Artificial light at night reveals hotspots and rapid development of industrial activity in the Arctic

Abstract

Climate warming enables easier access and operation in the Arctic, fostering industrial and urban development. However, there is no comprehensive pan-Arctic overview of industrial and urban development, which is crucial for the planning of sustainable development of the region. In this study, we utilize satellite-derived artificial light at night (ALAN) data to quantify the hotspots and the development of light-emitting human activity across the Arctic from 1992 to 2013. We find that out of 16.4 million km ² analyzed a total area of 839,710 km ² (5.14%) is lit by human activity with an annual increase of 4.8%. The European Arctic and the oil and gas extraction regions in Russia and Alaska are hotspots of ALAN with up to a third of the land area lit, while the Canadian Arctic remains dark to a large extent. On average, only 15% of lit area in the Arctic contains human settlement, indicating that artificial light is largely attributable to industrial human activity. With this study, we provide a standardized approach to spatially assess human industrial activity across the Arctic, independent from economic data. Our results provide a crucial baseline for sustainable development and conservation planning across the highly vulnerable Arctic region.

Full text

PNAS 2024 Vol. 121 No. 44 e2322269121 https://doi.org/10.1073/pnas.2322269121 1 of 7
RESEARCH ARTICLE
|
Significance
Documenting the human activity
in the Arctic is a crucial step
toward the sustainable
development of the region.
Dividing human activity to its
components of urbanization
and industrialization enhances
our understanding of the
mechanisms, heterogeneity, and
concentration of the economic
activity of dierent regions. Here,
we provide a pan-Arctic scale
assessment of industrial human
activity, detecting spatial
hotspots and a high annual
development rate that adds
pressure on the vulnerable Arctic
ecosystems that are already
threatened by strong climate
change and where the smallest
anthropogenic disturbance
persists for decades. Our study
provides the basis for
investigating further economic
and ecological questions in
relation to industrial human
activity in the Arctic.
The authors declare no competing interest.
This article is a PNAS Direct Submission.
Copyright © 2024 the Author(s). Published by PNAS.
This open access article is distributed under Creative
Commons Attribution- NonCommercial- NoDerivatives
License 4.0 (CC BY- NC- ND).
Although PNAS asks authors to adhere to United Nations
naming conventions for maps (https://www.un.org/
geospatial/mapsgeo), our policy is to publish maps as
provided by the authors.
1To whom correspondence may be addressed. Email:
cengiz.akandil@ieu.uzh.ch or gabriela.schaepman@ieu.
uzh.ch.
This article contains supporting information online at
https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.
2322269121/- /DCSupplemental.
Published October 21, 2024.
ENVIRONMENTAL SCIENCES
SUSTAINABILITY SCIENCE
Artificial light at night reveals hotspots and rapid development
of industrial activity in the Arctic
CengizAkandila,1 , ElenaPlekhanovaa,b, NilsRietzea, JacquelineOehria, MiguelO.Románc, ZhuosenWangd,e, VolkerC.Radelof,
and GabrielaSchaepman- Struba,1
Aliations are included on p. 7.
Edited by Gregory Asner, Arizona State University, Hilo, HI; received December 19, 2023; accepted September 5, 2024
Climate warming enables easier access and operation in the Arctic, fostering industrial
and urban development. However, there is no comprehensive pan- Arctic overview of
industrial and urban development, which is crucial for the planning of sustainable
development of the region. In this study, we utilize satellite- derived artificial light at
night (ALAN) data to quantify the hotspots and the development of light- emitting
human activity across the Arctic from 1992 to 2013. We find that out of 16.4 million
km2 analyzed a total area of 839,710 km2 (5.14%) is lit by human activity with an
annual increase of 4.8%. e European Arctic and the oil and gas extraction regions in
Russia and Alaska are hotspots of ALAN with up to a third of the land area lit, while the
Canadian Arctic remains dark to a large extent. On average, only 15% of lit area in the
Arctic contains human settlement, indicating that artificial light is largely attributable
to industrial human activity. With this study, we provide a standardized approach to
spatially assess human industrial activity across the Arctic, independent from economic
data. Our results provide a crucial baseline for sustainable development and conservation
planning across the highly vulnerable Arctic region.
nighttime light | human activity | extractive industries | urbanization
Amplied climate warming facilitates access and development in the Arctic, which is con-
strained by harsh environmental conditions ( 1 ). is might threaten the sustainable devel-
opment in the Arctic, that can be dened as “development that improves health, well-being,
and security of Arctic communities and residents while conserving ecosystem structures,
functions, and resources” ( 2 ). For example, Arctic cities are more prone to experience
boom-bust cycles, which are based on resource extraction accompanied by a lack of diver-
sication ( 3 ). Further, direct anthropogenic impacts on Arctic ecosystems might exceed
( 4 ) or at least exacerbate the eects of climate change in the coming decades ( 5 ). If the
growth rate of industrial development between 1940 and 1990 is maintained, 50 to 80%
of the Arctic may reach critical levels of anthropogenic disturbance by 2050 ( 6 ). e main
driving force behind the industrial development in the Arctic is extractive industries ( 7 ).
For example, as of 2007, the Arctic contributed 10% of the world oil and 25% of the gas
extraction, to which Russia contributed 80% and 99%, respectively ( 8 ). Similarly, in
response to increasing global demand ( 9 ), mining has expanded in the Arctic. However,
the environmental eects of extractive industries in the Arctic, including landscape changes
and pollution, are long lasting ( 8 , 9 ) making the spatial footprint of extractive industries
and their light pollution a critical component to address the sustainability across the Arctic.
Unfortunately, measuring industrial development in the Arctic is challenging due to data
scarcity, and the large geographical extent. Industrial development is usually quantied by
the gross domestic product (GDP) and related indicators; however, this approach has many
limitations. GDP is based on market prices ( 7 ), hence a spike in the oil price might increase
GDP without an increase in oil extraction and related activity. is is especially critical for
the Arctic where industrial activity is concentrated on hydrocarbons and mining. Further,
scal transparency is an important concern for Russia ( 10 ) and public access to Russian
economic data has been blocked in 2022, making GDP-based assessment of development
impossible for this major part of the Arctic. Finally, GDP data are mostly provided at
national scale, while industrial development varies at regional and subregional scales ( 7 ).
Analysis of industrial human activity and infrastructure for selected regions shows that total
disturbed area signicantly exceeds the planned area for oil elds of Prudhoe Bay ( 11 , 12 ),
and the Bovanenkovo ( 13 ) oil elds show rapid expansion of infrastructure which trans-
forms the Arctic tundra ecosystems. At pan-Arctic scale, the mapping of current distribution
of infrastructure is limited to the Arctic coast within a 100 km buer ( 14 ). Although it is
critical to document and assess past and current industrial activity to support sustainable
OPEN ACCESS
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.

2 of 7 https://doi.org/10.1073/pnas.2322269121 pnas.org
development in this rapidly warming region, there are no compre-
hensive data available to date at pan-Arctic scale that document
the light-emitting human activity.
Articial light at night (ALAN) observed from satellites pro-
vides a reliable measure of human activity ( 15 , 16 ). ALAN has
been analyzed to quantify regional dierences in socioeconomic
activity ( 17 ), map the Terrestrial Human Footprint ( 18 , 19 ), assess
the eects of human activities on vegetation cover change ( 20 ),
and changes in human activity during the COVID-19 pandemic
( 21 , 22 ). In the Arctic, ALAN indicates light-emitting activity by
people living in settlements, transportation, and industry but does
not include non-light-emitting human activities such as reindeer
herding and agriculture. Previous studies have shown urbanization
and related ALAN increases in other parts of the world ( 23 ) and
also in the Arctic urbanization is often following industrial devel-
opment ( 24 ). In this study, we provide a comprehensive assessment
of human activity at pan-Arctic scale. Specically, we quantied
the total area aected by ALAN in the pan-Arctic and the annual
increase in extent by extracting the pan-Arctic region from a global
annual consistent and corrected nighttime light (CCNL) dataset
based on Defense Meteorological Satellite Program (DMSP) sat-
ellite observations in the time period 1992 to 2013, at a spatial
resolution of 30 arc-second. CCNL is corrected for blooming and
saturation, which allows the user to run a trend analysis ( 25 ).
Currently, no ALAN data are available beyond 2013 with robust
correction for ephemeral lights (aurora borealis), which is essential
for the analysis of ALAN at high latitudes (SI Appendix, Figs. S1
and S2 and Text ). is analysis provides an insight into spatial
hotspots of human activity. Further, we analyzed how light inten-
sities changed through time for dierent regions in pan-Arctic.
To assess the mechanisms contributing to this development, we
determined how much of the ALAN can be attributed to human
settlement vs. industrial development in the pan-Arctic and for
dierent regions. Finally, dierences of ALAN development for
the two main extractive industries in the Arctic were exemplied
for oil & gas extraction and mining.
Results
Total Lit Area in the Arctic and Its Regional Distribution. We
analyzed two decades of ALAN from 1992 to 2013 to assess the
development of human activity in the terrestrial Arctic, as dened
by the area of economic regions (7, 26) (SIAppendix, TableS1).
We found that the total lit area by human activity in the Arctic
was 839,710 km
2
from 1992 to 2013 (5.14%, Table1). e North
American Arctic was the least- aected continent, only 0.30% of
the Canadian Arctic (10,813 km2) and 2.52% (38,581 km2)
of Alaska were lit. Arctic hotspots of total lit area occurred in
Russian oil & gas extraction regions and in the European Arctic.
e main oil- extracting region in the Russian Arctic, Khanty-
Mansi, had 178,681 km2 (33.30%) of its area lit. Kamchatka,
another region in the Russian Arctic, had a total lit area of only
5,610km2 (1.21%) (SIAppendix, TableS2). In the European Arctic
(excluding Greenland), 28.67% (159,085 km2) of the area (less
than Khanty Mansi alone) exhibited ALAN throughout the 22- y
period analyzed. Finland contributed most to that (50,782km2,
31.22%), followed by Norway (46,803 km
2
, 38.56%), and Sweden
(46,292 km
2
, 27.68%). ese results highlight the strong regional
concentration of light- emitting human activity in the Arctic.
We calculated the newly lit area by subtracting the lit area in
the base year (1993 for European Arctic, 1992 otherwise) from
the cumulative lit area of 1992 to 2013. During 1992 to 2013,
605,138 km2 (3.70%) of the terrestrial Arctic transformed from
dark to lit area. e Russian Arctic exhibited the largest increases
in lit area (439,048 km2 ) across the Arctic, especially in Khanty
Mansi (114,426 km2 ) and Yamal Nenets (107,837 km2 )
(SI Appendix, Table S2 ). We derived the annual increase in lit area
by calculating the slope of rst-order Autoregressive Integrated
Moving Average (ARIMA) model to account for temporal auto-
correlation ( 27 ). We divided the annual increase in lit area by the
base year’s lit area to quantify the annual rate of increase in lit area.
e modeled annual rate of increase in lit area in the terrestrial
Arctic was 4.80% (11,263 km2 ) from 1992 to 2013 ( Fig. 1A ).
Increases in Articial Light Intensity in the Arctic. We derived
the annual increase in regional aggregate cumulative light intensity
by calculating rst- order ARIMA. We divided the annual increase
in regional aggregate cumulative light intensity by the base year’s
aggregate cumulative light intensity to quantify the annual rate
of increase in regional aggregate cumulative light intensity, which
was 4.10% in the pan- Arctic (Fig.1B). At the regional level, the
strongest annual rate of increase in regional aggregate cumulative
light intensity of 37.41% occurred in the Nenets region.
We calculated spatial and temporal light intensity trends in the
terrestrial Arctic from 1992 to 2013 using rst-order ARIMA.
e light intensity trends signicantly increased in 1.41% of the
Table1. ALAN area and development during 1992 to 2013, in the pan- Arctic and for selected continental areas
Total area
(km2)
Total lit area
(km2) (percent of
total area
that is lit)
Newly lit area
(km2) (percent of
total area that is
newly lit)
Annual rate
of increase
in lit area
(* indicates
signicant)
Annual rate
of increase
in regional
aggregate
cumulative
light intensity
Percentage of lit
area containing
human settlement;
average (and
range) across years
pan-Arctic 16,352,106 839,710 (5.14%) 605,138 (3.70%) 4.80%* 4.10%* 15.0%
(12.7 to 18.9%)
Russian Arctic 9,387,892 629,897 (6.71%) 439,048 (4.68%) 4.17%* 3.49%* 12.4%
(10.3 to 15.9%)
European Arctic
(excl. Greenland)
554,810 159,085 (28.67%) 106,108† (19.13%) 4.53%*† 3.82%†* 21.9%
(18.1 to 26.8%)
North American
Arctic
5,082,640 49,378 (0.97%) 35,618
(0.70%)
3.60%* 3.51%* 25.4%
(22.7 to 31.1%)
Total lit area is the sum of all area lit; annual rate of increase in lit area is based on the ARIMA trend of annual lit area and * indicates signicant values based on P value of 0.05; annual
rate of increase in regional aggregate cumulative light intensity is based on ARIMA trend of annual regional aggregate cumulative light intensity and* indicates signicant values based on
P value of 0.05; † represents base year 1993 instead of 1992 due to missing data; percentage of lit area containing human settlement is calculated for 1995, 2000, 2005, and 2010, when
human settlement data are available, indicated number is average of the annual values and min and max values are indicated in the parentheses.
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.

PNAS 2024 Vol. 121 No. 44 e2322269121 https://doi.org/10.1073/pnas.2322269121 3 of 7
terrestrial Arctic, while only 0.28% of the area exhibited a signif-
icant decline (P -value < 0.05) from 1992 to 2013 (SI Appendix,
Table S2 ). Again, growth rates were spatially highly heterogeneous
( Fig. 2 ). e European Arctic had the largest relative area with
positive light intensity trends (9.71% of total area), especially
Norway. e Russian Arctic had a much smaller relative area with
positive light intensity trends (1.74% of total area), and a relatively
high area that decreased (0.46%). At the regional scale, the largest
relative area with a decline in light intensity trends in the entire
Arctic occurred in Khanty Mansi (3.34% of total area), the heart
of oil extraction, where there was also increase in 9.08% of the
area (SI Appendix, Figs. S7–S33 ).
Human Settlement as Minor Contributor to Total Lit Area in the
Arctic. We used the Global Human Settlement Layer (GHSL)
population grid, a dataset based on population census data, to
calculate the percentage of lit area containing human settlement
by dividing lit area inhabited by humans to the lit area for 1995,
2000, 2005, and 2010 and report the average and the min- max
range of these years [(Table1 and SIAppendix, TableS2]. On
average, 15% of lit area contained human settlement in the Arctic
but with large regional dierences, ranging from 4.7% in Yamal-
Nenets to 37.5% in Faroe Islands. On average, in North America,
25.4% of lit area contained human settlement, followed by Europe
(21.9%) and Russia (12.4%). In oil & gas extracting regions of
Russia (Nenets, Yamal Nenets, and Khanty Mansi), the percentage
of lit area containing human settlement ranged from 4.7 to 10.5%,
the lowest 3 regions in the pan- Arctic (SIAppendix, TableS2).
Overall, this suggests that human settlements played a minor role
in contributing to ALAN in the Arctic, compared to industrial
and extraction activities.
ALAN Development in Areas of Oil & Gas Extraction versus
Mineral Extraction. Extractive industries are the main light- emitting
industrial activity in the Arctic causing physical disturbance and
pollution with long- lasting impacts (8, 9). However, we found that
the spatial footprints of light pollution of the oil and gas industry
was much more widespread than that of the mining industry. Below
we rst present four case studies of dierent extractive industries and
then compare their spatial footprint of light pollution.
Samotlor is located in Khanty-Mansi, Russia ( Fig. 3A ), and is
one of the largest oil elds in the world ( 28 ). Its development
started in 1967, with rst extraction in 1969, reaching a peak in
1980 with 3.2 million barrels per day, dropping to less than
1 million barrels per day in the 1990s ( 28 ). More than 20,000
wells have been drilled in this eld together with 5,911 km of oil
pipelines and 1,923 km of paved roads ( 28 ). e Vankorskoye oil
elds are in the Krasnoyarsk region, Russia ( Fig. 3B ). Oil extrac-
tion started in 2009, with 150 km of ineld pipelines, 60 km of
gas pipelines, 100 km of motorways, 700 km of power transmis-
sion lines, and camps for 2000 oil eld workers built and ultimately
a total of 425 wells ( 29 ). Norilsk is one of the largest industrial
slope = 11,263 km2/year
200,000
300,000
400,000
500,000
600,000
1995 2000 2005 2010
year
Lit area (km2)
slope = 261,954 DN/year
6,000,000
9,000,000
12,000,000
1995 2000 2005 2010
year
Total light intensity (DN)
AB
Fig. 1. (A) Change in annual lit area in the terrestrial Arctic and its trend along
with 95% CI over 22 y, (B) Change in annual aggregate cumulative light intensity
and its trend along with 95% condence over 22 y (both trendlines are based
on ARIMA).
A
BC
Unlit
Lit
Increasing
Decreasing
Fig. 2. Continental distribution of ALAN intensity trend map, showing unlit areas, lit areas, and signicantly increasing or decreasing ALAN intensity (P- value < 0.05).
(A) Russian Arctic, (B) North American Arctic, and (C) European Arctic (excl. Greenland, Iceland, and Faroe Islands). (Made with Natural Earth) (Interactive map
on https://arcg.is/H1bPK).
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.

4 of 7 https://doi.org/10.1073/pnas.2322269121 pnas.org
mining towns in the world ( Fig. 3B ) and Nornickel runs the min-
ing operations. As of 2021, even after decades of mining, Nornickel
remains the world’s largest high-grade nickel extractor, the largest
palladium, the 4th largest platinum, the 5th largest rhodium and
the 12th largest copper extractor and reserves allow to continue
operation at current extraction rates for more than 75 y ( 30 ). e
Red Dog mine in Alaska is the 2nd largest zinc extractor in the
world as of 2018, operating in an open pit mine 80 km away from
Chukchi Sea ( 31 ) ( Fig. 3C ). Construction began in 1987, and the
mine was operational since 1989 ( 31 ).
e Vankorskoye oil elds had a total lit area of around 12,500 km2 ,
although extraction only started in 2009. e total lit area in the main
oil and gas extracting regions in the Russian Arctic (Khanty-Mansi,
Yamal-Nenets, and Nenets) was 339,430 km2 , almost the size of
Germany. ese three subregions contributed 40.4% of the total lit
area in the pan-Arctic. In contrast, the Norilsk and Red Dog mines
have been operating for decades, and yet the total lit area in Norilsk
(including the city) was only 5,130 km2 , and at the Red Dog Mine
135 km2 .
Discussion
We analyzed 22 y of ALAN data to quantify the industrial human
activity across the Arctic and found that more than 800,000 km2
were aected by light pollution and corresponding human activity.
However, only 15% of the lit area contains human settlements,
indicating that industrial activity is the major cause for ALAN
and its increases in both area and intensity. We show that ALAN
development in the Arctic is rapid, quite independent of human
settlement, and shows large dierences depending on the regional
economic activity, with oil and gas extraction regions being the
most important hotspots.
Our ndings suggest the predominant importance of industrial
activities across the Arctic to ALAN development. However, this con-
tribution varies strongly between regions. Nenets, Yamal-Nenets, and
Khanty-Mansi are the heart of the oil and gas extraction ( 32 ), they
have the smallest fraction of lit area containing human settlement
among the pan-Arctic regions, despite hosting some of the largest
cities in the Arctic such as Surgut and Novy-Urengoi. Although indus-
trialization was the main driver behind the establishment of these cities
( 24 ), their contribution to lit area is dwarfed by the ongoing heavy
industrial activity in these regions. Urbanization in the European
Arctic is characterized by population growth occurring mainly within
the urban centers lacking suburban population ( 33 ). e extensive
signicant increase in ALAN intensity trend in 9.71% of the area in
the European Arctic—up to 29.07% in Tromso—indicates the level
of human activity. However, the percentage of lit area containing
human settlement in the European Arctic is only 21.9%—for Tromso,
it is 10.6%—which indicates industrial development as the driving
force of human activity development rather than urbanization. e
North American Arctic on the other hand shows signicant increase
in ALAN intensity in only 0.21% of the area, while 25.4% of the lit
area contains human settlement, indicating comparatively lower level
of industrialization contributing to ALAN development.
We use ALAN as indicator of human activity in the terrestrial
Arctic, but data uncertainties require cautiousness. Snow albedo
eects might enhance articial light eects, but that means it also
should be included as part of the light pollution of human activity.
8
0
N
6
0
N
c) b)
a)
Unlit
Lit
Increasing
Decreasing
A
CB
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
S
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
A
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
M
MM
M
M
M
M
M
M
M
M
M
M
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
L
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
R
S
A
M
O
T
L
O
R
V
A
N
K
O
R
V
A
N
K
O
R
N
O
R
I
L
S
K
N
O
R
I
L
S
K
R
E
D
D
O
G
R
E
D
D
O
G
N
0 10050
km
N
0 10050
km
N
0 10050
km
Fig. 3. ALAN intensity trend maps in major oil & gas versus mineral extracting regions in the Arctic showing unlit areas, lit areas, and increasing or decreasing
ALAN intensity trend (P- value < 0.05). (A) Part of the oil & gas extraction region of Khanty- Mansi including Samotlor oil elds in the southeast (B) Red Dog mine,
(C) Vankorskoye oil elds and Norilsk mine. All maps are at the same spatial scale, indicating large dierences in total lit area between oil and gas extraction and
mining industries. (Made with Natural Earth).
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.

PNAS 2024 Vol. 121 No. 44 e2322269121 https://doi.org/10.1073/pnas.2322269121 5 of 7
ere can be an overestimation of lit area from satellites compared
to actual lit area on the ground mainly due to large pixel size and
geolocation errors ( 34 ) or residuals in the corrected dataset sourced
by auroras and blooming eects, yet also a signicant underesti-
mation of the area in underdeveloped regions ( 35 ). However, while
the total lit area might overestimate the direct impact by infrastruc-
ture, it is a reliable indicator of both hotspots of current human
activity in the Arctic and of development over time. Accordingly,
we use the area of ALAN as an indicator of the total area impacted
by human activities and its spatial distribution. However, the pres-
ence of roads and buildings provides access to the surrounding
areas, and cumulative eects of settlements, roads, and pipe- and
powerlines on ecosystems can extend up to 40 km in the Arctic
( 6 ). In the highly vulnerable permafrost landscape and tundra eco-
system, even just repeated trampling by humans and certainly
tracks by tundra vehicles can exert long-term environmental eects
reaching well beyond the lit area detected from satellites.
e pan-Arctic expansion of human activity has important
implications for biodiversity and ecosystem functions. e Arctic
has a strong seasonal variation in natural light conditions, with
very low natural light levels during wintertime. is variation of
light has led to physiological adaptations of endemic Arctic spe-
cies. Under controlled constant light conditions, circadian activity
patterns of Svalbard ptarmigans (Lagopus muta hyperborea ) weaken,
which is a concern because ptarmigans likely trigger their seasonal
biology (white winter plumage vs. brown summer plumage) based
on circadian processes ( 36 ). Arctic reindeer (Rangifer tarandus )
adapt their eyes to extreme blue color of the twilight that occurs
during the Arctic winter, which allows reindeer to nd food and
escape predators ( 37 ). In a recent synthesis study not limited to
the Arctic, melatonin secretion showed 36% decrease on average
in response to ALAN across 31 wild vertebrate species ( 38 ). For
the plants, ALAN also delays the coloring of leaves and breaking
leaf buds ( 39 ), which is critical for the Arctic species where the
growing season is limited. Further, the area lit by human activities
may serve as a spatial indicator of other disturbances of ecosystems.
e expansion of invasive species in the Arctic has been linked to
disturbance of the landscape by human activity such as increased
shipping, oil and gas exploration, mineral exploration, and asso-
ciated infrastructure development ( 40 ). Hence, the spatial distri-
bution of ALAN may be a proxy for invasion risk linked to
industrial human activity. Pollution by extractive industries is
another concern. Inland water bodies near oil and gas extraction
sites are under stress due to leaks from aging or poorly maintained
pipelines. Indeed, the oil content in bottom sediments of some of
these water bodies can reach up to 111 g/kg ( 41 ). As many of
these inland water bodies are outside of the area of the articial
light emitted by the extraction infrastructure, our results probably
underestimate the spatial impact related to such pollution eects.
e eects of rapid climate change in the Arctic require rapid
adaptation by local communities and the industrial development
might further increase adaptation needs ( 42 ). For example,
Yamal-Nenets has the largest reindeer herding area in the world
and due to climate change, herders already need to shift their
grazing grounds regularly ( 42 ). e heavy increase in industrial
development in Yamal-Nenets limits the options for adaptation
of reindeer herders due to pollution and reduction in the total
grazing area. We showed here the spatial extent and concentration
of the industrial human activity in the Arctic, including the heavy
industrial development in the Yamal Peninsula. e spatial vari-
ability and current hotspots of industrial development are critical
components of building capacity for resilience and adaptability
of the Arctic communities to achieve sustainable development
goals in the Arctic.
ALAN data are an important information source to indicate
the spatial footprint of extractive industries. During the develop-
ment of new oil and gas elds, the cost of transportation in the
Arctic is substantial due to harsh environmental conditions.
However, once a pipeline or transportation route is established,
they economically facilitate further expansion of new elds, espe-
cially if the extraction in the main elds declines ( 8 ). erefore,
dierent industrial activities based on dierent phases of the oil
extraction life cycle occur at the same time, such as extraction in
the main elds, infrastructure development for new elds at the
margins, and exploration in far regions ( 8 ), all of which we capture
in our results. Especially Khanty Mansi, one of the oldest oil
extraction regions, has both some of the most widespread signi-
cant increasing and decreasing trends in ALAN intensities among
all Arctic regions, indicating above-described abandonment.
Mineral extraction on the other hand is more concentrated at the
core area of extraction and represents typically long-term devel-
opment. Extractive industries play a major role for the Russian
economy, and fossil fuel subsidies in Russia in 2015 amounted to
551 billion USD which considers the eects of negative externality
components such as air pollution and climate change, while ignor-
ing changes in land cover sourced from fossil fuel extraction elds,
methane emissions from old and current oil wells, and light pol-
lution ( 43 ). Our ndings suggest that the spatial footprint of light
pollution is substantial for oil & gas elds in the Arctic and needs
to be integrated to quantify the externalities from fossil fuel extrac-
tion. However, there are many more biophysical feedbacks sourced
from extractive industries ( 42 ), and these externalities must be
accounted for to estimate the real cost of industrial activity in the
Arctic on society and the environment. e presented approach
based on ALAN might be a supportive tool in this process.
Although our study covers more than two decades, it is crucial
to carry this analysis to the present to assess the most recent devel-
opments. Ideally such analysis will be based on data with higher
spatial resolution and radiometric calibration, such as VIIRS
nighttime lights. e methodology for the correction of aurora
eects in VIIRS nighttime light data has recently been established
( 44 ). Once the entire data series is processed, it will oer oppor-
tunities to analyze light-emitting human activity in the Arctic.
e social, ecological, and economic systems in the Arctic are
tightly coupled and interconnected ( 42 ). A social–ecological sys-
tems framework to analyze the rapidly changing Arctic is needed
to ensure the sustainable development of the Arctic ( 42 ). Our
nding that ALAN has increased rapidly in both area and intensity
in many parts of the Arctic indicates that industrial development
is a major factor directly impacting the terrestrial Arctic ecosystems
and sustainability of the region.
Materials and Methods
Study Area. The study was carried out in different terrestrial Arctic regions,
which were selected based on the Arctic Human Development Report (AHDR)
of 2015 (26) to allow for a comparative analysis among regions and countries
(SIAppendix, TablesS1 and S2). We further regrouped the countries as Russian,
European, and North American Arctic to highlight the larger regional differences.
For Russia, AHDR describes Taimyr and Evenk separately, but we used Krasnoyarsk
as Taimyr was merged with Krasnoyarsk in 2006. Also, Koryak was merged with
Kamchatka; therefore, we used Kamchatka.
Data. ALAN data from the DMSP/Operational Linescan System are provided as a
yearly composite from 1992 to 2013 between 65°S and 75°N at 30 arc- second
pixel resolution. DMSP can detect light sources with radiances as low as 1.54 ×
10−9W cm−2 sr−1 µm−1 and indicates light intensity (45). We used the term ALAN
because DMSP is corrected for auroras which reduces nighttime light to ALAN
(SIAppendix, Text).
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.

6 of 7 https://doi.org/10.1073/pnas.2322269121 pnas.org
For our time- series analysis, we utilized “A CCNL dataset” (25), that addresses
three problems of the original DMSP dataset: 1) 6 Different satellites were used
from 1992 to 2013, which resulted in interannual inconsistency throughout years
due to lack of onboard calibration, satellite shift, and sensor degradation (46).
2) DMSP can detect radiances as low as 1.54 × 10−9W cm−2 sr−1 µm−1 and as high
as 3.17 × 10−9 W cm−2 sr−1 µm−1 and translates them to digital numbers (DN) on
a six- bit scale, where zero corresponds to no detectable light and 63 corresponds
to highest light detectable (45). The low radiometric resolution creates saturation
problems (i.e., radiances above detectable are assigned a DN value of 63). This issue
is especially important when studying gas flares from the extractive industries in this
study. 3) Blooming, which is primarily caused by atmospheric conditions and data
resampling. Correction of these problems allows the users to run a trend analysis.
When visually assessing the CCNL, we noticed several clusters of low intensity
lit areas 2 to 4 km
2
in diameter, which were lit only in 1 y across the 1992 to 2013
period and appeared at some distance from the previously lit areas. Those clusters
sometimes appeared in the areas very unlikely lit by human activity, therefore
we treated them as outliers. Specifically, we implemented an algorithm to filter
out all the pixels that were lit only once, that were >5 km away from a pixel that
was lit at least twice. We selected the 5- km threshold to retain the areas that were
expanding from the previously lit but filter out the outliers. The filtered- out pixels
accounted for approximately 3% of the total lit area. We then used the corrected
dataset as the basis for our study.
In CCNL dataset, three regions (Finnmark, Nordland, Tromso) had missing data
for 1992. Therefore, we calculated newly lit area, annual rate of increase in lit area,
spatial and temporal light intensity trends, and annual rate of increase in regional
aggregate cumulative light intensity for these three regions with 1993 as base year.
Gas flaring is a common practice to burn excess gas that cannot be shipped
to markets and one of the main sources of artificial lights in oil and gas extrac-
tion (47, 48). Even though most of socioeconomic studies exclude gas flaring
(burning of excessive gas) in calculation of related human activities (45), we
decided to keep it because it is an important part of the light pollution in the
pan- Arctic. Accordingly, we included gas flares when quantifying the area affected
by industrial human activity.
The GHSL is produced by the European Commission’s Joint Research Center.
GHSL is based on Gridded Population of the World (GPW) version 4.11 (49).
GPWv4 is based on population and housing censuses and disaggregated to create
the distribution of human population indicated as the number of people per
grid cell. We expect the GHSL to capture most human- settlement light- emitting
activities because especially smaller settlements are restricted in their spatial
extent following the “ideal Arctic town” model which is a highly concentrated
town structure to avoid inconvenience due to harsh climatic conditions (i.e., the
Svappavaara settlement or Resolute Bay developed by the architect Erskine) (50).
ALAN is not used in producing GHSL, so the datasets are independent. GHSL
is prepared from 1975 to 2030 for every 5 y and we used 1995, 2000, 2005,
and 2010, with 30 arcsec resolution and WGS84 projection, which match years,
resolution, and the projection of the ALAN data.
For the regional analyses, we downloaded regional shapefiles from the
Database of Global Administrative Areas (GADM 4.1) (51).
For the interactive map, we downloaded the global mining areas dataset
(52) along with Global Oil and Gas Extraction Tracker from Global Energy Monitor
(53), and for the coal mines Global Coal Mine Tracker from Global Energy Monitor (54).
Statistical Analysis. All statistical analyses were conducted in R [version 4.3.1;
R: The R Project for Statistical Computing (r- project.org)]. All maps were created
using ArcGIS version 3.0 by ESRI. We used “terra” and “raster” packages to
compile, resample, and summarize spatial data. For area calculations, we used
cellSize() from “terra” package. All ARIMA analyses were conducted with arima()
function from the “stats” package.
We calculated cumulative ALAN map by summing up the ALAN DN values for
every year from 1992 to 2013. We defined values above zero as “lit” while zero
values as “unlit.” The total lit area was the total area of lit pixels in cumulative
map. Newly lit area was defined as area that was lit in cumulative ALAN map, but
unlit in the base year (1993 for European Arctic and 1992 otherwise). We used the
cumulative map instead of 2013 ALAN data due to the possibility of discontinuous
industrial activity such as oil or mineral exploration in earlier years that ceased
by 2013. To estimate the annual trend of the lit area we fitted a first- order ARIMA
that corrects for temporal autocorrelation (Fig.1A) (27).
We calculated annual rate of increase in lit area from the resulting ARIMA
slope by the following formula:
Annual rate of increase in lit area
=
Area based on ARIMA slope
Area lit
base year
∗100%
.
For each year and region, we calculated the aggregate cumulative DN values
to assess the changes in the light intensity. To estimate the annual trend of the
light intensity we fitted a first- order ARIMA (Fig.1B). Annual rate of increase in
regional aggregate cumulative light intensity [also named “sum of NTL” in Zhao
etal. (25)] was calculated with the below formula:
Annual rate of increase in regional aggregate cumulative light intensity
=Aggregate cumulative light intensity based on ARIMA slope
Aggregate cumulative light intensit y
base year
∗100%
.
ALAN Trend Map Calculation. We calculated pixel- level ALAN intensity trend maps
by applying first- order ARIMA to every pixel of ALAN dataset from 1992 to 2013 across
the pan- Arctic. We derived the slope of the trendline and tested the significance of
the slope (P < 0.05). In the map, areas with significant positive slope in intensity were
labeled as “increasing” and areas with significant negative slope as “decreasing.” Lit
areas without significant trends were named “lit.” We initially performed the analysis
with Theil- Sen model (which is robust to outliers) but decided against using it because
we think that the outliers are an important part of the data due to the highly dynamic
nature of the changing human activity in the Arctic, such as development of new
infrastructure or pausing of a mine when the commodity prices go down.
Attribution of ALAN to Global Human Settlement. The GHSL is available in
5- y intervals for 1975 to 2030, albeit with different map projections and resolu-
tions. We downloaded the global data for the years 1995, 2000, 2005, and 2010
in WGS84 projection and a resolution of 30 arcsec, matching the projection and
resolution of the ALAN data. We used the resample function in the terra package
in R with “near” method for GHSL data to align the grids and extents of the two
datasets and avoid linear traces (when we resampled the ALAN data instead of
GHSL, our results changed by only 1%). To determine the relationship between
GHSL and ALAN, we calculated the percentage of lit area containing human set-
tlement among total lit area. Note, the calculation was based only on area and
not on light intensity or human population per pixel. We calculated this ratio for
1995, 2000, 2005, and 2010 and reported the average and the range.
Creating Maps and Calculating the Area of Extractive Industries in the
Case Studies. We used Cylindrical Equal Area projection, centered on 60°N for
less distortion of high latitude areas. Cylindrical equal area ensures that the area is
preserved throughout the map but at the cost of varying distance across latitudes.
Therefore, we removed the map’s scale bar to account for the change in distance
if the region’s latitudinal range stretched more than ten degrees of latitude, e.g.,
a pixel (30 arcsec × 30 arcsec) at 60°N has a side length of ca. 463 m, whereas
a pixel at 70°N has a side length of ca. 316 m. Hence, a scale bar would not
appropriately represent the distances depicted on the map.
To compute the lit area for the case studies (i.e., Norilsk, Red Dog, and Vankor),
we 1) manually outlined an area of interest, 2) clipped the cumulative ALAN raster to
that area, 3) converted it to binary (0 = lit area, 1 = unlit), 4) polygonized the binary
raster, 5) calculated the area in km2, 6) and summarized within the region of interest.
Data, Materials, and Software Availability. Anonymized code, maps, interactive
ALAN map data have been deposited in github (https://github.com/PlekhanovaElena/
ALAN_Arctic) (55).
ACKNOWLEDGMENTS.
We thank Dr. Urs Stampfli and Barbara Stampfli for their
support to this study through the University of Zurich (UZH) Foundation. This
study was supported by Barbara and Urs Stampfli via UZH Foundation, NASA
grant #80NSSC18K0678 (The Science of Terra/Aqua/Suomi- NPP) and NASA
grant # 80NSSC22K0199 [NASA’s Suomi- National Polar- orbiting Partnership
(Suomi- NPP), and Joint Polar Satellite System (JPSS) Program]. This study made
use of infrastructure services provided by the Science IT team of the University
of Zurich (www.s3it.uzh.ch). We thank Anthony R. Ives for his contribution to the
methodology for accounting for temporal autocorrelation.
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.

PNAS 2024 Vol. 121 No. 44 e2322269121 https://doi.org/10.1073/pnas.2322269121 7 of 7
Author aliations: aDepartment of Evolutionary Biology and Environmental Studies, University
of Zurich, Zurich 8057, Switzerland; bLand Change Science Research Division, Dynamic
Macroecology group, Swiss Federal Research Institute for Forest, Snow, and Landscape,
Birmensdorf 8903, Switzerland; cEarth Sciences Division, NASA Goddard Space Flight Center,
Greenbelt, MD 20771; dEarth System Science Interdisciplinary Center, University of Maryland
College Park, College Park, MD 20742; eTerrestrial Information Systems Laboratory, NASA
Goddard Space Flight Center, Greenbelt, MD 20771; and fDepartment of Forest and Wildlife
Ecology, University of Wisconsin- Madison, Madison, WI 53706
Author contributions: C.A. and G.S.- S. conceptualized the study; C.A., E.P., N.R., J.O., M.O.R.,
Z.W., V.C.R., and G.S.- S. developed the methodology; C.A. and E.P. ran the formal analysis
along with validation, data curation, and implementation of the computer code; C.A. and
G.S.- S. wrote the draft; C.A., E.P., N.R., J.O., M.O.R., Z.W., V.C.R., and G.S.- S. reviewed the
paper; C.A., E.P., N.R., J.O., M.O.R., Z.W., V.C.R., and G.S.- S. edited the paper; C.A., E.P., and
N.R. made the visualizations, G.S.- S. supervised and administered the project; and C.A.,
M.O.R., and G.S.- S. acquired the funding; and C.A., E.P., N.R., J.O., M.O.R., Z.W., V.C.R., and
G.S.- S. wrote the paper.
1. Conservation of Arctic Flora and Fauna (CAFF), “Arctic Biodiversity Assessment: Report for Policy
Makers” (CAFF, Akureyri, Iceland, 2013). https://oaarchive.arctic- council.org/server/api/core/
bitstreams/564e2092- c358- 4673- 9e4a- fd40ee948252/content. Accessed 29 September 2023.
2. J. K. Graybill, A. N. Petrov, “Introduction to Arctic sustainability” in Arctic Sustainability, Key
Methodologies and Knowledge Domains, J. K. Graybill, A. N. Petrov, Eds. (Routledge, 2020), pp. 1–22.
3. M. Berman, R. W. Orttung, Measuring progress toward urban sustainability: Do global measures
work for arctic cities? Sustainability 12, 3708 (2020).
4. B. C. Forbes, J. J. Ebersole, B. Strandberg, Anthropogenic disturbance and patch dynamics in
circumpolar arctic ecosystems. Conserv. Biol. 15, 954–969 (2001).
5. Q. Yu, H. E. Epstein, R. Engstrom, N. Shiklomanov, D. Strelestskiy, Land cover and land use changes in
the oil and gas regions of Northwestern Siberia under changing climatic conditions. Environ. Res. Lett.
10, 124020 (2015).
6. United Nations Environment Programme, Globio: Global Methodology for Mapping Human Impacts
on the Biosphere: The Arctic 2050 Scenario and Global Application (UNEP- DEWA, 2001).
7. I. Mäenpää, S. Glomsrød, T. Wei, “Comparative analysis of Arctic economies at macro level” in The Economy
of the North 2015, S. Glomsrød, G. Duhaime, I. Aslaksen, Eds. (Statistics Norway, 2017), pp. 27–35.
8. J.- I. S. (Permanent P. of the I. P. O. John Calder (Chair, USA), Per Døvle (Vice- chair, Norway), Yuri
Tsaturov (Vice- chair, Russia), Russel Shearer (Canada), Ruth McKechnie (Canada), Morten Olsen
(Denmark), Outi Mähönen (Finland), Helgi Jensson (Iceland), Erik Syvertsen (Norway), Yngve
Brodin, Arctic Oil and Gas 2007 (2007).
9. K. Van Dam, A. Scheepstra, J. Gille, A. Stepien, T. Koivurova, “Mining in the European Arctic” in
Strategic Assessment of Development of the Arctic: An Assessment Conducted for the European Union
Report, S. Arnarsson et al., Eds. (Lapin yliopisto, Arktinen keskus, Rovaniemi, 2014), pp. 87–99.
10. International Monetary Fund, Russian federation: Fiscal transparency evaluation. IMF Staff Country
Reports (2014), p. 14.
11. D. A. Walker, M. K. Raynolds, M. Buchhorn, J. L. Peirce, Landscape and permafrost changes in the
Prudhoe Bay Oilfield, Alaska. Alaska Geobotany Center Publication AGC 14- 01 (2014), p. 84.
12. D. A. Walker et al., Cumulative impacts of oil fields on Northern Alaskan Landscapes. Science 238,
757–761 (1987).
13. T. Kumpula, A. Pajunen, E. Kaarlejärvi, B. C. Forbes, F. Stammler, Land use and land cover change in
Arctic Russia: Ecological and social implications of industrial development. Glob. Environ. Change
21, 550–562 (2011).
14. A. Bartsch, G. Pointner, T. Ingeman- Nielsen, W. Lu, Towards circumpolar mapping of arctic
settlements and infrastructure based on sentinel- 1 and sentinel- 2. Remote Sens. 12, 2368 (2020).
15. N. Levin et al., Remote sensing of night lights: A review and an outlook for the future. Remote Sens.
Environ. 237, 111443 (2020).
16. C. D. Elvidge, M. Zhizhin, T. Ghosh, F. C. Hsu, J. Taneja, Annual time series of global viirs nighttime
lights derived from monthly averages: 2012 to 2019. Remote Sens. 13, 1–14 (2021).
17. N. B. Weidmann, G. Theunissen, Estimating local inequality from nighttime lights. Remote Sens. 13,
1–17 (2021).
18. O. Venter et al., Sixteen years of change in the global terrestrial human footprint and implications
for biodiversity conservation. Nat. Commun. 7, 12558 (2016).
19. H. Mu et al., A global record of annual terrestrial Human Footprint dataset from 2000 to 2018.
Sci. Data 9, 176 (2022).
20. N. Xia, M. Li, L. Cheng, Mapping impacts of human activities from nighttime light on vegetation
cover changes in southeast Asia. Land 10, 1–23 (2021).
21. T. Lan et al., Quantifying spatiotemporal changes in human activities induced by COVID- 19
pandemic using daily nighttime light data. IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens. 14,
2740–2753 (2021).
22. E. C. Stokes, M. O. Román, Tracking COVID- 19 urban activity changes in the Middle East from
nighttime lights. Sci. Rep. 12, 1–13 (2022).
23. S. S. Satheendran, S. S. Chandran, J. C. Mathew, The evolution of lighting in South- West India from
night- time lights: 2012–2020. Spat. Inf. Res. 30, 261–277 (2022).
24. Y. Vaguet, Oil and Gas towns in Western Siberia: Past, present and future challenges. Nordregio 6,
125–132 (2013).
25. C. Zhao, X. Cao, X. Chen, X. Cui, A consistent and corrected nighttime light dataset (CCNL
1992–2013) from DMSP- OLS data. Sci. Data 9, 1–12 (2022).
26. L. Huskey, I. M
ӓ
enp
ӓӓ
, A. Pelyasov, “Economic systems” in Arctic Human Development Report
Regional Processes and Global Linkages, J. N. Larsen, G. Fondahl, Eds. (Nordic Council of Ministers,
2014), pp. 151–183.
27. A. R. Ives et al., Statistical inference for trends in spatiotemporal data. Remote Sens. Environ. 266,
112678 (2021).
28. JSC Samatlorneftegaz History. https://samotlorneftegaz.rosneft.com/about/Rosneft_today/Operational_
structure/Development_and_Production/Samotlorneftegaz/History/. Accessed 26 June 2023.
29. Vankor’s Millions. https://www.rosneft.com/press/today/item/175201/. Accessed 26 June 2023.
30. V. Potanin, S. Malyshev, “2021 Annual Report MMC Norilsk Nickel Navigating the Transition to a Net
Zero World” (Nornickel, Moscow, Russia, 2022). https://nornickel.com/upload/iblock/74e/2021_
annual_report_030622.pdf. Accessed 29 September 2023.
31. M. Berman, R. Loeffler, J. I. Schmidt, Long- term benefits to Indigenous communities of extractive
industry partnerships: Evaluating the Red Dog Mine. Resour. Policy 66, 101609 (2020).
32. A. N. Pilyasov, V. A. Tsukerman, Patterns and features of the formation of new oil and gas provinces in
the Russian Arctic (on the Example of the Nenets Autonomous Okrug). IOP Conf. Ser. Earth Environ. Sci.
459, 042081, (2020).
33. T. Nyseth, “Arctic urbanization: Modernity without cities” in Arctic Environmental Modernities:
From the Age of Polar Exploration to the Era of the Anthropocene, L.- A. Körber, S. MacKenzie,
A. Westerståhl Stenport, Eds. (Springer International Publishing, 2017), pp. 59–70.
34. C. D. Elvidge et al., “Preliminary results from nighttime lights change detection” in Proceedings of
the International Society of Photogrammetry and Remote Sensing, M. Moeller, E. Wentz, Eds. (Joint
Symposia Urban, Tempe, AZ, 2005), pp. 14–16.
35. P. Gao, T. Wu, Y. Ge, G. Yang, Y. Lu, Correcting the nighttime lighting data underestimation effect
based on light source detection and luminance reconstruction. Int. J. Appl. Earth Obs. Geoinf. 121,
103380 (2023).
36. D. Appenroth, G. C. Wagner, D. G. Hazlerigg, A. C. West, Evidence for circadian- based photoperiodic
timekeeping in Svalbard ptarmigan, the northernmost resident bird. Curr. Biol. 31, 2720–2727.e5
(2021).
37. R. A. E. Fosbury, G. Jeffery, Reindeer eyes seasonally adapt to ozone- blue Arctic twilight by tuning a
photonic tapetum lucidum. Proc. R. Soc. B Biol. Sci. 289, 20221002 (2022).
38. Y. Yang et al., Species sensitivities to artificial light at night: A phylogenetically controlled multilevel
meta- analysis on melatonin suppression. Econ. Lett. 27, e14387 (2024).
39. L. Meng et al., Artificial light at night: An underappreciated effect on phenology of deciduous woody
plants. PNAS Nexus 1, 1–10 (2022).
40. T. Barry, C. Price, Arctic biodiversity: From science to policy. J. Environ. Stud. Sci. 5, 283–287 (2015).
41. Y. A. Frank et al., Cleaning of oil- polluted bottom sediments of the boreal lake, Samotlor oil field,
North Russia: Case report. Water Sci. Technol. 82, 3062–3073 (2020).
42. M. Huitric et al., “What factors build or erode resilience in the Arctic?” in Arctic Resilience Report,
M. Carson, G. Perterson, Eds. (Stockholm Environment Institute and Stockholm Resilience Centre,
Stockholm, Sweden, 2016), pp. 96–125.
43. D. Coady, I. Parry, N.- P. Le, B. Shang, Global fossil fuel subsidies remain large: An update based on
country- level estimates. IMF Work. Pap. 19, 1 (2019).
44. V. Kalb, B. Kosar, Y. Collado- Vega, C. Davidson, Aurora detection from nighttime lights for earth and
space science applications. Earth Space Sci. 10, 1–16 (2023).
45. M. M. Bennett, L. C. Smith, Advances in using multitemporal night- time lights satellite imagery to
detect, estimate, and monitor socioeconomic dynamics. Remote Sens. Environ. 192, 176–197 (2017).
46. X. Li, Y. Zhou, M. Zhao, X. Zhao, A harmonized global nighttime light dataset 1992–2018. Sci. Data 7,
168 (2020).
47. C. D. Elvidge et al., A fifteen year record of global natural gas flaring derived from satellite data.
Energies 2, 595–622 (2009).
48. M. Zhizhin et al., Measuring gas flaring in Russia with multispectral VIIRS nightfire. Remote Sens.
13, 3078 (2021).
49. M. Schiavina, S. Freire, A. Carioli, K. MacManus, GHS- POP R2023A - GHS population
grid multitemporal (1975–2030). European Commission, Joint Research Centre (2023),
10.2905/2FF68A52-5B5B-4A22-8F40-C41DA8332CFE.
50. M. Kenny, “Urban planning in the Arctic: Historic uses & the potential for a resilient urban future”
in The Arctic Yearbook 2017, L. Heininen, H. Exner- Pirot, J. Plouffe, Eds. (Northern Research Forum,
2017), pp. 133–146.
51. GADM, Global Administrative Areas (version 4.1, University of California, Berkley [digital geospatial
data], 2022). http://www.gadm.org. Accessed 10 March 2023.
52. V. Maus et al., A global- scale data set of mining areas. Sci. Data 7, 1–13 (2020).
53. GOGET, Global Oil and Gas Extraction Tracker, Global Energy Monitor [digital geospatial data]. https://
globalenergymonitor.org/projects/global- oil- gas- extraction- tracker/. Accessed 10 October 2023.
54. GCMT, Global Coal Mine Tracker, Global Energy Monitor [digital geospatial data]. https://
globalenergymonitor.org/projects/global- coal- mine- tracker/. Accessed 10 October 2023.
55. E. Plekhanova, ALAN_Arctic. GitHub. https://github.com/PlekhanovaElena/ALAN_Arctic. Accessed
2October 2024.
Downloaded from https://www.pnas.org by Cengiz Akandil on October 31, 2024 from IP address 89.206.81.74.