286200
100
222000
77.5
64200
22.5
Nitrogen (NOX)
15900
100
1060
6.7
14840
93.3
Total nitrogen
142590
100
Yakutia, Magadan region and Chukotka (Arctic part)
Sulfur
70720
100
Nitrogen (NOX)
36400
100
Total nitrogen
104000
100
Table 3.9. Atmospheric deposition loads for sulfur and nitrogen compounds on the open sea areas of the Arctic Ocean.
Sulfur
NO3-N
Average flux , kg/km2/yr
Amount, t/yr
Average flux , kg/km2/yr
Amount, t/yr
Barents Sea
180 (110-220)
256000
60 (40-70)
85440
Kara Sea
90 (60-120)
79470
40 (25-65)
35300
Laptev Sea
30 (20-40)
19860
12 (5-25)
7900
East Siberian Sea
25 (20-30)
22800
10 (5-20)
9100
Chukchi Sea
30 (10-30)
11900
15 (10-20)
9000
Frequency of pH values, %
50
Kola Peninsula
Arkhangelsk region, Nenets
region, Komi republic
40
Yamalo-Nenets region
Taymir, Dolgano-Nenets region
30
Northern part of Yakutia, Kolima
Chukotka
20
Arctic Ocean islands
10
0
4.0 - 4.4
4.4 - 4.8
4.8 - 5.2
5.2 - 5.6
5.6 - 6.0
6.0 - 6.4
6.4 - 6.8
6.8 - 7.2
7.2 - 7.6
7.6 - 8.0
Figure 3.16. Acidity of snow cover across the Russian Arctic, 20012004.
24
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
concentric zones extending approximately 60 km from the
non-ferrous metal smelters on the Kola Peninsula and at
0 250 km
Laptev Sea
Norilsk. The tables and figure do not indicate acidic pre-
cipitation in winter over large areas of the Russian Arctic.
Tiksi
The average pH level in snow cover accumulated during
the polar winter (6-7 months) is typically above 5.0, al-
Markoka
Kolima
though acidic precipitation is sometimes seen in Taymir at
ana
Y
the outer limit of the area affected by industrial activities
in Norilsk.
Lena
The tables show that around 530 kt of sulfur, 89 kt of
nitrogen as nitrogen oxides, and 382 kt of total nitrogen are
Marha
Vilyi
Mirny
deposited on the continental region of the Russian Arctic
Aldan
Lensk
each year. The proportion estimated to have originated
Yakutsk
from Russian sources ranged from 12% (Arkhangelsk re-
gion, Komi, Tyumen region) to 7077% (Murmansk region,
pH
the northern Krasnoyarsk region). Annual deposition to
< 5.5
the seas of the Russian sector are estimated at 390 kt of
5.5 - 6.5
sulfur and 147 kt of nitrate nitrogen (Table 3.9).
6.5 - 7.0
Figure 3.16 shows that acidic snow cover occurred
7.0 - 7.6
in all regions. Acidic snow (snow with a pH of less than
Figure 3.17. pH in snow cover in Yakutia.
about 5.6) was most frequently observed in the Yamalo-
Nenetckiy region (65% of measurements). The lowest fre-
quency of acidic snow samples (1113%) was observed in
the Arkhangelsk region, Nenetckiy region, Komi republic,
and Taymir.
In seasonal snow in the area closest to the non-ferrous
metal smelters on the Kola Peninsula (Nikel, Zapolyarnyy,
underlying surface to the snow, and by the dissolution of
and Monchegorsk) pH levels are relatively high, at up to
contaminants in the snow cover during snowmelt.
6.06.8. This reflects neutralization by alkaline dust, which
Relatively high pH values (i.e., values near neutrality;
is a very important process in such areas. No seasonal
6.57.2) in snow cover in southeastern and eastern Yakutia
trends were evident in the pH values. Acidic snow with
reflect the arrival of cold air masses from the Pacific basin.
pH values of 4.6 to 5.2 is evident more than 100 km from
Snow acidification due to local sources has not been ob-
the non-ferrous smelters in the direction of the prevailing
served in Yakutia.
winds.
Because the pH of meltwater depends on the anion:cat-
The effects of emissions from Norilsk enterprises on
ion ratio and on the presence of alkaline compounds such
snow acidity are unclear. A snow acidity level of pH 6.0
as calcium and magnesium oxides which predominate in
dominates the area up to 250 km from Norilsk. Measure-
emissions, high pH values occur in snowmelt waters. High
ments at 12 monitoring sites up to 800 km to the south
dust levels in the air in industrial regions and a predomi-
and southeast of Norilsk found no pH values below 5.6.
nance of carbonates and calcium and magnesium oxides in
However, Norilsk emissions may have decreased the pH
the emissions, have lead to a sharp increase in pH values
value for snow cover to the west of Norilsk in the direc-
in some industrial areas: in the heavily impacted areas of
tion of Yamal.
western Yakutia snow pH values as high as 7.5 to 8.5 occur
The influence of sea salt on precipitation acidity is evi-
in some urban regions (Udachny: 7.85; Mirnyi: 8.2) and 8.0
dent in coastal areas of the Kola Peninsula, in the Arkhan-
to 9.4 in mining areas (the Mir diamond pipe: 9.45).
gelsk region, and in Chukotka. This is characterized by a
relatively small decrease in pH (to 5.0) compared to the
background level of 5.6.
Precipitation chemistry in Yakutia is mainly influenced
3.6. Pollution history from ice cores
by salts of continental origin. This is clear from the high
and lake sediments
calcium bicarbonate content, while salts of marine origin
comprise only 16% of the total salt content. Precipitation
in Yakutia has a pH close to 5.6 which is typical of unpol-
Whereas ice cores from Antarctic and Greenland ice sheets
luted precipitation (Figure 3.17). Emissions from the indus-
have been useful for extracting pollution history with high
trial sources (e.g., non-ferrous metal smelters) commonly
time resolution, ice cores from arctic ice caps have been
produce an alkaline reaction; thus precipitation and snow
considered difficult to use because summer melting and
cover in the vicinity of these sources have higher pH values
percolation of melt water are thought to blur the signal
than those further away. On the other hand, acidification
(Glowacki, 1997). However, it has recently been shown
of precipitation and snow cover occurs in areas remote
that by including a model for percolation effects (elution
from the sources (due to long range transport of sulfur
of ions), it is possible to reconstruct the `pristine' chemical
and nitrogen compounds) and the snow water here may
composition of ice caps such as Lomonosovfonna (78º51'
have low pH values.
N, 17º25' E, 1255 m above sea level) on Svalbard (Moore
The chemical composition of the snow cover is directly
et al., 2005). Even with 80% melting there is little distur-
dependant on the chemical composition of the precipita-
bance in chemical stratigraphy. This suggests that ionic
tion. Snow chemistry is also influenced by gas exchange
records from arctic ice caps are almost as reliable as those
between the snow cover and the atmosphere during the
from the large ice sheets where melting hardly occurs. This
long polar winter, by the transport of material from the
new type of knowledge and improved and more objective
25
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
techniques (e.g., Rasmussen et al., 2002) make it possible
NO concentration in ice, parts per billion
3
to interpret and compare trans-Arctic ice core chemical
500
Lomonosovfonna
records. However, nitrate and sulfate profiles from ice caps
in areas such as the Russian Arctic are still not available in
400
the open literature. Spatial variations in macroelements
in Severnaya Zemlya glaciers (Evseev et al., 2000) indicate
300
that averaged concentrations deposited on snow and ice
during warm stages of the Late Holocene and historical
200
epoch are 1.8 to 3 times higher than the corresponding
100
values for cold stages.
To interpret chemical signals in ice cores, knowledge
0
about deposition and post-depositional processes for the
different atmospheric compounds is required. Sharp et al.
500
(2002) studied the snow pack chemistry at the John Evans
Austfonna
Glacier at Ellesmere Island, Canada (79º40' N, 74º23' W)
400
A
and showed that the seasonal cycle in snow chemistry
L
closely reflects changes in the composition of the atmos-
300
pheric aerosol at Alert. There was some modification of ni-
trate concentrations by post-depositional processes. Mean
200
water-weighted solute concentrations in the snow pack are
largely independent of accumulation, while atmospheric
100
deposition tends to increase with accumulation. This sug-
gests that, for most species, wet deposition is the dominant
0
1800
1840
1880
1920
1960
2000
depositional process throughout the year. However, con-
centrations of calcium and potassium increase with both
Figure 3.18. Nitrate concentrations in ice cores from glaciers on Sval-
accumulation and elevation, implying an enhanced input
bard (Isaksson et al., 2003). The nitrate records from Lomonosovfonna
(adapted from Kekonen et al., 2002) and Austfonna (adapted from
from dry deposition of soil dust above 800 m elevation.
Watanabe et al., 2001) are for the period from 1800 to the present. The
Concentrations of sulfate are inversely related to accumula-
black line is a 25-year running mean.
tion, especially in the winter layer, suggesting a significant
input from non-precipitating events, such as dry deposition
or riming, during this period of very limited snowfall.
3.7. Modeling
Results of ice core drillings from Svalbard have been
published over the last ten years or so (e.g., Kekonen et
The Danish Eulerian Hemispheric Model (DEHM) system
al., 2002). Even though summer melting and percolation
consists of a weather forecast model, the PSU/NCAR Mes-
of melt water diffuses the signals, it is evident that use-
oscale Model version 5 (MM5) modeling subsystem (see
ful information about climate and pollution history can
Grell et al., 1994), which is driven by meteorological data
be obtained from these studies (Isaksson et al., 2003). The
from ECMWF (the European Centre for Medium-range
anthropogenic influence on the Svalbard environment is il-
Weather Forecasts), and a 3-dimensional atmospheric
lustrated by increased levels of non-sea-salt sulfate, nitrate,
transport model, the DEHM model. The model has a hori-
acidity, fly ash, and organic contaminants, particularly dur-
zontal resolution of 150 x 150 km and 20 vertical layers
ing the latter half of the 1900s (Figure 3.18). Decreased
and the coverage is close to hemispheric from nearly 10º
concentrations of some components in recent decades
N at the corners and 25º N at the midpoints of the model
probably reflect emissions reductions.
domain boundaries.
At Penny Ice Cap (67º15' N, 65º46' W) in the Canadian
The model system has been used to study transport
Arctic, Goto-Azuma et al. (2002) showed that summer melt-
of air pollution to the Arctic since 1991. There are several
ing resulted in disturbance of the seasonal signal. Sulfate
versions of this system. The original version of the DEHM
and nitrate concentrations at this site started to increase
model was developed for studying the long-range trans-
around 1900 and 1960 respectively. Bigler et al. (2002b)
port of SO2, SO4 and lead (Pb) to the Arctic (Christensen,
showed that, over the last 1200 years, significant changes
1997, 1999). The sulfur version was used in the first AMAP
in sulfate levels were present only during the industrial
assessment (see Kämäri, 1998) and the Pb version in the
era. Their Greenland ice core was taken at B20 (78º50' N,
AMAP heavy metals assessment (AMAP, 2005). The model
36º30' W, 2150 m above sea level).
was further developed to study transport, transformation,
To date, lake sediments have been considered of limited
and deposition of reactive and elemental mercury, and this
value as archives of the atmospheric pollution history since
version was also used in the heavy metals assessment, see
around 1800 because the temporal resolution has been too
also Christensen (2004) and Heidam et al. (2004). Other ver-
low. However, recent research shows that human activity
sions calculate the concentrations and deposition of vari-
has induced climate-driven regime shifts in lakes across the
ous pollutants (Frohn et al., 2002, 2003) through the inclu-
Arctic (Smol et al., 2005) and the effect of increased nitrogen
sion of an extensive chemistry scheme, and transport and
deposition cannot be excluded. Increased nitrate concen-
exchange of atmospheric carbon dioxide (Geels et al., 2004)
trations throughout the 1990s also correspond well with
and persistent organic pollutants (Hansen et al., 2004).
changes in diatom populations found in lake sediments
This assessment used both the original version of the
on Svalbard and at other arctic sites (Wolfe, University
model and the extensive chemical version of the model.
Centre on Svalbard pers. comm., 2005). Increased nutrient
The original version includes two species: SO2 and SO4 and
deposition and climate change are possible explanations
the chemical transformation of SO2 to SO4 is described by
for the consistent pattern of biological change observed
a simple linear transformation depending on time of year
(e.g., relative appearance of species).
and latitude. The extensive chemical version includes 63
26
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
species and more than 120 reactions that describe the chem-
3.7.1. Validation of the system for
istry of sulfur oxides, nitrogen oxides, reduced nitrogen
temporal trend analysis
(NHX), volatile organic compounds, and ozone. The simple
version of the DEHM model was used for the long-term
The chemical version of the DEHM model was run for the
calculations because it is up to 30 times faster than the
period 1991 to 2002 using actual emissions estimates for
version with the large chemical scheme. Emissions used as
each year. Mean air concentrations of SOX (i.e., SO2+SO4)
Fig_C$JA
input to the models are described in section 2.3.
and NO3 are shown in Figure 3.19 for 2000. The SOX results
show clearly the very high concentrations around Norilsk,
concentration in 2000, µg/m3
NO concentration in 2000, µg/m3
SOx
3
0.2 0.4 0.6 1.2 2.0 3.0 4.0 8.0 12.0
0.2 0.4 0.6 1.2 2.0 3.0 4.0 8.0 12.0
Figure 3.19. Mean surface air concentrations of sulfur oxides (i.e., sulfur dioxide plus sulfate) and nitrate in 2000.
Fig_C$JB
SO deposition in 2000, mg/m2/yr
NO deposition in 2000, mg/m2/yr
x
3
20
70 120 170 220 270 320 370 420
20
70 120 170 220 270 320 370 420
Figure 3.20. Total deposition of sulfur oxides (i.e., sulfur dioxide plus sulfate) and nitrate in 2000.
27
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
Measured SO -S, g/m3
Measured SO -S, g/m3
2
4
7
N = 97
N = 123
SO
means: calculated = 1.09
SO
means: calculated = 0.44
2
4
measured = 0.85
measured = 0.34
correlation = 0.63
NO47
1.0
correlation = 0.55
6
5
NO47NO47
NO47
NO47
NO47 FI22
4
IS91
NO47
IS91
IS91
IS91
IS91
FI22
FI22
NO47
RU01
RU01
NO47
NO47
NO47
NO47
NO47
IS91
3
0.5
FI22
FI96
RU01
FI22 NO47
NO47
RU01
NO30
RU13
NO47
NO47
RU01
FI22
RU01
NO30
FI22
FI22
RU13
SE05 FI22RU13
RU13
RU01
NO30
FI22
RU01
RU01
US04
US04
NO15
FI96
RU13
RU13
US04
NO30
SE05
NO47 RU13
RU01
RU13
NO30
NO30
FI96
FI96
2
US04
US04
US04
NO47
SE05
US04
CA42
IS02
SE05
SE05
SE05
NO30
NO30
NO15
NO30
FI96
NO15
RU01
NO15
NO15
NO15
RU01
IS02
NO30
NO15
CA42
CA42
RU01
CA42
IS02
SE05
NO15
RU01
RU01
DK10
SE05
DK10
DK10
NO15
SE05
RU01
RU01
IS02
IS02
IS02 NO42
NO42
NO42
NO42
CA42
IS02
DK10
NO42
NO15SE05
NO42
1
RU01
CA42 NO42
NO30
IS02
NO42
NO42
RU01 FI22
CA42
CA42
CA42
DK10 NO42
NO30
IS02
FI22
FI22
IS02
FI22
RU13
CA42
DK10
DK10
NO30
SE05
NO30
NO30
FI96
NO30
NO30
NO30
FI22
FI22
FI22
FI22
FI22
FI96
RU13 RU13
RU13
SE05
NO30
RU13
04 NO30
FI22
FI96
FI96
10
RU13
RU13
04
O42
NO15
04
04
04
0
O
O
1 SE05
SE05
FI96
2
42
RU13
SE05
2
SE05
42
O4
1 2
O42
O4
40
10
1 2
2
NO15
NO15
NO15
NO15
0
SE05
10
NO15
NO15
O15
0
0
1
2
3
4
5
6
7
0
0.5
1.0
Modeled SO -S, g/m3
Modeled SO -S, g/m3
2
4
Measured NO -N, g/m3
Measured NH -N, g/m3
3
4
N = 109
N = 92
NO
means: calculated = 0.15
0.3
NH
means: calculated = 0.18
3
4
measured = 0.05
measured = 0.16
correlation = 0.28
0.4
correlation = 0.74
RU13
RU13
RU13
0.3
RU01
NO47 NO47
0.2
SE05
FI22
RU13
NO47
NO47
SE05
RU13
NO47
RU13
RU01
RU01
RU01 RU01
NO47 RU01
SE05
SE05
RU01
SE05
SE05
RU01
0.2
RU01
SE05
FI22
RU13
NO30
NO30
NO30
FI22
FI22
NO47
SE05
FI22
NO30
NO30
SE05
NO15
NO30
NO30
NO30
SE05
NO30
FI22
NO30
FI22
NO15 FI22
NO30
NO15
NO15
0.1
NO15
NO30
FI22
FI96
NO30
NO15
NO30
NO47
NO42
NO42
SE05
FI22
NO30
FI22
NO30
NO47
FI22
FI22
FI96
FI96 FI96
NO15
NO42
NO30
NO47
NO30
NO47 NO47
FI96
FI22
FI22
0.1
NO15
FI22
FI22
NO30
RU01
FI22
NO47
CA42
US04
RU01
RU01
RU01 FI96
FI22
SE05
NO42
NO42
NO47
RU01
IS91
IS91
FI22
FI96
NO42
IS91
FI96
RU01
NO30
SE05
SE05
SE05
FI96
US04
US04
US04 NO42
IS91
RU01
FI96
SE05
RU01
NO15
SE05
SE05 SE05
NO42
US04
US04
NO42
IS91
SE05
A4
A 2
RU01
NO15
RU13
4
CA42
CA42
NO42
RU13
NO42
DK10
US04
NO42
NO42
NO15
NO15
NO15
DK10
DK10
IS91
A42
RU13
NO15
NO15
NO15
CA42
CA42
NO42
NO42
NO42
RU13
CA42
CA42
RU13
RU13
K10
RU13
NO42
RU13
NO42
DK10
CA42
CA42
CA42
CA42
CA42
CA42
DK10
DK10
CA42
DK10
CA42
CA42
DK10
DK10
DK10
DK10
DK10
0
0
0
0.1
0.2
0.3
0
0.1
0.2
0.3
0.4
Modeled NO -N, g/m3
Modeled NH -N, g/m3
3
4
Figure 3.21. Scatterplots comparing measured and modeled annual average concentrations of sulfur dioxide, sulfate, nitrate and ammonium at
arctic monitoring stations for 1991 to 2000. Labels on graphs refer to EMEP station codes.
and the high concentrations across the Kola Peninsula,
the two datasets for SO2 and SO4, and a particularly good
and the elevated concentrations associated with oil-related
correspondence for NH4. The model predicts the annual
activities at Barrow. The NO3 results do not indicate any
averages for most arctic stations well, but over predicts
local hot spots within the Arctic and that concentrations de-
NO3 concentrations for stations in the European Arctic.
crease from the source areas in Europe and North America
Figures 3.22, 3.23 and 3.24 show time series of measured
towards the Arctic. In Figure 3.20 total deposition of SOX
and modeled monthly mean concentrations of SO2, SO4 and
and NO3 for 2000 are shown, and the general patterns are
NO3 for some arctic monitoring stations. For Tustervann
similar to those for air concentrations in Figure 3.19.
(Figure 3.22) there is reasonable agreement between the
For arctic areas the model results were compared with
modeled and measured data for both SO2 and SO4, but poor
measurements from a range of background monitoring
agreement for NO3 concentrations. This is also the case for
stations (Figure 3.21). A scatter plot comparing measured
Zeppelin (Figure 3.23). The results for Janiskoski (Figure
and modeled annual average concentrations of SO2, SO4,
3.24) show that for this station the model overestimates
NO3 and NH4 for all arctic stations for each year between
the SO2 concentrations for most years, except after 1999.
1991 and 2000 shows a reasonable correspondence between
The main reason for this is that the spatial distribution
28
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
of official EMEP expert emissions for the Kola Peninsula
Zeppelin
was changed in 1999 before 1999 high emissions were
SO -S concentration, g/m3
2
attributed to the grid cell in which Janiskoski is placed
1.0
modeled
whereas after 1999 these emissions were allocated to the
measured
0.8
neighboring grid cell. These changes in the distribution of
0.6
emissions also result in a better performance of the mo-
del for SO
0.4
4. The model is able to predict well monthly
variations in SO2 and SO4 for a large part of the Arctic,
0.2
with most difficulties associated with the Russian stations
0
and at Alert and Denali in North America. Common to all
stations are poor results for NO3. The model predicts very
SO -S concentration, g/m3
4
well the higher NO3 concentrations at mid-latitudes, but
0.7
over predicts lower NO
0.6
3 concentrations at high latitudes
in the European Arctic.
0.5
0.4
0.3
3.7.2. Trend analysis based on
0.2
measurements at Station Nord
0.1
0
and DEHM model results
NO -N concentration, g/m3
3
Total sulfur (the sum of sulfur measured as sulfur dioxide
0.30
modeled
and sulfate) measured at Station Nord shows a strong sea-
measured
0.25
sonal variation with high values in the winter/spring and
0.20
very low values in the summer. In addition, the partition
0.15
of the two components changes during the winter/spring
0.10
season from a high content of sulfur dioxide during the
dark winter period to a very low content after polar sunrise
0.05
in early spring.
0
1992
1994
1996
1998
2000
Station Nord is virtually unaffected by local or regional
air pollution and can be considered a remote watchtower
Figure 3.23 Time series of measured and modeled monthly concentra-
tions of sulfur dioxide, sulfate, and nitrate at Zeppelin (Spitsbergen).
from where to follow the average emissions from a huge
emissions area in the eastern part of Europe and Russia.
Tustervann
Janiskoski
SO -S concentration, g/m3
2
SO -S concentration, g/m3
Zeppelin
2
1.0
modeled
Janiskoski
10
modeled
measured
Tustervann
0.8
measured
8
0.6
6
0.4
4
0.2
2
0
0
SO -S concentration, g/m3
4
SO -S concentration, g/m3
4
1.2
2.5
1.0
2.0
0.8
1.5
0.6
1.0
0.4
0.2
0.5
0
0
NO -N concentration, g/m3
NO -N concentration, g/m3
3
3
0.6
0.7
modeled
modeled
measured
measured
0.6
0.5
0.5
0.4
0.4
0.3
0.3
0.2
0.2
0.1
0.1
0
0
1992
1994
1996
1998
2000
1992
1994
1996
1998
2000
Figure 3.22. Time series of measured and modeled monthly concentra-
Figure 3.24. Time series of measured and modeled monthly concentra-
tions of sulfur dioxide, sulfate, and nitrate at Tustervann (Norway).
tions of sulfur dioxide, sulfate, and nitrate at Janiskoski (Russia).
29
Chapter 3 · Concentrations and Deposition of Acidifying Pollutants
A study of the general emissions trends is therefore pos-
SO -S (measured): SO -S (modeled)
x
x
sible using the long time series of sulfur measurements
1.4
made over the last decade. But, unfortunately, the meas-
1.2
ured concentrations depend strongly upon meteorological
conditions along the transport pathway to the Arctic. In
1.0
summer, emissions never get to the Arctic, whereas in win-
0.8
ter, transport times can change greatly from week to week.
Therefore, single measurements provide a poor reflection
0.6
of emissions in the source areas and a trend analysis on
the raw concentration data can give misleading results.
0.4
One way round this problem was to use the results of the
0.2
hemispheric Eulerian model DEHM.
The DEHM model with simple SO
0
2-SO4 chemistry
1990
1992
1994
1996
1998
2000
was run for the 11-year period from 1991 to 2001, with
emissions kept constant by using emissions data for 1990
only. All variations in modeled concentrations at Station
Figure 3.25. Ratios determined by regression analysis of measured
Nord were therefore due to daily changes in meteorologi-
values of sulfur (SOx-S) and constant emission model values for each
year between 1991 and 2000. The values are fitted by an exponential
cal conditions. Figure 3.25 shows the ratio SOx(measured):
function of time with a half-value period of 11±2 years.
SOx(modeled), calculated by weighted regression analysis
for each year between 1991 and 2000 (Wåhlin et al., 2002).
The values were fitted by an exponential function of time. It
is evident that this trial function does not disagree statisti-
between surface pressures of the subtropical highs at the
cally with the data points when the uncertainties (standard
Azores and the subpolar lows at Iceland. The AO is defined
deviations) are taken into account. The fitted function has
as the surface-level pressure anomalies for the North Pole.
an exponential decay of 11 years with an uncertainty of 2
As the NAO and AO are highly correlated only the NAO
years. If it is assumed that the changes in the ratio between
index was used in this assessment.
the measured and modeled concentrations of total SOX are
When the NAO is in a positive phase, negative low-pres-
due to changes in emissions only, these analyses indicate
sure anomalies over the Icelandic region and throughout
that the emissions, which contribute to the measured con-
the Arctic, together with positive high-pressure anomalies
centrations at Station Nord, decreased by a factor of two
across the subtropical Atlantic, tend to produce stronger
between 1991 and 2000.
westerly winds at mid-latitudes. The negative low-pres-
sure anomalies throughout the Arctic trigger transport epi-
sodes from the Russian Arctic into the atmosphere over
3.7.3. Effects of natural climate variations on
the Arctic Ocean.
long-range transport to the Arctic
For the European Arctic the total mean column level of
SO2+SO4 for January to April is negatively correlated with
Climate change may alter the atmospheric transport of
the NAO index (Figure 3.27), which means that during
contaminants to and within the Arctic. One way to investi-
periods with a negative NAO index there is higher sulfur
gate this is to study the influence of present natural climate
transport from Europe to the Arctic. For the Canadian-
variations (such as the North Atlantic Oscillation (NAO)
USA-eastern Siberian Arctic the total mean column level of
and/or the Arctic Oscillation (AO)) on transport of con-
SO2+SO4 for January to April is positively correlated with
taminants to and within the Arctic using an atmospheric
the NAO index (Figure 3.27), which means that during
transport model. The DEHM model has been used to study
periods with a positive NAO index there is more sulfur
natural climate variations related to the AO and NAO by
transport from eastern Russia into the Arctic and more
carrying out 24 years of model calculations.
transport from North America towards the area between
A large part of the climate variability over the northern
Canada and Greenland.
hemisphere is associated with variability in the NAO index
Climate models project that large positive NAO events
(Figure 3.26). The NAO index is defined as the difference
will occur more frequently in future, and that the rest of the
NAO index
3
2
1
0
-1
-2
-3
Figure 3.26. NAO index for
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
the period 1979 to 2002.
30
Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Mean SO -S concentration, g/m3
Total deposition, kt
x
0.8
4000
0.7
3500
0.6
3000
0.5
2500
0.4
2000
0.3
1500
0.2
1000
0.1
500
0
0
Mean NH -N concentration, g/m3
X
0.07
140
0.06
120
0.05
100
0.04
80
0.03
60
0.02
40
mg S/m2
0.01
20
0.5
1.0
2.0
4.0
6.0
8.0
10.0
0
0
Figure 3.27. The total mean column level of SO2+SO4 for the period
Mean NO -N concentration, g/m3
January to April for 1979 to 2002, with contour lines showing the cor-
Y
0.18
900
relation between the different monthly mean total column levels and
the NAO index.
0.16
800
0.14
700
0.12
600
NAO frequency distribution will change little (see Coppola
0.10
500
et al., 2005). This would result in less transport from Euro-
0.08
400
pean sources and more transport from sources in eastern
0.06
300
Russia and North America, resulting in lower concentra-
0.04
200
tions in the European Arctic and higher concentrations in
0.02
100
the Russian and North American Arctic.
0
0
1990
2000
2010
2020
concentration CLE
3.7.4. Scenarios
concentration MFR
deposition CLE
deposition MFR
The DEHM model with extensive chemistry was run for
six emissions scenarios. These emissions scenarios are for
Figure 3.28. Total mean concentrations and total depositions of sulfur
oxides, reduced nitrogen (NH
the years 1990, 2000, 2010 and 2020, with 2010 and 2020
X) and airborne oxidized nitrogen (NOy)
for the area north of Arctic Circle for six emissions scenarios.
having two different emissions scenarios: the `Maximum
technically Feasible Reduction' (MFR) scenario and the
`Current LEgislation' (CLE) scenario (see section 2.3 for
that even though the total emissions in the MFR scenario
further details). For each emissions scenario the DEHM
are a factor of 2 less than those in the CLE scenario this
model was run using the same meteorological input for
has only a minor effect on the total concentrations and
the period 1991 to 1993 in order to reduce the effects of
deposition of SOx and NOx in the Arctic. Since most of
meteorological variations on the model results.
the difference between the two scenarios is in the Asian
The model results (see Figure 3.28) confirm the de-
emissions, this indicates that future emissions in Asia are
crease in SOx concentrations calculated for Station Nord
likely to have only a small impact on acidification in the
(discussed in section 3.7.2) and estimate that both the mean
Arctic. Pollutants emitted in warmer regions such as Asia
SOX concentrations and the total sulfur deposition almost
are not transported directly to the cold Arctic (see section
halved between 1990 and 2000. The results for NOy are
2.2). It is emissions in Eurasia that contribute most to aci-
similar to those for SOx.
dification in the Arctic and so it is future changes in these
The outcome projected for the two future emissions
emissions that are likely to have the greatest impact. These
scenarios (CLE and MFR) is a small decrease in both con-
findings confirm the results in the previous AMAP assess-
centrations and deposition. Another important point is
ment (AMAP, 1998).
31
Chapter 4
Arctic Haze
Patricia Quinn, Betsy Andrews, Ellsworth Dutton, Glenn Shaw, and Tuija Ruoho-Airola
4.1. The arctic haze phenomenon
It has been more than 50 years since observations of a
strange haze, of unknown origin, were reported by pilots
flying in the Canadian and Alaskan Arctic (Greenaway,
1950; Mitchell, 1956). Based on measurements at McCall
Glacier in Alaska, Shaw and Wendler (1972) noted that
the turbidity maximized in spring. First measurements of
5 - 15 %
the vertical structure of the haze were made in an Alaskan
`bush' airplane with a hand-held sunphotometer (Shaw,
1975). At that time the origin of the haze was uncertain
and was attributed to ice crystals seeded by open leads or
blowing dust from riverbeds. It was only through `chemi-
Arctic
cal fingerprinting' of the haze that its anthropogenic source
10 - 40 %
5 - 25 %
Ocean
was revealed (Ottar et al., 1986; Rahn et al., 1977; Rahn and
McCaffrey, 1979; Rahn, 1989). By the late 1970s the anthro-
pogenic origin was clear but surprising since it was widely
North
North
believed that aerosol was generally not transported more
Pacific
Atlantic
than a few hundred kilometers from its source regions.
Ocean
Ocean
Experts from Europe and America convened at the first
Arctic Air Chemistry Symposium at Lillestrom, Norway
in 1978 and an informal measurement network was agreed
upon. Spatial gradients soon showed the direction of flow
and the surprisingly large extent of this anthropogenic
cloud of pollution. A combination of intensive field pro-
Arctic Front, Summer
Major south-to-north air transport
routes into the Arctic and the
grams and long-term measurements extending over the
percentage frequency of the winds in
past thirty years confirmed the early conclusions that the
Arctic Front, Winter
summer (orange) and winter (blue).
haze is anthropogenic in origin due to emissions from Eu-
rope and the former Soviet Union that are transported to
Figure 4.1. Mean position of the arctic air mass in winter (January) and
and trapped in the arctic air mass during the winter and
summer (July), superimposed on the percentage frequency of major
south-to-north transport routes into the Arctic in summer and winter
early spring (Figure 4.1).
(AMAP, 1998).
The haze comprises a varying mixture of sulfate and
particulate organic matter and, to a lesser extent, ammo-
nium, nitrate, dust, and black carbon (e.g., Li and Barrie,
1993; Quinn et al., 2002). It is also rich in certain heavy met-
als which has allowed for the identification of particular
inhibits turbulent transfer between atmospheric layers as
industrial sources (e.g., Shaw, 1983; Rahn, 1989). Particles
well as the formation of cloud systems and precipitation;
within the haze are well-aged with a mass median diam-
the major removal pathway for particulates from the at-
eter of about 0.2 m or less (e.g., Heintzenberg, 1980; Hoff
mosphere (Barrie et al., 1981; Shaw, 1981, 1995; Heintzen-
et al., 1983; Pacyna et al., 1984; Shaw, 1984; Clarke, 1989;
berg and Larssen, 1983). In addition, meridional transport
Leaitch et al., 1989; Trivett et al., 1989; Hillamo et al., 1993).
from the mid-latitudes to the Arctic intensifies during the
This particle size range is very efficient at scattering visible
winter and spring (Iversen and Joranger, 1985). The combi-
solar radiation since the peak in the particle surface-area
nation of these factors results in the transport of precursor
size distribution is near the maximum efficiency for Mie
gases and particulates to the Arctic and the trapping of the
scattering (Waggoner and Weiss, 1980; Shaw, 1987). The
pollutant haze for up to 15 to 30 days (Shaw, 1981, 1995).
haze is also weakly absorbing due to the presence of black
Aircraft and lidar measurements throughout the 1980s
carbon (e.g., Hansen and Rosen, 1984; Noone and Clarke,
and 1990s revealed that the haze occurs primarily in the
1988; Kahl and Hansen, 1989; Hopper et al., 1994). The
lowest five kilometers of the atmosphere and peaks in the
result of the strong scattering and weaker absorption is a
lowest two kilometers (Leaitch et al., 1984; Hoff, 1988; Pa-
noticeable reduction in visibility to a few kilometers or less.
cyna and Ottar, 1988; Barrie, 1996). Throughout the haze
The `weak' absorption may have large climatic influences
season, the pollution layers are highly inhomogeneous
when the dark colored haze spreads out over the highly re-
both vertically (tens of meters to 1 km thick) and spatially
flecting snow and ice pack of the Arctic. The highly reflect-
(20 to 200 km in horizontal extent) (Radke et al., 1984; Brock
ing surface enhances aerosol-radiative interactions due to
et al., 1989).
multiple scattering between the surface and the haze.
Recent aircraft measurements of sulfate aerosol using a
Several seasonally-dependent mechanisms contribute
high time resolution technique revealed detailed informa-
to the formation of arctic haze. Strong surface-based tem-
tion about the evolution of the vertical structure of the haze
perature inversions form in the polar night causing the
between February and May (Scheuer et al., 2003). During
atmosphere to stabilize. This cold and stable atmosphere
early February, significant enhancements in sulfate aero-
32
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
sol are confined near the surface (< 2 km) as long-range
pit and 75 m ice core in Greenland indicated that there is
transport from northern Eurasia occurs along low level,
a maximum in sulfate and black carbon deposition to the
sinking isentropes (Klonecki et al., 2003). As the haze sea-
surface during the arctic haze season although there can
son progresses, enhanced sulfate occurs at higher altitudes
be significant deposits throughout the year (Masclet et al.,
(up to at least 8 km). Since vertical mixing is prohibited by
2000). Similarly, sulfate concentrations as well as concen-
the persistent low-level inversion (Kahl, 1990), the higher
trations of other arctic haze tracers (lead, cadmium, and
altitude haze layers are thought to be due to transport into
arsenic) in snow at sixteen sites across northwest Alaska
the Arctic along vertically higher isentropes tracing back
were found to be highest in the later winter snow pack
to increasingly warmer source regions in northern Eurasia.
(Douglas and Sturm, 2004). The aerosols are removed
During early April, sulfate layers below 3 km begin to dis-
from the atmosphere either through dry deposition or
sipate due to the beginning of solar heating and resulting
wet deposition with the mechanism of the latter most
mixing near the surface. However, more stable isentropic
likely involving nucleation followed by precipitation in
transport continues at higher altitudes. By the end of May,
the form of ice crystals (Masclet et al., 2000). Since the tim-
both the lower and higher altitude sulfate enhancements
ing of the buildup of the haze and the snow pack is similar
are significantly decreased due to the continued break-up
and haze concentrations decline before the snow has fully
of the inversion and the return of wet deposition.
melted, it is likely that haze contaminants first enter the
Recent studies have provided evidence for an influence
ecosystem by being deposited in the snow (Douglas and
of natural climate variability on interannual changes in
Sturm, 2004). Contaminants deposited in the snow then
levels of arctic haze. Modeling the dispersion of anthro-
end up in the tundra and rivers as the snow melts. The
pogenic emissions from northern hemisphere continents,
network of snow measurements across northwest Alaska
Eckhardt et al. (2003) found that the North Atlantic Oscil-
showed spatially homogeneous concentrations of sulfate
lation (NAO) influences pollution transport into the Arc-
and trace elements suggesting little variability in atmos-
tic during the winter-spring haze season (see also section
pheric concentration or scavenging efficiency within the
3.7.3). During positive phases of the NAO, surface concen-
confines of the sampled region. By contrast, the acidity of
trations of modeled tracers in the arctic winter were found
the snow was much patchier indicating that competing
to be elevated by about 70% relative to negative phases.
acidifying and buffering sources determine the local pH.
This difference was mainly due to a change in pathways
For example, pH was consistently higher in the Brooks
of European pollution and, to a lesser extent, North Ameri-
Range than elsewhere due to mechanical weathering of
can pollution to the Arctic both of which are enhanced
carbonate rocks.
during positive NAO phases. In addition, during posi-
tive NAO phases, significant positive correlations between
the NAO and measured carbon monoxide concentrations
were found at three arctic monitoring stations (Spitsbergen,
4.2. Trends in arctic haze
Barrow, and Alert) confirming enhanced poleward trans-
4.2.1. Chemical composition
port of pollution from Europe, Asia, and North America.
Similar but weaker correlations between the NAO and
Arctic haze is marked by a dramatic increase in concentra-
measured carbon monoxide concentrations were found
tions of several key particulate pollutants during winter
for spring. Low correlations were found during summer
and early spring. The seasonal trend in the haze has been
and autumn.
detected at several monitoring sites in the Arctic includ-
During transport from the source regions to the Arctic,
ing Alert (82.46° N) in the Canadian Arctic, Station Nord
the pollutant-containing air masses have a high probability
in Greenland (81.4° N), Zeppelin (79° N) on the island of
of reaching saturation and nucleating and precipitating
Svalbard, Barrow in Alaska (71.3° N), Karasjok (69.5° N)
clouds. It is not understood how so much material gets
and Svanvik (69.45° N) in northern Norway, Oulanka (66.3°
through a strongly scavenging system (Bowling and Shaw,
N) in northern Finland, and Janiskoski (69° N) in western
1992).
Russia (see Figure 3.3). A time series for particulate sulfate
Arctic haze has been the subject of much study because
concentrations is shown for these eight monitoring sites
of its potential to change the short and longwave radiation
in Figure 4.2. Each site has a similar winter/early spring
balance of the Arctic, to affect visibility, and to provide a
increase in sulfate with maximum concentrations reaching
source of contaminants to arctic ecosystems. The near sur-
up to about 1 g S/m3. Monthly average concentrations in
face concentration of aerosols at most places in the Arctic
summer are less than 0.03 g S/m3. Non-marine sulfate
are about an order of magnitude lower than those found
(SO4*) makes up about 30% of the submicron mass dur-
at more polluted and industrialized locations. At the same
ing the haze season (Barrie et al., 1981; Quinn et al., 2000,
time, however, the affected areas are much larger in size
2002). Figure 4.3 shows the time series for particulate ni-
and the affected ecosystems in the high Arctic are thought
trate concentrations at Alert and Barrow, two sites which
to be quite sensitive to gaseous and aerosol contamina-
have a clear seasonal pattern for this species. Maximum
tion.
concentrations are near 0.04 g N/m3. Other species also
It is not known what fraction of the arctic haze con-
indicative of continental sources (ammonium and non-
taminants leave the Arctic and what fraction is deposited
marine potassium for biomass burning and magnesium
within the Arctic on land and sea surfaces. As the po-
and calcium for dust) have maximum concentrations in
lar night ends, some of the pollution that has accumu-
winter and spring indicating long range transport to the
lated is released to the mid-latitudes (Penkett et al., 1993;
Arctic (Quinn et al., 2002).
Heintzenberg et al., 2002). It is known that haze contami-
Natural aerosol chemical components display seasonal
nants (e.g., acidic sulfate and organics) end up in arctic
cycles quite different from anthropogenic components. Sea
ecosystems (Meijer et al., 2003; Wania, 2003) but the timing
salt concentrations are highest at Alert and Barrow from
and mechanism of the scavenging from the atmosphere
November through February (Quinn et al., 2002). The win-
is not well understood. Measurements from a 5 m snow
ter maximum has been attributed to seasonally high winds
33
Chapter 4 · Arctic Haze
Arctic
Subarctic
SO -S concentration in air, g/m3
4
Barrow (B)
Karasjok (K)
1.5
2.5
2.0
1.0
1.5
1.0
0.5
0.5
0
0
Alert (A)
Svanvik (Sv)
1.5
1.5
1.0
1.0
0.5
0.5
0
0
Nord (N)
Janiskoski (J)
1.5
1.5
1.0
1.0
0.5
0.5
0
0
Zeppelin (Z)
Oulanka (O)
1.5
1.5
(Ny-Ålesund)
1.0
1.0
0.5
0.5
0
0
1/82 1/84 1/86 1/88 1/90 1/92 1/94 1/96 1/98 1/00 1/02 1/04
1/80 1/82 1/84 1/86 1/88 1/90 1/92 1/94 1/96 1/98 1/00 1/02
Month/year
Figure 4.2. Time series of monthly averaged particulate sulfate concentrations at eight arctic monitoring sites. Data made available for Alert by
the Canadian National Atmospheric Chemistry (NAtChem) Database and Analysis System, for Barrow by NOAA PMEL (http://saga.pmel.noaa.
gov/data/), and for the other stations by EMEP (http://www.emep.int/).
SO *-S in air, g/m3
NO -N, g/m3
4
3
B
1.0
0.06
Alert
SO *
0.8
4
A
NO
N
3
0.04
0.6
Z
J
K Sv
O
0.4
0.02
0.2
0
0
1/82
1/84
1/86
1/88
1/90
1/92
1/94
1/96
1/98
1/00
1/02
0.4
0.04
Barrow
0.3
0.03
0.2
0.02
Figure 4.3. Time series of monthly
averaged particulate sulfate and
0.1
0.01
nitrate concentrations at Alert
(Canada) and Barrow (Alaska).
0
0
Data sources as for Figure 4.2.
1/98
1/99
1/00
1/01
1/02
1/03
1/04
Month/year
in high-latitude source regions of the Pacific and Atlantic
increase in late June as the sea ice recedes and phytoplank-
Oceans and long-range transport to the Arctic (Sturges
ton productivity in surface waters begins. This is also when
and Barrie, 1988; Sirois and Barrie, 1999). Supermicron sea
DMS that has been trapped under the ice is released (Ferek
salt aerosol peaks during the summer months at Barrow
et al., 1995).
when the ice pack extent is at a minimum (Quinn et al.,
Two years of measurements at a site in northern Fin-
2002). Atmospheric methanesulfonic acid (MSA-) is de-
land revealed that particulate organic matter made up,
rived solely from the oxidation of biogenically produced
on average, 22% of the total fine aerosol mass (range 3 to
dimethylsulfide (DMS). Concentrations of MSA- begin to
69%) (Ricard et al., 2002). Correlation of particulate organic
34
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
matter with sulfate* was low indicating that they had dif-
the trend, slope estimate, and percentage change over the
ferent sources. Since particulate organic matter displayed
measurement period are given in Table 4.1. For sites with a
a seasonal cycle with maximum concentrations in summer,
significant trend (based on the Mann-Kendall test), sulfate
it is most likely to result from biogenic emissions and/or
concentrations decreased by 30 to 60% between 1990 and
enhanced oxidation processes. A small increase in organic
present. The decreasing trend in sulfate at Alert detected
acids as early as February/March may indicate photooxi-
through the 1990s has continued into the present century.
dation at polar sunrise, as was pointed out for Alert by
In addition, sulfate has decreased significantly at Zeppelin,
Kawamura et al. (1996).
Karasjok, and Oulanka.
The longest record of sulfate concentrations in the Arc-
Corresponding data are shown for nitrate in Figure
tic (1980 to present at Alert, Canada) revealed no change in
4.5 and Table 4.2. In contrast to sulfate, nitrate increased
sulfate concentrations during the 1980s (Sirois and Barrie,
during April at Alert for the long-term period of 1981 to
1999). This lack of a trend was attributed to little change
2004 and during March for the shorter term more recent
in emissions in the former Soviet Union between 1985 and
period of 1990 to 2004. Between 1990 and 2004, nitrate con-
1990. Beginning in 1991, sulfate and other measured an-
centrations increased by about 50%. More measurements
thropogenic constituents (lead, zinc, copper, excess vana-
are needed at both Alert and Barrow to verify these trends,
dium and manganese, and ammonium) began to decline
however.
suggesting that reduced industrial activity in the early
years of the new Eurasian republics had led to lower levels
of pollutants reaching the Arctic.
4.2.2. Optical properties
A combined modeling and measurement analysis of
sulfate concentrations at Station Nord in northern Green-
The seasonality and trends in arctic haze are clearly seen
land indicated a decreasing trend throughout the 1990s
in time series data of light absorption and scattering by
(Heidam et al., 2004). The analysis was able to account for
scatter in measured concentrations due to changing me-
Table 4.1. Change in monthly average sulfate concentrations at a range
teorology. With the meteorological variability removed, it
of arctic monitoring stations for March and April. Significance of the
was possible to attribute the decrease in concentrations to a
trend, Sen's slope estimate, and percentage change per period are
reduction in emissions. The model that was used estimated
listed.
that more than 70% of the sulfur measured at Station Nord
a
Slopeb Percentage
change
was emitted from the area making up the former Soviet
over period
Union which indicated that emissions from the region de-
Alert
SO4*
March 19812002 0.01
-0.066
- 66
creased significantly during the 1990s. This result is sup-
April 19812003
0.001
-0.079
- 71
ported by the 50% decrease in Russian sulfur emissions
March 19902002 0.001
-0.088
- 59
reported to EMEP during the 1990s (Vestreng, 2003). It is
April 19902003
0.01
-0.086
- 63
not clear from this analysis how reductions from Western
Nord
SO4
March 19942002
Europe and North America influenced sulfate concentra-
April 19942002
Zeppelin
SO
tions at Station Nord.
4
March 19902003 0.1
-0.094
-33
April 19902003
0.1
-0.079
-27
Sulfate concentrations also decreased during the 1990s
Barrow
SO4*
March 19982004
at several other sites in the Arctic. Data for updating trend
April 19982004
analyses through the first few years of the 21st century are
Karasjok
SO4
March 19782003 0.001
-0.044
- 80
now available at many arctic sites.
April 19782003
0.001
-0.021
- 59
March 19902003 0.1
-0.017
- 40
As pointed out by Macdonald et al. (2005), the detection
April 19902003
0.01
-0.025
- 48
of recent trends in the Arctic is difficult due to the combina-
Svanvik
SO4
March 19932000
tion of short or incomplete data records at some sites and
April 19932000
interference from natural variations on seasonal, annual,
Janiskoski SO4
March 19912002
and decadal time scales. To remove seasonal variability
April 19912002
from the trend analyses, this assessment has focused on
Oulanka
SO4
March 19902002 0.1
-0.034
- 45
April 19902002
0.01
-0.043
- 56
average monthly concentrations for March and April. The
a
existence of a monotonic increasing or decreasing trend in
Significance level, , of the Mann-Kendall test. A significance level
of 0.001 indicates a 0.1% probability of no trend. No value indicates a
the time series was tested with the nonparametric Mann-
significance level of >0.1; b Sen's nonparametric method was used to
Kendall test at significance levels p<0.001, p<0.01, and
estimate the slope of the existing trend as change per year (g m3/yr).
p<0.1 as a two-tailed test (Gilbert, 1987). The estimate for
the slope of a linear trend was calculated with the non-
Table 4.2. Change in monthly average nitrate concentrations at Alert
parametric Sen's method (Sen, 1968a). The Mann-Kendall
(Canada) and Barrow (Alaska) for March and April. Significance of
test is suitable for cases where the trend may be assumed
the trend, Sen's slope estimate, and percentage change per period
are listed.
to be monotonic such that no seasonal or other cycle is
a
present in the data. In the Mann-Kendall test, missing data
Slopeb
Percentage
change
values are allowed and the data need not conform to any
over period
particular distribution. The Sen's slope is the median of the
Alert
NO3
March 19812000
slopes calculated from all pairs of values in the data series.
April 19812000
0.1 0.0006
67
The Sen's method is not greatly affected by data outliers
March 19902000
0.05 0.0008
50
and can be used when data are missing (Salmi et al., 2002).
April 19902000
Trends are only reported for a significance level of <0.1, i.e.,
Barrow
NO3
March 19982004
April 19982004
when the probability of no trend is 10% or less.
Average monthly concentrations of sulfate for March
a Significance level, , of the Mann-Kendall test. A significance level
of 0.001 indicates a 0.1% probability of no trend. No value indicates a
and April are shown in Figure 4.4 along with the Sen's slope
significance level of >0.1; b Sen's nonparametric method was used to
estimates for a significance level of <0.1. The significance of
estimate the slope of the existing trend as change per year (g m3/yr).
35
Chapter 4 · Arctic Haze
SO -S concentration in air, g/m3
4
Barrow (B)
Karasjok (K)
0.4
3. 0
0. 3
2. 0
0. 2
1. 0
0. 1
0
0
Alert (A)
Svanvik (Sv)
1. 5
1. 2
1.0
0. 8
0. 5
0. 4
0
0
0. 8
Nord (N)
Janiskoski (J)
1. 5
0. 6
1. 0
0.4
0. 2
0. 5
0
0
Zeppelin (Z)
Oulanka (O)
0. 6
1.2
(Ny-Ålesund)
0.4
0.8
April SO4
March SO4
Long-term trend March
0. 2
Long-term trend April
0.4
Short-term trend March
Short-term trend April
0
0
1975
1980
1985
1990
1995
2000
2005
1975
1980
1985
1990
1995
2000
2005
Figure 4.4. Monthly averaged sulfate concentrations for March and April at eight arctic monitoring sites and Sen's slope estimates for the long
term (approximately 1980 through the available data) and short term (approximately 1990 through the available data) trends. Trend lines are not
shown for > 0.1. Data sources as for Figure 4.2.
B
NO -N concentration in air, g/m3
3
Alert (A)
0.04
A
N
0.03
Z
J
K Sv
O
0.02
0.01
0
Barrow (B)
0.04
0.03
0.02
April NO3
March NO
0.01
3
Figure 4.5. Monthly averaged nitrate concentrations for March and
Long-term trend April
April at Alert (Canada) and Barrow (Alaska) and Sen's slope estimates
Short-term trend March
for the long term (approximately1980 through the available data) and
0
short term (approximately 1990 through the available data) trends.
1980
1985
1990
1995
2000
2005
Trend lines are not shown for > 0.1. Data sources as for Figure 4.2.
36
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Scattering, 1/Mm
30
Barrow
20
10
0
Absorption, 1/Mm
2.5
Barrow
2.0
1.5
1.0
0.5
0
Black carbon, ng/m3
250
Alert
200
150
100
50
1/77
1/79
1/81
1/83
1/85
1/87
1/89
1/91
1/93
1/95
1/97
1/99
1/01
1/03
Month/year
Figure 4.6. Monthly averaged light scattering and absorption at 550 nm by <10 m aerosol at Barrow (Alaska) and black carbon concentrations at
Alert (Canada). Data sources as for Figure 4.2.
aerosols measured at the surface (Figure 4.6) (Bodhaine,
measurement period of 1977 to 2005. Breaking the period
1989) and in total column aerosol optical depth (Dutton et
into two smaller intervals reveals a decreasing trend from
al., 1984; Herber et al., 2002). Bodhaine and Dutton (1993)
1982 to 1996 which reverses to yield an increasing trend
reported that both aerosol scattering and optical depth
from 1997 to 2005. The increasing trend for March is signifi-
measurements at Barrow peaked in 1982 followed by a
cant at the 0.05 level indicating that there is a probability
factor of two decrease between 1982 and 1992. The decrease
of 95% that the trend exists.
was only apparent during March and April corresponding
A similar analysis for light absorption at Barrow in-
to the time of year when arctic haze is most pronounced.
dicates an overall significant decrease between 1988 and
It was hypothesized that the decrease in the haze was
2005 for March. Trends between 1997 and 2005 are unclear
most likely to be due to a combination of a reduction in
(Figures 4.7 and 4.8; Table 4.3). Sharma et al. (2006) reported
the output of pollution aerosol by Europe and the former
an increasing trend in absorption for the winter months
Soviet Union and stricter pollution controls in Western
(defined as January through April) at Barrow between 2000
Europe. The decreases in aerosol scattering and optical
and 2003.
depth at Barrow during this ten-year period are not equal
Black carbon concentrations measured at Alert show a
to the known reductions in sulfate emissions, however,
decreasing trend for both March and April over the meas-
indicating that other factors such as changes in transport
urement period of 1990 through 2001 (Figures 4.7 and 4.8;
processes could have played a role (Jaffe et al., 1995). Since
Table 4.3). Sharma et al. (2004) reported a decrease in black
the NAO was in a positive phase between the early 1980s
carbon (the main light absorber in the arctic atmosphere)
and 1990s, implying enhanced transport of pollutants to
of 56% for the winter/spring season from 1989 to 2002 at
the Arctic (Hurrell and van Loon, 1997), it appears not to
Alert. However, Sharma et al. (2006) reported an increase
have been responsible for the observed decrease in arctic
during winter (January to April) beginning in 2000.
haze during this period.
A recent study using a general circulation model sug-
An update of the monthly averaged light scattering
gested that one of the major sources of arctic soot today
data analysis originally performed by Bodhaine and Dut-
is southern Asia (Koch and Hansen, 2005) due to increas-
ton (1993) is shown in Figures 4.7 and 4.8 with data at
ing emissions from industrial and biofuel combustion.
Barrow extending through 2005. Also shown are Sen's
This suggestion has been refuted by Stohl (2006), how-
slope estimates for three periods: 19772005, 19821996,
ever, based on the long passage from pollution sources in
and 19972005. The significance of the detected trends,
southern and eastern Asia to the Arctic relative to more
slope estimates, and percentage change in scattering for
rapid transport from Europe and northern Asia. Using a
these periods are given in Table 4.3. For March, there is a
particle dispersion model, it was found that, for a trans-
significant decreasing trend in scattering over the entire
port time of five days, the southern Asia black carbon
37
Chapter 4 · Arctic Haze
contribution is about 1 to 2% of the European source
Table 4.3. Change in monthly average light scattering and absorption
contribution near the surface and 3 to 5% for the total
at Barrow (Alaska) and black carbon at Alert (Canada) for March and
air column.
April. Significance of the trend, Sen's slope estimate, and percentage
change during the period are listed.
Results reported by Stohl (2006) identify boreal and
a
temperate forest fires, especially Siberian fires, as a sig-
Slopeb
Change
during the
nificant source of black carbon during the summer. In in-
period %
tense fire years, boreal forest fires may be the dominant
Barrow
c
sp
March 19772005
0.01
-2.4 E-07
- 41
source of black carbon for the Arctic. Measurements at
April 19772005
Alert and Barrow show there is a strong seasonal cycle in
March 19821996
0.01
-5.7 E-07
- 42
potassium*, a tracer for biomass burning, with minimum
April 19821996
0.01
-6.6 E-7
- 56
values in the summer and maximum values in the winter
March 19972005
0.05
7.8 E-07
46
April 19972005
(e.g., Quinn et al., 2000). Based on these data, the impact
of summertime forest fire emissions on low altitude sur-
Barrow
c
ap
March 19882005
0.1
-3.9 E-08
-61
face sites within the Arctic is relatively small compared to
April 19882005
winter emissions. There is evidence, however, of pyrocu-
March 19881996
April 19881996
mulonimbus injection of smoke from boreal forest fires to
March 19972005
the upper atmosphere (Fromm et al., 2005). In addition,
April 19972005
biomass burning signatures have been observed in the
snow at the high altitude (3200 m) site of Summit, Green-
Alert
BCd
March 19902001
0.001 -17
- 88
land (Dibb et al., 1996). The fraction of this material that
April 19902001
0.05
-8.3
-63
March 19972001
0.1
-3.9
-20
is deposited to lower elevations throughout the Arctic is
April 19972001
unknown. More measurements coupled with modeling
a Significance level, , of the Mann-Kendall test. A significance level
studies are required to identify sources of black carbon to
of 0.001 indicates a 0.1% probability of no trend. No value indicates
the Arctic and to assess trends in black carbon and light
a significance level of > 0.1; b Sen's nonparametric method was
absorption by aerosols.
used to estimate the slope of the existing trend as change per year
An extension of the Barrow aerosol optical depth (AOD)
(mm/yr for scattering and absorption, ng m3/yr for black carbon);
c <10 m aerosol; d BC: black carbon.
data through 2002 shows a continued decrease through the
mid-1990s (Figure 4.9). Monthly averaged values of AOD
anomalies (relative to a base of non-volcanic years) for
March show a continued decline through 2002. However,
the AOD anomalies for April indicate an increase between
Scattering, 1/Mm
Scattering, 1/Mm
30
30
Barrow
Barrow
March averages
April averages
20
20
1982-1996 Trend ( =0.01)
1982-1996 Trend ( =0.01)
1977-2005 Trend ( =0.01)
10
10
1997-2005 Trend ( =0.05)
0
0
Absorption, 1/Mm
Absorption, 1/Mm
2.5
1.0
Barrow
Barrow
2.0
0.8
1.5
0.6
1988-2005 Trend ( =0.1)
1.0
0.4
0.5
0.2
0
0
Black carbon, ng/m3
Black carbon, ng/m3
300
300
Alert
Alert
250
250
200
200
150
1990-2001 Trend ( =0.05)
150
1990-2001 Trend ( =0.001)
100
100
1997-2001 Trend ( =0.1)
50
50
0
0
1975
1980
1985
1990
1995
2000
2005
1975
1980
1985
1990
1995
2000
2005
Figure 4.7. Monthly averaged concentrations for March of light scat-
Figure 4.8. Monthly averaged concentrations for April of light scat-
tering and light absorption at 550 nm for <10 m aerosol at Barrow
tering and light absorption at 550 nm for <10 m aerosol at Barrow
(Alaska) and black carbon concentrations at Alert (Canada). Lines
(Alaska) and black carbon concentrations at Alert (Canada). Lines
indicate the Sen's slope estimate for the periods indicated and the
indicate the Sen's slope estimate for the periods indicated and the
value indicates the significance level of the trend ( = 0.001 indicates
value indicates the significance level of the trend ( = 0.001 indicates
there is a 0.1% probability that the trend does not exist). Trend lines
there is a 0.1% probability that the trend does not exist). Trend lines are
are not shown for >0.1. Data sources as for Figure 4.2.
not shown for >0.1. Data sources as for Figure 4.2.
38
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
Aerosol optical depth anomaly
0.20
March
0.15
0.10
0.05
0
-0.05
-0.10
0.20
April
0.15
0.10
0.05
0
-0.05
-0.10
1/78
1/80
1/82
1/84
1/86
1/88
1/90
1/92
1/94
1/96
1/98
1/00
1/02
Month/year
Figure 4.9. Monthly averaged aerosol optical depth anomalies at Barrow (Alaska) for March and April. The anomalies are relative to a base of
non-volcanic years. Data from 1992 and 1993 were removed due to stratospheric aerosol influx from the Pinatubo eruption in 1991. Vertical lines
represent 1 standard deviation of the monthly mean (data provided by NOAA CMDL).
1998 and 2001 where the currently available data record
0.2 K/day were measured by Valero et al. (1989) during
ends. In contrast to the Barrow trend through the 1990s,
AGASP (Arctic Gas and Aerosol Sampling Program) II
Herber et al. (2002) reported a slightly increasing trend in
and by Treffeisen et al. (2004) during the ASTAR (Arctic
AOD (1% per year) at Koldewey station in Ny-Ålesund,
Study of Tropospheric Aerosols, Clouds, and Radiation)
Spitsbergen, between 1991 and 1999.
2000 campaign in Svalbard. The AASE (Airborne Arctic
Stratospheric Expedition) II flights in winter 1992 revealed
soot-contaminated arctic aerosols at altitudes of 1.5 km.
Pueschel and Kinne (1995) calculated that this layer of
4.3. Effects of aerosol on the climate
aerosols could heat the earthatmosphere system above
system in the Arctic
surfaces of high solar albedo (ice/snow) even for single
4.3.1. Direct effects
scattering albedos as high as 0.98. Hence, a modest amount
of black carbon in the haze layers can result in a measur-
The direct effects of aerosols on the radiation balance in
able contribution to diabatic heating.
the Arctic are due to the absorption and scattering of ra-
MacCracken et al. (1986) estimated that the cooling of
diation by the aerosol. The Arctic is thought to be par-
the surface due to absorption of solar radiation by the haze
ticularly sensitive to changes in radiative fluxes imposed
layers would be compensated by infrared emission from
by aerosols because of the small amount of solar energy
the atmosphere to the surface. The infrared emission is
normally absorbed in polar regions (Valero et al., 1989).
expected to dominate during the polar night when long-
Arctic haze is present as a layer of light-absorbing mate-
wave radiation controls the energy budget of the Arctic. If
rial over a highly reflective ice/snow surface. Shaw and
the haze particles deliquesce thereby taking up water and
Stamnes (1980) first realized that the absorbing nature of
growing to cloud droplet or ice crystal size, their longwave
arctic haze would have a significant impact on the energy
impact will be enhanced. Measurements made on Svalbard
balance of the Arctic. Several early calculations using 1-
when the sun was below the horizon indicate that arctic
D radiative transfer models estimated that the diurnally
haze can have a measurable direct thermal radiative forc-
averaged atmospheric warming due to the layer ranged
ing altering the flux of downward longwave radiation by
between 2 and 20 W/m2, with a corresponding depletion
up to +3 to +4.7 W/m2 and the outgoing longwave radia-
of the solar flux at the surface of 0.2 to 6 W/m2 (Porch and
tion by -0.23 to +1.17 W/m2 (Ritter et al., 2005).
MacCracken, 1982; Leighton, 1983; Blanchet and List, 1987;
The vertical distribution of the absorbing haze layers
Emery et al., 1992; Shaw et al., 1993). These estimates agreed
does not affect the radiation budget at the top and bottom
with direct measurements from wideband sun photom-
of the atmosphere (Cess, 1983) but may affect atmospheric
eters (Mendonca et al., 1981). Heating rates of about 0.1 to
circulation and climate feedback processes.
39
Chapter 4 · Arctic Haze
4.3.2. Indirect effects
aerosols (Blanchet and Girard, 1995). Measurements cor-
roborate this finding. Borys (1989) reported that arctic haze
The indirect effects of aerosols on irradiances in the Arctic
aerosol had lower ice nuclei concentrations, a lower ice nu-
result from the impact of aerosol particles on the micro-
clei to total aerosol fraction, and slower ice nucleation rates
physical properties of clouds. Enhanced aerosol particle
than aerosol from the remote unpolluted troposphere. The
concentrations increase solar cloud albedo by increasing
reduction in ice nuclei leads to a decrease in the ice crystal
the number concentration and decreasing the average size
number concentration and an increase in the mean size of
of cloud droplets provided the liquid water content in the
ice crystals (Girard et al., 2005). As a result, the sedimenta-
clouds remains constant (Twomey, 1977). An increase in
tion and precipitation rates of ice crystals increase leading
the number of pollution aerosol particles that act as cloud
to an increase in the lower troposphere dehydration rate
condensation nuclei will affect arctic stratus and stratocu-
and a decrease in the downwelling infrared irradiances
mulus by increasing the cloud droplet number concentra-
from the cloud. Using a 1-D simulation and observations
tion which results in more radiation being reflected back
from Alert (Canada), Girard et al. (2005) found that a cloud
to space (Albrecht, 1989; Twomey, 1991). The relatively
radiative forcing of -9 W/m2 may occur locally as a result of
low aerosol number concentrations in the Arctic results
the enhanced dehydration rate produced by sulfate aerosol.
in a large percentage of particles activating during cloud
The mechanism by which ice nuclei concentrations are de-
formation. Hence, changes in aerosol properties are likely
creased in the presence of sulfuric acid aerosol has yet to be
to have a significant impact on microphysical and optical
explained and warrants further research. If this mechanism
cloud properties. As the cloud droplet number concentra-
applies to much of the Arctic, it could explain the cooling
tion increases, cloud droplet size decreases which reduces
tendency in the eastern high Arctic during winter.
drizzle formation and increases cloud coverage and life-
Because of the combination of the static stability of the
time potentially leading to less deposition of haze contami-
arctic atmosphere, the persistence of low level clouds, and
nants within the Arctic (Hobbs and Rangno, 1998).
the relatively long lifetime of aerosols during the haze sea-
Garrett et al. (2004) showed that low-level arctic clouds
son, the impact of aerosols on cloud microphysical and
are highly sensitive to particles that undergo long-range
optical properties may be greater in the Arctic than else-
transport during winter and early spring. The sensitivity
where on earth (Curry, 1995; Garrett et al., 2004). The win-
was detected as higher cloud droplet number concentra-
ter/spring occurrence of arctic haze events enables the
tions and smaller cloud droplet effective radii compared
study of anthropogenic influences against a very clean
to summertime clouds exposed to particles nucleated in
atmospheric background. In other regions of the globe,
the Arctic from local biogenic sources. In addition, arctic
a reliable distinction between natural and anthropogenic
stratus appears to be more sensitive to pollutant particles
effects is more difficult. In this sense, the Arctic is a natural
than clouds outside the Arctic. The most significant effect
laboratory for studying the anthropogenic portion of the
of the change in cloud properties due to arctic haze may be
aerosolcloudradiation interactions.
on cloud emissivity. A decrease in droplet effective radius
in these optically thin clouds will increase the infrared
optical depth and thus the infrared emissivity (Curry and
4.3.3. Surface albedo
Herman, 1985; Garrett et al., 2002). The result is expected
to be an increase in downwelling infrared irradiances from
Surface albedo affects the magnitude and sign of climate
the cloud and an increase in the rate of springtime snow
forcing by aerosols. Absorbing soot deposited onto the sur-
pack melting (Zhang et al., 1996).
face via wet and dry deposition affects the surface radia-
According to observations during the SHEBA (Surface
tion budget by enhancing absorption of solar radiation at
Heat Budget of the Arctic Ocean) experiment, supercooled
the ground and reducing the surface albedo (Warren and
cloud droplets are common in the Arctic even at tempera-
Wiscombe, 1980) (Figure 4.10). Clarke and Noone (1985)
tures of -20 °C or lower (Curry, 1995). The sulfate-containing
found a 1 to 3% reduction in snow albedo due to deposited
pollution aerosol within arctic haze is also thought to affect
black carbon with another factor of three reduction as the
ice nucleation. Models estimate that aerosols containing
snow ages and black carbon becomes more concentrated.
sulfuric acid produce fewer ice nuclei than nearly insoluble
Hansen and Nazarenko (2004) estimated that soot con-
a)
b)
Figure 4.10. Impact of soot deposited onto snow and ice surfaces in the Arctic. Polar ice reflects light from the sun back to space (a). As the ice
begins to melt, less light is reflected and more is absorbed by the oceans and surrounding land leading to an increase in overall temperature and
further melting. Darker, soot-covered ice reflects even less light and, thus, enhances the warming (b) (Source: NASA).
40
AMAP Assessment 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
tamination of snow in the Arctic and the corresponding
rapid industrialization of China, the increasing amounts of
decrease in surface albedo yields a positive hemispheric
pollution being transported over long distances, and indi-
radiative forcing of +0.3 W/m2. The resulting warming
cations of increasing concentrations of absorbing aerosol
may lead to the melting of ice and may be contributing to
in the Arctic more research is warranted to document the
earlier snowmelts on tundra in Siberia, Alaska, Canada,
contribution of this source to arctic haze and to determine
and Scandinavia (Foster et al., 1992).
its climate impact on the Arctic. A warming climate has
Clearly, the radiative impacts of pollutant aerosols in
been forecast to result in large increases in the areal extent
the Arctic are complex. Complex feedbacks between aero-
of fires within Russian and Canadian boreal forests (Stocks
sols, clouds, radiation, sea ice, and vertical and horizontal
et al., 1998). Hence, boreal forest fires are another source
transport processes complicate the impact as do potentially
to be monitored to determine their impact on black car-
competing effects of direct and indirect forcing. As a result,
bon concentrations in atmospheric aerosol as well as black
the magnitude and sign of the forcing for the Arctic are not
carbon that is deposited onto snow and ice surfaces.
yet well understood.
Other key atmospheric species have a distinct season-
ality in the Arctic. There is evidence of the enrichment of
halogens in arctic air masses in late winter and spring.
4.4. Summary
Since these compounds tend to peak later in the year,
it is thought that they are produced photochemically. More
Based on measurements of sulfate aerosol a main con-
research is required to partition their sources (e.g., anthro-
stituent of arctic haze and light scattering and extinction,
pogenic, especially coal combustion vs. marine), to inves-
the amount of haze and haze precursors reaching the Arctic
tigate their numerous and complex chemical pathways,
was either relatively constant or decreasing between the
and to assess their environmental impacts. Of special
1980s and early 1990s (Bodhaine and Dutton, 1993; Sirois
note is iodine, which shows a bimodal seasonal behavior,
and Barrie, 1999; Heidam et al., 2004). The updated trends
peaking in both spring and autumn (Sturges and Shaw,
in light scattering presented here show a continued de-
1983).
crease through the late 1990s with an increase in the first
The direct radiative effect of arctic haze has been es-
years of the 21st century at Barrow (Alaska). There also is
timated with 1-D radiative transfer models which find
evidence, although not as strong, of an increasing trend
a warming in the atmosphere due to absorption of solar
in black carbon over this period at Alert (Canada). Sulfate
radiation and a concurrent cooling at the surface. These
appears to have continued decreasing into the 21st century
estimates are highly sensitive to the assumed properties of
based on measurements at Alert in the Canadian Arctic,
the aerosol in the haze. Despite the many research activi-
Zeppelin on the island of Svalbard, Karasjok in northern
ties devoted to the characterization of arctic haze since the
Norway, and Oulanka in northern Finland. On the other
1970s, measurements of arctic aerosols are not extensive or
hand, nitrate appears to be increasing at Alert with an
well distributed in space or time, which limits the accuracy
unclear trend at Barrow. Continued measurements cou-
of the estimates of both the direct and indirect radiative
pled with chemical transport models are required to better
forcing. Treffeisen et al. (2004) designed an approach based
define emerging trends and to assess their causes.
on cluster analysis for integrating aircraft, ground-based,
Arctic haze is generally understood to consist of an-
and long-term data sets for use in 3-D climate models.
thropogenically-generated material and has often been at-
The accurate evaluation of climate forcing by arctic haze
tributed to sources in central Eurasia (Shaw, 1983). There
requires such data sets coupled with 3-D climate models
are examples, however, of Asian dust entering the Alaskan
that consider both direct and indirect effects. In particular,
sector of the Arctic from as long ago as the mid-1970s
3-D models are required to assess the complex feedbacks
(e.g., Rahn et al., 1981). Recent modeling studies yield
between aerosols, clouds, radiation, sea ice, and dynamic
conflicting results on whether southern Asia is a signifi-
transport and to quantify climate forcing due to arctic haze
cant source of pollutants to the Arctic or not. Given the
(Girard et al., 2005).
Document Outline
- AMAP ASSESSMENT 2006: Acidifying Pollutants, Arctic Haze, and Acidification in the Arctic
- Contents
- Preface
- Acknowledgements
- Executive Summary to the Arctic Pollution 2006 Ministerial Report
- Chapter 1: Introduction
- Chapter 2: Sources of Acidifying Pollutants and Arctic Haze Precursors
- 2.1. Sources within the Arctic
- 2.1.1. Stationary sources: industry and energy
- 2.1.2. Local air pollution in Russian cities
- 2.1.3. Oil and gas activities
- 2.1.4. Shipping activities
- 2.1.5. Natural sources within the Arctic: wildfires
- 2.2. Sources outside the Arctic and atmospheric transport to the Arctic
- 2.3. Emissions estimates used in modeling
- Chapter 3: Concentrations and Deposition of Acidifying Pollutants
- 3.1. Atmospheric and transport processes for air pollutants in the Arctic
- 3.1.1. Sulfur
- 3.1.2. Nitrogen
- 3.2. Distribution of monitoring stations
- 3.3. Concentrations, distribution, and trends in air and precipitation
- 3.3.1. Air
- 3.3.2. Precipitation
- 3.3.2.1. General pattern
- 3.3.2.2. Russian Arctic
- 3.4. Episodes and exposure to sulfur and nitrogen
- 3.5. Concentrations in seasonal snow cover
- 3.5.1. General pattern
- 3.5.2. Russian Arctic
- 3.6. Pollution history from ice cores and lake sediments
- 3.7. Modeling
- 3.7.1. Validation of the system for temporal trend analysis
- 3.7.2. Trend analysis based on measurements at Station Nord and DEHM model results
- 3.7.3. Effects of natural climate variations on long-range transport to the Arctic
- 3.7.4. Scenarios
- Chapter 4: Arctic Haze
- 4.1. The arctic haze phenomenon
- 4.2. Trends in arctic haze
- 4.2.1. Chemical composition
- 4.2.2. Optical properties
- 4.3. Effects of aerosol on the climatesystem in the Arctic
- 4.3.1. Direct effects
- 4.3.2. Indirect effects
- 4.3.3. Surface albedo
- 4.4. Summary