ArticlePDF Available

Quantifying tourism booms and the increasing footprint in the Arctic with social media data

PLOS
PLOS One
Authors:

Abstract and Figures

Arctic tourism has rapidly increased in the past two decades. We used social media data to examine localized tourism booms and quantify the spatial expansion of the Arctic tourism footprint. We extracted geotagged locations from over 800,000 photos on Flickr and mapped these across space and time. We critically examine the use of social media as a data source in data-poor regions, and find that while social media data is not suitable as an early warning system of tourism growth in less visited parts of the world, it can be used to map changes at large spatial scales. Our results show that the footprint of summer tourism quadrupled and winter tourism increased by over 600% between 2006 and 2016, although large areas of the Arctic remain untouched by tourism. This rapid increase in the tourism footprint raises concerns about the impacts and sustainability of tourism on Arctic ecosystems and communities. This boom is set to continue, as new parts of the Arctic are being opened to tourism by melting sea ice, new airports and continued promotion of the Arctic as a ‘last chance to see’ destination. Arctic societies face complex decisions about whether this ongoing growth is socially and environmentally sustainable.
This content is subject to copyright.
RESEARCH ARTICLE
Quantifying tourism booms and the
increasing footprint in the Arctic with social
media data
Claire A. RungeID
1
*, Remi M. Daigle
2
, Vera H. Hausner
1
1Arctic Sustainability Lab, Department of Arctic and Marine Biology, UiT The Arctic University of Norway,
Tromsø, Norway, 2De
´partement de biologie, ’Universite
´Laval, Que
´bec, Canada
*claire.runge@uqconnect.edu.au,twitter@Claire_Runge
Abstract
Arctic tourism has rapidly increased in the past two decades. We used social media data to
examine localized tourism booms and quantify the spatial expansion of the Arctic tourism
footprint. We extracted geotagged locations from over 800,000 photos on Flickr and
mapped these across space and time. We critically examine the use of social media as a
data source in data-poor regions, and find that while social media data is not suitable as an
early warning system of tourism growth in less visited parts of the world, it can be used to
map changes at large spatial scales. Our results show that the footprint of summer tourism
quadrupled and winter tourism increased by over 600% between 2006 and 2016, although
large areas of the Arctic remain untouched by tourism. This rapid increase in the tourism
footprint raises concerns about the impacts and sustainability of tourism on Arctic ecosys-
tems and communities. This boom is set to continue, as new parts of the Arctic are being
opened to tourism by melting sea ice, new airports and continued promotion of the Arctic as
a ‘last chance to see’ destination. Arctic societies face complex decisions about whether this
ongoing growth is socially and environmentally sustainable.
Introduction
The number of tourists visiting the Arctic has increased dramatically over the past two decades
[1,2], reflecting a rise in tourism globally over the past 50 years [3]. While this could bring
alternative livelihoods to Arctic communities, concerns over the social and environmental sus-
tainability of the rate and scale of the tourism boom are growing across the Arctic [46]. Tour-
ism can have both positive and negative impacts on the natural environment and on host
communities. Direct effects of tourism (e.g. transporting, accommodating, and feeding tour-
ists) and the indirect socioeconomic change brought about by the tourism industry (e.g. influx
of seasonal workers) drive increased habitat loss [7], resource use, and carbon emissions [8]
across the world, in addition to the localized consequences of nature-based tourism and recre-
ation activities on the natural environment, such as injury, death, or disturbance of wildlife or
damage to vegetation [9]. Understanding, at a local scale, how and where tourism booms are
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 1 / 14
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Runge CA, Daigle RM, Hausner VH
(2020) Quantifying tourism booms and the
increasing footprint in the Arctic with social media
data. PLoS ONE 15(1): e0227189. https://doi.org/
10.1371/journal.pone.0227189
Editor: Wenwu Tang, University of North Carolina
at Charlotte, UNITED STATES
Received: June 14, 2019
Accepted: December 9, 2019
Published: January 16, 2020
Peer Review History: PLOS recognizes the
benefits of transparency in the peer review
process; therefore, we enable the publication of
all of the content of peer review and author
responses alongside final, published articles. The
editorial history of this article is available here:
https://doi.org/10.1371/journal.pone.0227189
Copyright: ©2020 Runge et al. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: The data underlying
the results presented in the study are available
from Flickr (www.flickr.com). To aid decision-
making we have made the code, rasters and maps
of tourism intensity across the Arctic and tables of
distributed across landscapes is crucial for conserving Arctic environments and for managing
impacts on host communities.
A major challenge for planning and managing sustainable tourism growth in the Arctic lies
in the difficulties of mapping where tourists go and how they use landscapes and ecosystems.
Data on spatial visitation and trends is sparse in the Arctic. While statistics on hotel stays and
transport use are now commonly collected by government and tourism management organi-
zations, data on where tourists go during the day and what they do there is rarely collected.
Social media provides a useful source of information on tourist visitation patterns to better
pinpoint the needs of tourists and target actions to manage tourism impacts [10]. Passively
crowdsourced and high resolution information from social media (volunteered geographic
information; VGI) can be used to rapidly generate maps of the multiple destinations visited by
tourists across large areas and over time [11]. Social media data has been well demonstrated to
be useful for mapping and monitoring at a range of spatial scales across the world. Such data
has been used to map the distribution of tourists in protected areas [1214], and within cities
[11] and can be used to inform a variety of aspects of tourism and landscape management,
including to identify peaks of visitation to attractions [15,16], map environmental impacts [17]
and to estimate the landscape values, nature-based experiences and ecosystem services appre-
ciated by tourists [1820]. Most of these studies use the social media platform, Flickr, which
has been shown to correlate well with visitor statistics at different scales [12,15,16,21].
Similar to many nature-based tourism destinations in developing countries, parts of the
Arctic in Scandinavia, Iceland, Faroe Islands and Alaska have experienced unprecedented
growth in the number of tourists in recent years, and Greenland, Canada and the Russian Arc-
tic are likely to be the next tourism frontiers [22,23]. Rapid and localized booms in tourism
can overwhelm local capacity (and desire) to host visitors, particularly in small, remote com-
munities such as those found in the Arctic and many parts of the developing world [2326].
The ability to rapidly identify sites in the early stages of a boom would support better adapta-
tion and planning to pre-emptively address many of the sustainability challenges brought
about by rapid increases in nature-based tourism. It would allow local communities and
national tourism organizations to proactively direct resources to sites where special manage-
ment such as provisioning of restrooms and waste disposal, better signage, safety and disaster
management, parking, and trail maintenance may be required. Social media data has been
demonstrated as an early-warning system for rapidly detecting booms and busts in such
diverse applications as disaster management [27,28] and disease control [29]. Methods for
such ‘event detection’ are rapidly evolving [30,31]. These methods rely on large amounts of
high temporal resolution data (‘big data’). The suitability and limitations of social media data
for detecting events in data-sparse regions has yet to be tested. We investigate whether Flickr
data can be used as an early warning system to detect localized booms in tourism in the Arctic
and similar data-deficient regions.
Tourism growth can influence the spatial use of landscapes in various ways. Here, we dem-
onstrate how the tourist footprint on Arctic landscapes has changed over the past 14 years at a
pan-Arctic scale, and examine the management implications of those changes. We test two
hypotheses drawn from theories of tourism and economic geography [32,33] 1) that tourists
visit the same sites through time (overall footprint is unchanged but magnitude of use at each
site has increased) 2) that tourists have spread throughout the landscape (overall footprint
increased but magnitude of use at any one site is constant). These two patterns of tourism
growth have very different implications for both the social and environmental sustainability of
tourism and the satisfaction of the tourist experience. For instance, if tourists avoid busy areas
and self-segregate across landscapes [34], environmental and social impacts will be more wide-
spread but lower intensity. Alternately, the negative impacts of tourism can be localized by
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 2 / 14
the footprint across time publicly and freely
available for download at doi:10.18710/QEOFPY.
Funding: The work was funded by grants awarded
to VHH from FRAM - High North Research Centre
for Climate and the Environment through the
Flagship MIKON (Project RConnected; https://
www.framcentre.com/) and the Arctic Belmont
Forum Arctic Observing and Research for
Sustainability (Project CONNECT; https://www.
belmontforum.org/). The Norwegian collaboration
was financed by Norwegian Research Council
grant 247474 (https://www.forskningsradet.no/en).
The funders had no role in study design, data
collection and analysis, decision to publish, or
preparation of the manuscript.
Competing interests: The authors have declared
that no competing interests exist.
channeling tourists into high use sites and away from sensitive communities and ecosystems.
We examine seasonal differences in the spatial patterns of Arctic visitation, and explore how
infrastructure such as roads, airports and ports influences these patterns, with a view to
informing how upcoming infrastructure development could contribute to tourist growth.
Materials and methods
Extraction of data from Flickr
We first extracted geotagged and publicly shared photo metadata for over 2 million photos
from Flickr (www.flickr.com) for the region north of latitude 60˚N. Photo metadata included
location and date that each photo was taken, user id (key coded by Flickr), image URL, Flickr-
and user- generated image tags, and user-generated image title. Data was extracted from the
Flickr API (https://www.flickr.com/services/api/) on 4 December 2017. Due to an issue with
the data download we re-extracted photos for Iceland (bounded by -27˚ to -12˚ longitude and
62˚ to 68˚ latitude) on 11 January 2018. Hourly metrics of the number of photos uploaded to
Flickr globally between to January 1 2004 and December 31 2017 were obtained from the
Flickr API on 07 May 2018. We used the R package ‘flickRgeo’ [35] which provides an R wrap-
per for the Flickr API.
For the purposes of this study we define our study region, ‘Arctic’, as the region bounded
by the Arctic Council AMAP boundaries [36] and confined to areas north of latitude 60˚N.
We constrain the study region using an environmental rather than political definition as the
study is focused primarily on impacts on Arctic landscapes. We excluded photos from the
extracted dataset that were taken outside this study region. We also excluded photos that were
missing urls or geotag coordinates, had null coordinates (0,0), and photos taken prior to Janu-
ary 1 2004, or after December 31 2017. We excluded photos by users with only 1 or 2 photos
within the study region as they are likely to represent people who are just trialing Flickr by
uploading a random photo rather than a photo representing a genuine tourist. These ‘test
users’ account for approximately 36% of users in the Flickr dataset but just 0.95% of photos.
Further details on the choice of 2 photos as a threshold for exclusion are included in S1 Appen-
dix. The final dataset contained a total of 805,684 geotagged photos with metadata from 13,596
unique users.
Mapping the seasonal distribution of Arctic tourism
To map the relative intensity of visitation across the Arctic in summer and winter, we first cre-
ated square spatial grids (rasters) at 10km and 100km resolutions. We then calculated the
photo-unit-days in each grid cell for summer and for winter aggregating data for each season
across all years (2004–2017). We defined the months of May to October as ‘summer’ and
November through April (of the following year) as ‘winter’ (e.g. “winter 2016” includes the
months November and December in 2016 and January through April in 2017). Photo-unit-
days (PUD) is an established metric of tourism visitation [15] that accounts for the biases in
social media data introduced by differences in the number of photographs uploaded by differ-
ent users. For example, three PUD can represent either a site visited on three separate days
throughout the year by a single person, or by three separate people on a single day. This is the
conventional approach for working with this type of social media data because it corresponds
to empirical user data collected by tourism sites that are often based on daily user access fees.
For example, if three users access a park with daily user access fees on the same day, or a single
user accesses that park on 3 separate days, both visitation scenarios appear identical in terms
of empirical visitation rates (i.e. fees collected) as well as PUD.
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 3 / 14
Pan-Arctic trends in the footprint of Arctic tourism over time
To quantify the ‘footprint’ of Arctic tourism (the percentage of the Arctic visited by tourists),
we created a hexagonal spatial grid covering the entire Arctic with a 5km diameter resolution.
We chose this resolution after examining the trade-off between commission errors and com-
puting efficiency (S2 Appendix). We allocated each cell a value of 1 (visited) if any user had
taken a photograph in that cell in a given year (2004–2017) and season (summer, winter), and
0 (unvisited) if not. The number of people using Flickr globally changed over time as the popu-
larity of the platform waxed and waned. Relying on raw metrics of the number of Arctic Flickr
users, photographs or photo-unit-days will thus result in biased estimates of tourism trends
across time. The global and Arctic trends in Flickr use across time and the number of photos
sampled in each year can be found in S3 Appendix. We calculated the footprint in three ways.
Firstly, the Uncorrected Arctic footprint uses all available Flickr records (566205 summer;
228667 winter) and shows the full extent of Flickr users’ footprint, but includes the bias from
the global change in Flickr usage between 2004 and 2017. The Global-bias corrected subsample
removes the global pattern of Flickr usage to represent a less source-biased view of the rise in
the footprint of tourism in the Arctic. This was done by selecting a random sample of Arctic
photos based on the change in number of photos submitted to Flickr globally using 2004 values
as a baseline. For example, the year with the lowest global usage (2004), we kept all the photo
records. If the global Flickr usage doubled in a particular year relative to 2004, we sampled half
of the available records for that particular year. The number of records sampled for each
month can be found in Table S3 Appendix (total 36546 summer records, 12262 winter).
Finally, the Equal sample size also removes the global bias and provides a measure of the
change in the relative footprint per-tourist across time. This was done by randomly selecting
an equal number of photographs for each year from which to calculate the footprint (total
15652 records summer, 4018 records winter). Numbers for 2017 should be treated with cau-
tion and are likely underestimates as we extracted data from the Flickr API on 4 December
2017 and the average lag time between photos being taken and uploaded is 2 weeks, but can be
longer [13].
Modelling the influence of accessibility on seasonal patterns of Arctic
landscape use
In order identify the effect of accessibility on the presence of tourists in different seasons, we
first divided the study region into a hexagonal spatial grid with a 10km diameter resolution.
We chose to use a lower resolution than that used in the footprint analysis above as the large
number of cells at 5km made the models computationally intractable. We then calculated foot-
print in each cell for summer and for winter by allocating these cells a value of 1 if any photo-
graph had been taken in that cell in that season (at any time from 2004 and 2017), and 0 if not,
similar to the Uncorrected Arctic footprint described above. We removed cells that fell within
Russia due to sparse coverage of this region by Flickr data, and from marine areas. We then
modelled the footprint against variables describing the accessibility of each cell: log of distance
to airports, log distance to ports, log distance to populated areas, log distance to road, the
square root of the length of road within a grid cell, whether the cell overlapped any protected
area, and country as a fixed effect. We chose these accessibility variables after examining a
wider set of candidate variables for correlation. The models took the form of a binomial gener-
alized additive model with logit link of footprint as a response variable. In order to account for
spatial autocorrelation we included a fitted thin-plate spline on the variables latitude and longi-
tude of each cell centroid. Smooths, intercept, slope and confidence intervals were estimated
by a restricted maximum likelihood (REML) estimator and methods for large datasets (‘bam’
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 4 / 14
function in R’s mgcv package). The model formula is:
ln Pfootprint
1Pfootprint
!
¼b0þf1ðlongÞ þ f2ðlatÞ þ b1log10 ðairportdistÞ þ b2log10 ðportdistÞ
þb3log10ðpopulateddistÞ þ b4log10 ðroaddist Þ þ b5ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
roadlength
qþb6PA
þg7Country þε
where f
1
,f
2
are fitted thin-plate spline smooth functions on the variables latitude and longitude
of each cell centroid, airport
dist
is the distance to airports, port
dist
is the distance to ports, popu-
lated
dist
is the distance to populated areas, road
dist
is the distance to road, road
length
is the length
of road within a grid cell, and PA is whether the cell overlapped any protected area (0 if false, 1
if true). Random effects are represented by β’s and the fixed effect is represented by .
The location of airports, ports and populated areas, and country boundaries were extracted
from Natural Earth (www.naturalearthdata.com) using the R package ‘rnaturalearth’ [37].
Roads were extracted from Global Roads Inventory Project [38]. Protected area boundaries
were drawn from CAFF [39] and supplemented with data from Protected Planet [40]. We ran
separate models for summer and winter as the presence of snow and ice limits access to rural
areas in winter. The summer model had 90,750 cells with no photos, and 6,554 with photos.
The winter model had 93,990 cells with no photos, and 3,314 with photos.
Local trends (booms and busts) in Arctic tourism
We investigated the suitability of Flickr data to be used to identify local booms and busts in
tourism in the Arctic, a data deficient region. We first divided the landscape into a square grid
cells and calculated the photo-unit-days in each cell for each year. We then performed linear
regression modelling to identify trends in PUD in each cell between 2012 and 2017 and test
their statistical significance. We ran models for cell diameters ranging from 500m to 100km to
examine the effect of data availability on trend detection. We modelled only cells that were vis-
ited in at least two of the six years between 2012 and 2017. This timeframe was chosen as global
Flickr usage remained reasonably constant during this period (Fig A in S3 Appendix).
Unless otherwise stated, all analysis was conducted in R version 3.4.2 [41] using the ‘tidy-
verse’ [42], ‘sf’ [43], and ‘mgcv’ [44] packages. All spatial data was projected to North Pole Azi-
muthal Lambert equal area (EPSG:102017) for analysis. Code associated with the project is
available at doi:10.18710/QEOFPY.
Results
Increase in Arctic summer and winter tourism footprint
We find that the overall footprint of tourism and area used per tourist has increased since 2004
(Fig 1). In the 10 years between 2006 and 2016, the uncorrected footprint increased from
0.066% to 0.385% of the Arctic in summer and 0.015% to 0.173% in winter. After correcting
these figures to account for the increased proportion of tourists captured in the analysis over
that time (i.e. accounting for the rise in Flickr use globally), the footprint increased by 374% in
summer (0.029% of Arctic in 2006, 0.109% in 2016) and 634% in winter (0.006% of Arctic in
2006, 0.036% in 2016). The relative-footprint-per-tourist also increased over that period (sum-
mer: 0.028% of Arctic in 2006 to 0.043% in 2016; winter: 0.007% in 2006 to 0.012% in 2016).
We caution that this metric is not the absolute per-tourist footprint, rather it should be inter-
preted as a relative indication of how the footprint of a fixed (and unquantified) number of
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 5 / 14
tourists has changed across time. These estimates are robust to uncertainty introduced by ran-
dom sampling in the Global-bias corrected and Equal sample size methods (S4 Appendix).
Growth in Arctic visitation over time
The total number of photos on Flickr increased between 2004 and 2017, both globally and
in the Arctic. Global uploads of photos to Flickr steadily increased between 2004 and 2008,
slowed between 2008 and 2012 before a slight upsurge in 2012, plateaued between 2013 and
2015 and has declined slightly since then (Fig A in S3 Appendix). The Arctic represents an
increasing share of Flickr’s yearly photo traffic (Table in S3 Appendix). The number of pho-
tos uploaded in the Arctic shows an exponential growth between 2004 and 2013, and has
remained roughly steady since then (Fig B in S3 Appendix), with this trend overlaid on a
seasonal trend in visitation (Fig C in S3 Appendix). Across the Arctic, July and August were
the most popular months to photograph the Arctic (Fig D in S3 Appendix). Nonetheless, a
large number of users visited during the Arctic winter (Table 1; 29.8% of all photos; 40,272
photo-unit-days, 53.3% of summer PUD). At the extreme, visitation in Greenland is con-
centrated in the summer months (91% of photos). In contrast to the rest of the Arctic, visita-
tion in northern Finland peaks in winter (61.3% of photos). Winter and Christmas are a key
part of the branding of tourism to northern Finland, and places such as ‘Santa Claus’s
Fig 1. Footprint of Arctic tourism. The total footprint of Arctic tourism measured from Flickr data increased between 2004 and 2017 (Uncorrected Arctic footprint,
darker blue), even after adjusting for the global rise in Flickr use during this period (Global-bias corrected, green). The relative footprint per tourist (Equal sample size,
pale blue) also increased slightly over this time. Similar trends are seen in summer and winter, though the tourism footprint in winter is approximately half the
magnitude of that in summer. The footprint is defined as the percentage of 5km hexagonal grid cells within the Arctic region visited by at least one Flickr user per year.
The 2017 decline should be interpreted with caution as it may in part be an artefact of the timing of our data download and the lag between photos being taken and their
being uploaded to Flickr.
https://doi.org/10.1371/journal.pone.0227189.g001
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 6 / 14
village’ in Rovaniemi, Lappland, attract the majority of the visitors to the region (Fig 2).
These metrics of Arctic visitation derived from Flickr show good agreement with official
metrics of tourism visitation (S1 Table), consistent with previous research on social media
data [12,15,21].
Table 1. Seasonal variation in Arctic visitation, estimated from Flickr data (2004 to 2017).
Overall number of
photos (% of all photos)
Summer (% of all
photos for that region)
Winter (% of all
photos for that
region)
Number of photo-
unit-days (PUD)
Summer (% of total
PUD for that region)
Winter (% of total
PUD for that region)
Whole Arctic 805,684 70.2 29.8 115, 775 65.2 34.8
Iceland 377,817 (46.9) 70.0 30.0 51,500 66.4 33.6
Alaskan Arctic 150,325 (18.7) 76.0 24.0 21,903 67.0 33.0
Norwegian
Arctic
122,387 (15.2) 68.2 31.8 20,779 63.7 36.3
Canadian
Arctic
43,466 (5.4) 68.4 31.6 7,713 66.8 33.2
Finnish Arctic 22,164 (2.8) 38.7 61.3 3,716 43.2 56.8
Faroe Islands 20,259 (2.5) 78.1 21.9 2,226 68.0 32.0
Greenland 19,944 (2.5) 91.0 9.0 1,952 78.8 21.2
Swedish Arctic 18,303 (2.3) 53.5 46.5 2,809 56.9 43.1
Svalbard & Jan
Mayen
16,282 (2.0) 71.3 28.7 1,321 72.6 27.4
Russian Arctic 13,321 (1.7) 64.2 35.8 2,228 60.7 39.3
https://doi.org/10.1371/journal.pone.0227189.t001
Fig 2. Seasonal maps of Arctic tourism (2004–2017) displayed at 10km resolution. The guide maps (right) are displayed at 100km resolution. Spatial patterns of
tourism are strongly governed by air, road and sea access, with few tourists venturing far from populated areas in winter. A photo-unit-day value of 14 corresponds to
one Flickr user visiting the cell per year. Country borders are modified from Natural Earth CC PD.
https://doi.org/10.1371/journal.pone.0227189.g002
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 7 / 14
Flickr data not suitable as early-warning system
We investigated the potential for Flickr data to be used to identify local trends in tourism over
time. The Arctic is a data-poor region and we found that only a small number of places were
photographed more than once per year (Fig in S2 Appendix). Only 22 10km grid cells were vis-
ited by Flickr users more than 50 times (i.e. once a week) in 2017. Although annual tourism
growth is documented as ranging from 5–20% across the Arctic, due to these data limitations
regression models of trends in visitation between 2012 and 2017 were able to detect significant
trends at just a handful of sites (S1 Fig). Most sites lacked sufficient data to detect trends even
when aggregated to 10km grid cells, a relatively large scale compared to the scale required for
management decisions.
Spatial patterns of Arctic visitation change across the seasons
The spatial pattern of tourists’ landscape use differed between summer and winter. Flickr users
ventured further north and into marine areas to a greater extent during the summer months
(Fig 2). Visitation was influenced by access and often, though not always, concentrated in rec-
ognized tourism hotspots (Fig 2). The main hotspots of tourism fall along coastal roads in Ice-
land, in the fjords and islands of northern Norway, and in protected areas and along roads in
North America (Fig 2). We note that although the size of the Alaskan tourism market eclipses
that of the rest of the Arctic, including Iceland, few cruises travel further north than Anchorage
(61.2˚N), and the majority of this tourism thus falls outside our study region which is bounded
to the south at 60˚N.
Accessibility drives Arctic visitation
We found that accessibility has a significant effect on the distribution of Flickr users
throughout the Arctic, with the presence of tourists decreasing rapidly as distance from
transport infrastructure and populated areas increases, and increasing with the length of
road in a given cell (Table 2). The summer accessibility model explained 47.3% of the devi-
ance in the presence of tourists (adjusted R
2
0.448, AIC 25347). The winter accessibility
model explained 51.4% of the deviance in the presence of tourists (adjusted R
2
= 0.436, AIC
of 14117). The variable square root of length of road in cell had the largest explanatory
power of the accessibility variables (model without this variable had adjusted R
2
0.409 sum-
mer, 0.393 winter, deviance explained 44.7% summer, 48.8% winter). The protected area
term of the winter model was significant (p = 0.000670) in the full model, but not significant
in more parsimonious models. Removal of this term only slightly decreased the deviance
explained (ΔAIC = 9, Δdf = -1.0239, Δdeviance = -10.706, Pr(>Chi) = 0.001117), indicating
that tourists were no more or less likely to visit protected areas than non-protected areas in
winter. All other terms were significant in both models. We removed the variable that had
the least explanatory power, log distance to port, from the summer model, and the winter
model without the protected area term. This reduced the model fit of the summer model
(ΔAIC = 145, Δdf = -1.0881, Δdeviance = -146.57, Pr(>Chi) = 2.066x10
-16
) but had minimal
effect on the winter model (ΔAIC = 25, Δdf = -1.1028, Δdeviance = -27.406), though the
large number of data points in the model meant that this variable was significant at 95%
confidence (2.02x10
-7
). Removal of all other terms substantially decreased the deviance
explained by the models. Visual examination of model residuals did not reveal any residual
spatial autocorrelation in any of the models. Plots of model fit and partial plots of model
variables are included in S5 Appendix.
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 8 / 14
Discussion
The footprint of tourism on the Arctic environment has almost quadrupled over the past
decade, from 0.03% of the Arctic in summer 2006 to 0.11% of the Arctic in summer 2016 (Fig
1), and the winter footprint has increased by over 600%. Despite this dramatic expansion,
large areas of the Arctic still remain free from tourists (Fig 2). Arctic tourists tend to congre-
gate in a handful of highly visited sites (Fig 2) with a long tail of places that are visited only
occasionally, following the power law seen in other parts of the world [45]. The recent tourism
boom across the Arctic has led to widespread concerns over the effects of tourism on Arctic
ecosystems and communities, and the sustainability of Arctic tourism [4,5]. Arctic tourism is
often marketed around ideas of pristine, untouched nature, and the overall growth in tourist
numbers and the concurrent increase in infrastructure to support that growth presents chal-
lenges for maintaining environmental and social sustainability. There are indications that
many popular sites are reaching capacity, and that tourists are beginning to experience disap-
pointment and frustration around the large number of visitors present [46]. The footprint per
tourist has also increased (Fig 1), indicating that tourists are now visiting a wider variety of
places. This may be either due to self-segregation by tourists seeking an undisturbed experi-
ence [46], or the marketing of a wider variety of tourist attractions and nature-based activities
as Arctic tourism matured over the past decade [47].
These patterns of visitation present both advantages for the management of tourists and
challenges for the sharing of the economic benefits of the recent and ongoing Arctic tourism
Table 2. Model coefficients for binomial generalized additive models of the effect of accessibility on the tourist footprint in summer and winter in the Arctic. The
intercept represents Norway, unprotected. The protected area term was not included in the winter model.
Summer Winter
Variable Estimate of coefficient (±standard
error)
z-value Pr(>|z|) Estimate of coefficient (±standard
error)
z-value Pr(>|z|)
(Intercept) 8.720 ±0.619 14.094 <2.00 x
10
16
7.830 ±0.925 8.468 <2.00 x
10
16
Norway
Canada -5.707 ±0.632 -9.035 <2.00 x
10
16
-5.690 ±0.961 -5.923 3.17x 10
9
Finland -0.474 ±0.159 -2.988 0.0028 0.979 ±0.171 5.728 1.01 x 10
8
Faroe Islands 5.802 ±22.647 0.256 0.798 -5.557 ±0.970 -5.731 1.00 x 10
8
Greenland -5.167 ±0.590 -8.757 <2.00 x
10
16
-4.805 ±0.893 -5.381 7.42 x 10
8
Iceland -4.082 ±0.624 -6.540 6.15 x 10
11
-3.712 ±0.914 -4.063 4.85 x 10
5
Svalbard & Jan Mayen -0.059 ±0.267 -0.221 0.825 -1.524 ±0.390 -3.939 8.18 x 10
5
Sweden -0.673 ±0.134 -5.007 5.52 x 10
07
0.022 ±0.150 0.147 0.883
USA (Alaska) -4.682 ±0.630 -7.433 1.06 x 10
13
-5.647 ±0.954 -5.917 3.28 x 10
9
Protected area TRUE 0.841 ±0.042 19.919 <2.00 x
10
16
Square root of length of road in
cell
0.805 ±0.026 31.326 <2.00 x
10
16
0.574 ±0.023 24.717 <2.00 x
10
16
Log distance to road -0.278 ±0.019 -14.912 <2.00 x
10
16
-0.278 ±0.025 -11.217 <2.00 x
10
16
Log distance to airports -0.220 ±0.027 -8.135 4.13 x 10
16
-0.337 ±0.034 -9.855 <2.00 x
10
16
Log distance to ports -0.381 ±0.031 -12.481 <2.00 x
10
16
-0.229 ±0.038 -5.970 2.37 x 10
9
Log distance to populated places -0.532 ±0.022 -24.243 <2.00 x
10
16
-0.603 ±0.028 -21.244 <2.00 x
10
16
https://doi.org/10.1371/journal.pone.0227189.t002
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 9 / 14
boom. Management of tourists and their impacts is often easier where they congregate in a few
small areas as resources can be allocated to these high priority areas bringing economies of
scale. This is particularly important in small, rural communities where both human and finan-
cial capacity to manage tourist impacts can be limited [24]. Though environmental impacts
can be locally high in these well-visited areas, requiring thoughtful management, aggregation
of tourists in a few small areas can reduce the impacts of tourism across the wider landscape
and channel resources into efficient management at the high use sites to sustain tourist satis-
faction despite crowding. Channeling visitors into such ‘sacrificial sites’ may have net benefits
at the landscape scale in places like the Arctic, where wildlife and vegetation are highly sensi-
tive to disturbance and take a long time to recover from human impacts.
The tourism footprint is strongly associated with access, and is particularly dependent on
the presence of roads and airports. This is not surprising, but important to keep in mind when
planning for growth in tourism. The proposed construction of three transatlantic airports in
Greenland has a high potential of boosting tourism in host communities that do not have suffi-
cient capacity to sustainably manage this growth. The sustainability of tourism in Greenland,
and similar developing places around the world depends heavily on building community
capacity that sustains natural resources and local culture, and that protects vulnerable sites and
species from the expansion of tourists into new locations [6]. The spatial extent of the winter
footprint of tourism is about 40% lower than that in summer, with tourists gathering closer to
airports and towns. Managers therefore need to be especially aware of the expansion of tourists
into vulnerable sites and the greater use of protected areas in the summer season. Ports have
only minimal influence on the presence of tourists on land.
While social media data may be useful for rapidly detecting localized booms in tourism in
highly visited regions [48], our analysis indicates that Flickr data is of limited use for identify-
ing local tourism booms in data deficient regions such as in the Arctic. Low rates of visitation
across most of the Arctic combined with the small proportion of visitors that use Flickr [15]
means that just a handful of Arctic locations are visited by Flickr users more than once a
month (Fig C in S1 Fig). One of the few regions where we detected statistically significant
increases in visitation was the Lofoten islands of northern Norway. There, our analysis con-
firmed qualitative trends already noticed by tourism agencies in the region. Twitter has a
higher user base and may be a better source of fine-scale temporal data in the parts of the Arc-
tic with high-speed mobile data coverage [18,49,50]. Changes in ownership of the social media
platforms and changes to data access rules means social media data from other suitable plat-
forms such as Panoramio and Instagram were no longer freely available to academic research-
ers at the time of analysis. Quantitative analysis of social media data requires specialized
computing and technical skills that are not normally available to local tourism management
agencies. Maintaining dialogue between tourism management bodies and local communities
and tour operators remains the most pragmatic way to detect and respond to fine-scale tour-
ism trends in areas where data and technical capacity are limited.
Conclusions
The recent and rapid increase in the footprint of tourists on the Arctic that we document here
is concerning. Upcoming investments in transport infrastructure in places like Greenland and
the promotion of remote areas of the Arctic as tourist destinations, such as Franz Josef in Rus-
sia, will drive a further expansion of the tourist footprint in this unique part of the world. Des-
tinations are also increasingly been marketed as ‘last chance tourism’ attracting visitors to
venture into previously unexplored areas to experience Arctic ecosystems and species at risk of
disappearing [4,51]. For instance, in Hudson Bay in Canada the small community of Churchill
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 10 / 14
where polar bears spend increasingly more time on shore due to climate change, have experi-
enced a rapid influx of tourists [4]. Wildlife viewing of vulnerable species, such as polar bears,
narwhals and beluga whales, is putting additional pressure on species threatened by climate
change [52,53]. Our high resolution and seasonal maps of Arctic tourism allow tourist man-
agement bodies and environmental organizations to pinpoint the places visited by tourists and
the relative magnitude of visitation across the Arctic and to detect landscape-wide trends in
visitation that need to be managed. Strategic and thoughtful assessment of whether this ongo-
ing growth in Arctic tourism is sustainable or desirable for Arctic ecosystems and communi-
ties is urgently required.
Supporting information
S1 Appendix. Sensitivity analysis of photo exclusion threshold.
(PDF)
S2 Appendix. Sensitivity analysis of resolution.
(PDF)
S3 Appendix. Global and Arctic Flickr trends.
(PDF)
S4 Appendix. Uncertainty around footprint estimates.
(PDF)
S5 Appendix. GAM model performance.
(PDF)
S1 Fig. Annual maps of tourism growth.
(PDF)
S1 Table. Comparison with official visitor statistics.
(PDF)
Author Contributions
Conceptualization: Claire A. Runge, Vera H. Hausner.
Data curation: Claire A. Runge.
Formal analysis: Claire A. Runge, Remi M. Daigle.
Funding acquisition: Vera H. Hausner.
Methodology: Claire A. Runge, Remi M. Daigle.
Software: Remi M. Daigle.
Visualization: Claire A. Runge, Remi M. Daigle.
Writing original draft: Claire A. Runge.
Writing review & editing: Claire A. Runge, Remi M. Daigle, Vera H. Hausner.
References
1. Maher PT. Tourism Futures in the Arctic. In: Latola K, Savela H, editors. The Interconnected Arctic—
UArctic Congress 2016. Cham: Springer International Publishing; 2017. pp. 213–220. https://doi.org/
10.1007/978-3-319-57532-2_22
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 11 / 14
2. Fay G, Karlsdo
´ttir A. Social indicators for Arctic tourism: observing trends and assessing data. Polar
Geogr. 2011; 34: 63–86. https://doi.org/10.1080/1088937X.2011.585779
3. Scott D, Go
¨ssling S. What could the next 40 years hold for global tourism? Tour Recreat Res. 2015; 40:
269–285. https://doi.org/10.1080/02508281.2015.1075739
4. Lemelin H, Dawson J, Stewart EJ, Maher P, Lueck M. Last-chance tourism: the boom, doom, and
gloom of visiting vanishing destinations. Curr Issues Tour. 2010; 13: 477–493. https://doi.org/10.1080/
13683500903406367
5. Pashkevich A, Stjernstro
¨m O, Lundmark L. Nature-based tourism, conservation and institutional gover-
nance: a case study from the Russian Arctic. Polar J. 2016; 6: 112–130. https://doi.org/10.1080/
2154896X.2016.1171000
6. Kidd A, Monz CA, Hausner VH, Schmidt JI, Clark D. Nature-based tourism, resource dependency, and
resilience of Arctic communities: Framing complex issues in a changing environment. J Sustain Tour.
Forthcoming; https://doi.org/10.1080/09669582.2019.1612905
7. Brandt JS, Buckley RC. A global systematic review of empirical evidence of ecotourism impacts on for-
ests in biodiversity hotspots. Curr Opin Environ Sustain. 2018; 32: 112–118. https://doi.org/10.1016/j.
cosust.2018.04.004
8. Becken S, Patterson M. Measuring national carbon dioxide emissions from tourism as a key step
towards achieving sustainable tourism. J Sustain Tour. 2006; 14: 323–338. https://doi.org/10.2167/
jost547.0
9. Tolvanen A, Kangas K. Tourism, biodiversity and protected areas–review from northern Fennoscandia.
J Environ Manage. 2016; 169: 58–66. https://doi.org/10.1016/j.jenvman.2015.12.011 PMID: 26720330
10. Leung D, Law R, Hoof H van, Buhalis D. Social media in tourism and hospitality: a literature review. J
Travel Tour Mark. 2013; 30: 3–22. https://doi.org/10.1080/10548408.2013.750919
11. Garcı
´a-Palomares JC, Gutie
´rrez J,
´nguez C. Identification of tourist hot spots based on social net-
works: A comparative analysis of European metropolises using photo-sharing services and GIS.Appl
Geogr. 2015; 63: 408–417. https://doi.org/10.1016/j.apgeog.2015.08.002
12. Levin N, Lechner AM, Brown G. An evaluation of crowdsourced information for assessing the visitation
and perceived importance of protected areas. Appl Geogr. 2017; 79: 115–126. https://doi.org/10.1016/j.
apgeog.2016.12.009
13. Walden-Schreiner C, Rossi SD, Barros A, Pickering C, Leung Y-F. Using crowd-sourced photos to
assess seasonal patterns of visitor use in mountain-protected areas. Ambio. 2018; 1–13. https://doi.
org/10.1007/s13280-018-1020-4 PMID: 29435731
14. Kim Y, Kim C, Lee DK, Lee H, Andrada RIIT. Quantifying nature-based tourism in protected areas in
developing countries by using social big data. Tour Manag. 2019; 72: 249–256. https://doi.org/10.1016/
j.tourman.2018.12.005
15. Wood SA, Guerry AD, Silver JM, Lacayo M. Using social media to quantify nature-based tourism and
recreation. Sci Rep. 2013; 3: 2976. https://doi.org/10.1038/srep02976 PMID: 24131963
16. Mancini F, Coghill GM, Lusseau D. Using social media to quantify spatial and temporal dynamics of
nature-based recreational activities. PLOS ONE. 2018; 13: e0200565. https://doi.org/10.1371/journal.
pone.0200565 PMID: 30011294
17. Hale BW. Mapping potential environmental impacts from tourists using data from social media: a case
study in the Westfjords of Iceland. Environ Manage. 2018; 62: 446–457. https://doi.org/10.1007/
s00267-018-1056-z PMID: 29736767
18. Zanten BT van Berkel DBV, Meentemeyer RK Smith JW, Tieskens KF Verburg PH. Continental-scale
quantification of landscape values using social media data. Proc Natl Acad Sci. 2016; 201614158.
https://doi.org/10.1073/pnas.1614158113 PMID: 27799537
19. Oteros-Rozas E, Martı
´n-Lo
´pez B, Fagerholm N, Bieling C, Plieninger T. Using social media photos to
explore the relation between cultural ecosystem services and landscape features across five European
sites. Ecol Indic. 2017; https://doi.org/10.1016/j.ecolind.2017.02.009
20. Hausmann A, Toivonen T, Slotow R, Tenkanen H, Moilanen A, Heikinheimo V, et al. Social media data
can be used to understand tourists’ preferences for nature-based experiences in protected areas. Con-
serv Lett. 2018; 11: e12343. https://doi.org/10.1111/conl.12343
21. Barchiesi D, Moat HS, Alis C, Bishop S, Preis T. quantifying international travel flows using Flickr.
PLOS ONE. 2015; 10: e0128470. https://doi.org/10.1371/journal.pone.0128470 PMID: 26147500
22. Stewart EJ, Draper D. Sustainable cruise tourism in Arctic Canada: an integrated coastal management
approach. J Tour Manag. 2006; 3: 77–88. https://doi.org/10.3727/154427306779435210
23. Lamers M, Pashkevich A. Short-circuiting cruise tourism practices along the Russian Barents Sea
coast? The case of Arkhangelsk. Curr Issues Tour. 2015; 21: 440–454. https://doi.org/10.1080/
13683500.2015.1092947
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 12 / 14
24. Amundsen H. Illusions of Resilience? An analysis of community responses to change in Northern Nor-
way. Ecol Soc. 2012; 17. Available: https://www.jstor.org/stable/26269221
25. Lasserre F, Têtu P-L. The cruise tourism industry in the Canadian Arctic: analysis of activities and per-
ceptions of cruise ship operators. Polar Rec. 2015; 51: 24–38. https://doi.org/10.1017/
S0032247413000508
26. Stewart EJ, Dawson J, Draper D. Cruise tourism and residents in Arctic Canada: development of a resi-
dent attitude typology. J Hosp Tour Manag. 2011; 18: 95–106. https://doi.org/10.1375/jhtm.18.1.95
27. Panteras G, Wise S, Lu X, Croitoru A, Crooks A, Stefanidis A. triangulating social multimedia content
for event localization using Flickr and Twitter. Trans GIS. 2014; 19: 694–715. https://doi.org/10.1111/
tgis.12122
28. Albuquerque JP de, Herfort B, Brenning A, Zipf A. A geographic approach for combining social media
and authoritative data towards identifying useful information for disaster management. Int J Geogr Inf
Sci. 2015; 29: 667–689. https://doi.org/10.1080/13658816.2014.996567
29. Gao Y, Wang S, Padmanabhan A, Yin J, Cao G. Mapping spatiotemporal patterns of events using
social media: a case study of influenza trends. Int J Geogr Inf Sci. 2018; 32: 425–449. https://doi.org/10.
1080/13658816.2017.1406943
30. Helwig NE, Gao Y, Wang S, Ma P. Analyzing spatiotemporal trends in social media data via smoothing
spline analysis of variance. Spat Stat. 2015; 14: 491–504. https://doi.org/10.1016/j.spasta.2015.09.002
31. Goswami A, Kumar A. A survey of event detection techniques in online social networks. Soc Netw Anal
Min. 2016; 6: 107. https://doi.org/10.1007/s13278-016-0414-1
32. Oppermann M. Tourism space in developing countries. Ann Tour Res. 1993; 20: 535–556. https://doi.
org/10.1016/0160-7383(93)90008-Q
33. Ma M, Hassink R. An evolutionary perspective on tourism area development. Ann Tour Res. 2013; 41:
89–109. https://doi.org/10.1016/j.annals.2012.12.004
34. Butler RW. The concept of a tourist area cycle of evolution: implications for management of resources.
Can Geogr Ge
´ographe Can. 1980; 24: 5–12. https://doi.org/10.1111/j.1541-0064.1980.tb00970.x
35. Daigle R, Dunnington D. flickRgeotag v0.0.1. Zenodo. http://doi.org/10.5281/zenodo.1314717 [Inter-
net]. 2018. Available: https://github.com/remi-daigle/flickRgeotag/tree/v0.0.1
36. Arctic Assessment and Monitoring Programme (AMAP). Boundary for AMAP working group of the Arc-
tic Council [Internet]. 2017. Available: http://geo.abds.is/geonetwork/srv/eng/catalog.search#/
metadata/9BB31CE7-9AE2-453D-9234-0BAEB5C8A33C
37. South A. rnaturalearth: world map data from Natural Earth. R package version 0.1.0. [Internet]. 2017.
Available: https://CRAN.R-project.org/package=rnaturalearth
38. Meijer JR, Huijbregts MAJ, Schotten KCGJ, Schipper AM. Global patterns of current and future road
infrastructure. Environ Res Lett. 2018; 13: 064006. https://doi.org/10.1088/1748-9326/aabd42
39. Conservation of Arctic Flora and Fauna (CAFF). Arctic protected areas– 2017 [Internet]. 2017. Avail-
able: http://geo.abds.is/geonetwork/srv/eng/catalog.search#/metadata/2e56ee1f-50a9-4983-88f4-
edaa8588950d
40. UNEP-WCMC and IUCN. Protected Planet: The World Database of Protected Areas [Internet]. 2018.
Available: www.protectedplanet.net
41. R Core Team. R: A language and environment for statistical computing. R Foundation for Statistical
Computing, Vienna, Austria. [Internet]. 2017. Available: http://www.R-project.org/
42. Wickam H. tidyverse: easily install and load the “tidyverse”. R package version 1.2.1. [Internet]. 2017.
Available: https://CRAN.R-project.org/package=tidyverse
43. Pebesma E. Simple Features for R: Standardized support for spatial vector data. R J. 2018; Available:
https://journal.r-project.org/archive/2018/RJ-2018-009/
44. Wood SN. Fast stable restricted maximum likelihood and marginal likelihood estimation of semipara-
metric generalized linear models. J R Stat Soc B. 2011; 73: 3–36.
45. Schlieder C, Matyas C. photographing a city: an analysis of place concepts based on spatial choices.
Spat Cogn Comput. 2009; 9: 212–228. https://doi.org/10.1080/13875860903121848
46. Csa
´goly Z, þo
´rsdo
´ttir AD, O
´lafsdo
´ttir R. Tourism changing the edge of the wild. J Outdoor Recreat
Tour. 2017; 17: 1–8. https://doi.org/10.1016/j.jort.2016.10.004
47. Fredman P, Tyrva
¨inen L. Frontiers in nature-based tourism. Scand J Hosp Tour. 2010; 10: 177–189.
https://doi.org/10.1080/15022250.2010.502365
48. Girardin F, Fiore FD, Ratti C, Blat J. Leveraging explicitly disclosed location information to understand
tourist dynamics: a case study. J Locat Based Serv. 2008; 2: 41–56. https://doi.org/10.1080/
17489720802261138
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 13 / 14
49. Li L, Goodchild MF, Xu B. Spatial, temporal, and socioeconomic patterns in the use of Twitter and Flickr.
Cartogr Geogr Inf Sci. 2013; 40: 61–77. https://doi.org/10.1080/15230406.2013.777139
50. Tenkanen H, Di Minin E, Heikinheimo V, Hausmann A, Herbst M, Kajala L, et al. Instagram, Flickr, or
Twitter: Assessing the usability of social media data for visitor monitoring in protected areas. Sci Rep.
2017; 7. https://doi.org/10.1038/s41598-017-18007-4 PMID: 29242619
51. Groulx M, Lemieux C, Dawson J, Stewart E, Yudina O. Motivations to engage in last chance tourism in
the Churchill Wildlife Management Area and Wapusk National Park: the role of place identity and nature
relatedness. J Sustain Tour. 2016; 24: 1523–1540. https://doi.org/10.1080/09669582.2015.1134556
52. Atwood TC, Marcot BG, Douglas DC, Amstrup SC, Rode KD, Durner GM, et al. Forecasting the relative
influence of environmental and anthropogenic stressors on polar bears. Ecosphere. 2016; 7: e01370.
https://doi.org/10.1002/ecs2.1370
53. Halliday WD, Têtu P-L, Dawson J, Insley SJ, Hilliard RC. Tourist vessel traffic in important whale areas
in the western Canadian Arctic: Risks and possible management solutions. Mar Policy. 2018; 97: 72–
81. https://doi.org/10.1016/j.marpol.2018.08.035
Quantifying the Arctic tourism footprint with social media data
PLOS ONE | https://doi.org/10.1371/journal.pone.0227189 January 16, 2020 14 / 14
... Empirical studies have been published that analyze Arctic coverage in traditional media (Gritsensko, 2012;Pincus and Ali, 2016;Landriault, 2020), yet comprehensive assessment of social media coverage has escaped scholarly investigation. Rather, studies linking social media and the Arctic focused on specific phenomena taking place in the region, for example, tourism, Inuit activism, and development (Castleton, 2018;Spence, 2019;Runge et al, 2020), without looking at the overall discussion about the Arctic on social media. This article will remedy this knowledge gap and focus on how the Arctic region has been framed and which topics were salient on one social media platform: Twitter/X. ...
... For Castleton (2018), Facebook has played a role in expressing Inuit identity, while for Rodgers and Scobie (2015) and Wachowich and Scobie (2010) social media have allowed Inuit activists to defend traditional practices, for example with the use of the #sealfie. Others, such as Aldao and Mihalic (2020) and Runge, et al. (2020), used social media data as a data source to document Arctic tourism trends and patterns. ...
Article
Full-text available
Social media use has grown in popularity in recent years, becoming a primary source of information for many. Several scholarly inquiries have analyzed how the Arctic region has been portrayed in traditional media. However, no study has comprehensively detailed how the region has been presented on social media. The objective of this article is to sketch the contours of the Arctic discussion on Twitter and to inquire whether significant geopolitical events impact the nature of the online discussion. Using tweets on Arctic issues published between January 2020 and August 2022, we assessed the timing, prevalence, and nature of messages about the circumpolar North. Overall, the Arctic conversation on Twitter is first and foremost an Arctic climate conversation, focusing on climate change, Arctic sea ice, and permafrost thawing. Climate issues are the most salient ones and are treated independently from other topics by online users. We assessed whether the Russian invasion of Ukraine changed this dominance. We found that the Arctic Twittersphere remained still predominantly focused on climate issues, although the invasion increased Arctic military security discussions dissociated from other diplomatic or natural-resources considerations.
... While the preceding chapters illustrate how the nature-culture binary unfolds in everyday land use, legal texts and wildlife management, this one examines the significant role of the private sector in shaping wilderness imaginaries and practices. This is done using the example of nature-based tourism targeted at international markets, which is of growing economic importance to many sparsely populated regions across northernmost Europe (Runge et al. 2020). Politically, the travel sector is promoted as a beneficial economic activity for areas that are rich in natural amenities but lack the potential to attract other industries. ...
Chapter
Full-text available
This chapter unravels the nature–society dichotomy in the context of nature-based tourism. Drawing upon Neil Smith’s production of nature thesis, the chapter examines how Arctic-themed resort enclaves in northernmost Sweden, Finland and Norway, as well as Iceland, produce wilderness for international markets. Even though the studied venues evoke imaginaries of pristine wilderness to attract travellers, these touristic environments are a product of globally circulating ideas of Arctic nature coupled with tourism businesses’ need to create exchange value. As such, the key characteristic of this wilderness is that locally specific flora and fauna are overlooked in favour of distant landscape sceneries and the skies above. Unlike all other human sensations involving wilderness, the view is the easiest to commodify and is incorporated by the Arctic-themed enclaves into a highly standardised, safe and comfortable accommodation environment. Thus, wilderness is largely confined to an indoor experience in which the visual
... Additionally, the Arctic is warming as a result of climate change at a much faster rate than most of the globe (AMAP 2021; Rantanen et al. 2022). The last decade has also seen rapid development of extractive resource sectors (e.g., mining; Bartsch et al. 2021), commercial fishing (Fauchald et al. 2021), shipping and port development (Dawson et al. 2018), tourism (Runge et al. 2020), and military activity (Depledge and Kennedy-Pipe 2018). Given these issues and the importance of Arctic biodiversity to people and the planet, it is paramount to understand how threats will impact Arctic biodiversity. ...
Article
Full-text available
Arctic biodiversity is under threat from both climate-induced environmental change and anthropogenic activity. However, the rapid rate of change and the challenging conditions for studying Arctic environments mean that many research questions must be answered before we can strategically allocate resources for management. Addressing threats to biodiversity in the Arctic is further complicated by the region's complex geopolitics, as eight countries claim jurisdiction over the area, with multiple local considerations such as Indigenous sovereignty and resource rights. Here, we identify research priorities to serve as a starting point for addressing the most pressing threats to Arctic biodiversity. We began by collecting pressing research questions about Arctic biodiversity, thematizing them as either threats or actions, and then categorizing them further into 18 groups. Then, drawing on cross-disciplinary and global expertise of professionals in Arctic science, management, and policy, we considered the barriers to answering these questions and proposed potential solutions that could be implemented if barriers were overcome. Overall, our horizon scan provides an expert assessment of threats (e.g., species’ responses to climate change) and actions (e.g., a lack of fundamental information regarding Arctic biodiversity) needing attention and is intended to guide future conservation action within the Arctic.
... Groulx et al.'s (2019) study on the two cases of LCT Churchill (polar bear viewing) and the Jasper National Park (glacier tourism) in Canada also supported this paradox of tourists attaching a high 'value' to endangered landscapes/biota yet refusing to become self-critical regarding GHG emissions from their own travel [47]. Regarding tourism in the Arctic, the general backdrop of a rapid, several-fold increase in visitor numbers over the last decade or so [48] must also be kept in mind, as this creates a real and tangible footprint on the Arctic environment that cannot merely be wished away with good thoughts. ...
Article
Full-text available
Last Chance Tourism (LCT) is an increasingly popular phenomenon whereby tourists seek encounters with vanishing landscapes, cultures, and endangered species. However, there are concerns that it is not sufficiently ecologically informed, has a large carbon footprint, and may put further pressure on vulnerable ecosystems and communities. This review specifically focuses on the Arctic, which is a major global frontier for LCT and is at the forefront of disruptive and accelerating climate change. It draws on theoretical insights from the Ecological Grief concept to chart a new research focus as well as a pathway to share empathy, concern, and sorrow between scientists, communities, and visitors. Key literature sources on LCT and Ecological Grief were selected from major international scientific journals and monographs. The major findings of the study are (i) the Arctic cryosphere is a life-sustaining entity and disruptive changes in its mechanisms currently threaten the unique ecologies and culture of the region and (ii) LCT must be attentive to the emotive accounts of loss and grief associated with cryospheric change and emplace both human and non-human voices in the narrative. These findings are relevant for LCT researchers, tourism planners, and conscious travelers in the Arctic who prioritize destination sustainability.
Article
Full-text available
This paper reviews current climate change projections for northern Canada and considers what these mean for infrastructure development in the proposed Canadian Northern Corridor (CNC). We focus on chokepoints along the corridor’s notional route and estimate future costs of infrastructure along the chokepoints. We draw upon climate change projections at the end of the century (2100) using information from several climate variables sourced on the CMIP6 and CMIP5 reports. Climate variables include means and extreme values for temperature, precipitation, wind and their indirect impacts on physical features: permafrost, freezing rain and wildfires. In terms of infrastructure costs, we investigate investment costs and the useful life of nine sectors within transportation, energy and buildings infrastructures. The findings of our analysis show that mean temperatures within the CNC area could increase by 10.9oC, and precipitation by 45 per cent by 2100. Climate change could create chokepoints along the CNC route, affecting key areas essential for transportation flow. Central regions of the corridor are projected to have a higher probability of receiving concomitant impacts on several chokepoints, including combined threats from the increasing frequency of wildfires, freezing rain and permafrost thaw. Adding a climatic layer to investment costs within CNC chokepoints can increase infrastructure costs by more than 101 per cent. Transportation engineering infrastructure, electric power infrastructure and the institutional buildings sectors are most likely to be impacted. Just considering a climate layer to current infrastructure increases costs by more than 12billionforseveralhazardssuchasfreezingprecipitation(especiallyAlbertaandBC),12 billion for several hazards such as freezing precipitation (especially Alberta and BC), 7 billion for wildfires (especially BC) and more than $400 million for permafrost (especially Alberta and BC). Infrastructure built along the CNC route will need to be designed to remain functional under different climatic conditions that predominate today. Chokepoints will dictate how buildings and transportation infrastructure should be planned.
Article
Full-text available
Indicators can be powerful tools to measure progress towards achieving societal goals, and many indicators have been developed for sustainability goals nationally and internationally. When indicators are developed solely through top-down approaches without engaging local knowledge, they often fall short of capturing local perceptions and concerns relevant to decision-making. The aim of this project was to co-create a Coastal Barometer and its component indicators, together with local knowledge-holders and communities, using the Ocean Health Index as a framework. As a first step, local knowledge-holders from six communities dispersed across Northern Norway were invited to articulate local sustainability goals and the required knowledge for eight broad topics pertaining to global OHI goals: small-scale fisheries, food production, sense of place, tourism, food production, clean waters, carbon storage and economy and livelihoods. In this paper, our main focus is the co-design phase of the Coastal Barometer, namely eliciting sustainability goals. We thus, present locally desired sustainability goals and sub-goals and the process of eliciting these goals. We also include suggestions from local knowledge-holders on how to reach these goals (i.e. proposed management measures), along with researchers’ assessment of data availability (part of the co-production process) for developing indicators to measure progress towards these goals. Finally, we discuss the benefits and challenges of co-developing sustainability goals and indicators with local knowledge-holders. We conclude that co-design can increase the quality of sustainability assessments by enriching the view of coastal sustainability. This knowledge can subsequently be used to align indicators with local sustainability goals as well as to the local context where the indicators will be applied.
Thesis
Full-text available
As primeiras regulamentações ambientais em países desenvolvidos forçaram empresas a fazerem altos investimento medidas de proteção, porém quando esses custos começaram a ser repassados houve a confirmação de que havia o interesse público de pagar mais por um produto ou serviço com tal, assim os executivos começaram a testar estratégias de atração para esse setor, surgindo o termo pays-to-be-green. Com os investidores sendo atraídos por esses resultados econômicos-financeiros de empresas com responsabilidade social e financeira, surgiu então o termo Enrironmental, Social and Governance ESG, como um método para avaliação de empresas, a partir dos políticas e resultados dentro desses três pilares para a análise de ativos dentro das bolsas de valores. Como o ESG se demonstrou um meio para capitalização de recursos financeiros e Os aeroportos que são infraestruturas logísticas importantes, mas que possuem difícil capacidade de atingir retorno sobre o investimento e necessitarem de liquidez. Este estudo propõe investigar como as ações ESG, com os dados disponíveis para o cálculo de emissão direta de GEE para os aeroportos, poderá implicar no valor do patrimônio líquido das concessões. Foram utilizados como estudo de caso os aeroportos de Brasília (SBBR), Galeão (SBGL) e Campinas (SBKP). Por meio de um procedimento econométrico verificou-se por meio de uma regressão não-linear utilizando o método de Mínimos Quadrados Ordinários (MQO) quais são os coeficientes das variáveis econômicas, financeiras, de produção e ESG que podem estimar o valor do patrimônio líquido. Tais resultados também são comparados a análise de regressões feitas para ativos dos setores de energia, transportes e real estate. Os resultados obtidos sugerem que o coeficiente da variável de emissão direta de gases de efeito estufa é negativo para o atraçaõ de patrimônio líquido, o que em tese deveria acompanhar o resultado da movimentação de aeronaves que demonstrou sinal positivo. Este estudo provê evidências empíricas sobre a performance financeira de investimentos ambientais para o mercado brasileiro e aeroportos, sendo uma contribuição para a literatura.
Article
Full-text available
Spatial visitation patterns and its features on nature-based tourism are difficult to assess using only a field-based survey, which is costly and labor intensive. However, understanding of a protected area's visitation status is critical, as it can strongly influence the sustainability of natural resources. Hence, it is important to identify 'where people visit' and 'why people visit,' to evaluate the features attractive to tourists. In this regard, we proposed and applied social big data to investigate nature-based tourism in an ASEAN Heritage Park. Overall, our research was able to effectively illustrate spatial patterns of visitation using 10 years of Flickr geo-tagged photographs. Hotspots of high visitation were identified, while revealing the local spatial impact of distributed attributes. This study offers insights into the applicability of social big data to protected-area management and its potential in reinforcing existing field-based participatory approaches.
Article
Full-text available
Big data offer a great opportunity for nature-based recreation (NbR) mapping and evaluation. However, it is important to determine when and how it is appropriate to use this resource. We used Scotland as a case study to validate the use of data from Flickr as an indicator of NbR on a national scale and at several regional spatial and temporal resolutions. We compared Flickr photographs to visitor statistics in the Cairngorms National Park (CNP) and determined whether temporal variability in photo counts could be explained by known annual estimates of CNP visitor numbers. We then used a unique recent national survey of nature recreation in Scotland to determine whether the spatial distribution of Flickr photos could be explained by known spatial variability in nature use. Following this validation work, we used Flickr data to identify hotspots of wildlife watching in Scotland and investigated how they changed between 2005 and 2015. We found that spatial and temporal patterns in Flickr count are explained by measures of visitation obtained through surveys and that this relationship is reliable down to a 10 Km scale resolution. Our findings have implications for planning and management of NbR as they suggest that photographs uploaded on Flickr reflect patterns of NbR at spatial and temporal scales that are relevant for ecosystem management.
Article
Full-text available
With tourism increasing in remote regions, it is important to be able to estimate potential environmental impacts from the tourists in order to plan and manage natural areas. This study combines measures of ecological sensitivity with data from publicly available geotagged photographs posted on the social media site Flickr to assess the vulnerability of the locations frequented by foreign tourists in the Westfjords region of Iceland between 2014 and 2016. The results suggest that tourists cluster primarily around six hotspots that represented some of the major known tourist destinations of the region. Although tourists generally frequented areas with lower ecological sensitivity and rarely went far beyond the main roads, one of the hotspots was in an area of higher ecological sensitivity. Further, tourists also appeared to have higher intensity stays when they entered areas of higher ecological sensitivity. Overall, these findings highlight the usefulness of combining data from social media in assessing potential environmental impacts of tourism. However, natural resource managers should be aware of limitations in the use of such data.
Article
Full-text available
Georeferenced information on road infrastructure is essential for spatial planning, socio-economic assessments and environmental impact analyses. Yet current global road maps are typically outdated or characterized by spatial bias in coverage. In the Global Roads Inventory Project we gathered, harmonized and integrated nearly 60 geospatial datasets on road infrastructure into a global roads dataset. The resulting dataset covers 222 countries and includes over 21 million km of roads, which is two to three times the total length in the currently best available country-based global roads datasets. We then related total road length per country to country area, population density, GDP and OECD membership, resulting in a regression model with adjusted R 2 of 0.90, and found that that the highest road densities are associated with densely populated and wealthier countries. Applying our regression model to future population densities and GDP estimates from the Shared Socioeconomic Pathway (SSP) scenarios, we obtained a tentative estimate of 3.0–4.7 million km additional road length for the year 2050. Large increases in road length were projected for developing nations in some of the world's last remaining wilderness areas, such as the Amazon, the Congo basin and New Guinea. This highlights the need for accurate spatial road datasets to underpin strategic spatial planning in order to reduce the impacts of roads in remaining pristine ecosystems.
Article
Full-text available
Managing protected areas effectively requires information about patterns of visitor use, but these data are often limited. We explore how geotagged photos on Flickr, a popular photo-sharing social-media site, can generate hotspot maps and distribution models of temporal and spatial patterns of use in two mountain-protected areas of high conservation value. In Aconcagua Provincial Park (Argentina), two routes to the summit of Aconcagua were used in summer, but most visitors stayed close to the main road, using formal and informal walking trails and the Visitor Centre, while in winter, there was very limited visitation. In Kosciuszko National Park (Australia), alpine walking trails were popular in summer, but in winter, most visitors stayed in the lower altitude ski resorts and ski trails. Results demonstrate the usefulness of social-media data alone as well as a complement for visitor monitoring, providing spatial and temporal information for site-specific and park-level management of visitors and potential impacts in conservation areas.
Article
Full-text available
Tracking spatial and temporal trends of events (e.g. disease outbreaks and natural disasters) is important for situation awareness and timely response. Social media, with increasing popularity, provide an effective way to collect event-related data from massive populations and thus a significant opportunity to dynamically monitor events as they emerge and evolve. While existing research has demonstrated the value of social media as sensors in event detection, estimating potential time spans and influenced areas of an event from social media remains challenging. Challenges include the unstable volumes of available data, the spatial heterogeneity of event activities and social media data, and the data sparsity. This paper describes a systematic approach to detecting potential spatiotemporal patterns of events by resolving these challenges through several interrelated strategies: using kernel density estimation for smoothed social media intensity surfaces; utilizing event-unrelated social media posts to help map relative event prevalence; and normalizing event indicators based on historical fluctuation. This approach generates event indicator maps and significance maps explaining spatiotemporal variations of event prevalence to identify space-time regions with potentially abnormal event activities. The approach has been applied to detect influenza activity patterns in the conterminous US using Twitter data. A set of experiments demonstrated that our approach produces high-resolution influenza activity maps that could be explained by available ground truth data.
Article
Current research on tourism in the Arctic has focused largely on the extent, location, and type of tourism activities that occur in the region. Recently, challenges have been identified that the tourism industry is likely to face in the wake of global changes, including climate change. Related research, conducted within and outside of the Arctic, suggests that rural communities can become economically dependent on natural resource extraction (e.g. oil, gas, timber harvesting, and mining of minerals) and non-extractive resources (e.g. nature-based recreation and tourism), limiting diversification and potentially threatening resilience of rural communities. In the western USA, communities have become dependent on both extractive and non-extractive natural resource activities including nature-based tourism; however, it is less clear whether a similar situation is occurring in Arctic communities. In this article, we propose a framework and indicators to analyze the potential dependence of Arctic communities on nature-based tourism and the resilience of Arctic communities to potential boom–bust cycles of nature-based tourism. To do so, we examine the current state-of-knowledge about tourism and nature-based tourism in the Arctic through the lens of boom–bust dynamics and social-ecological systems.
Article
Simple features are a standardized way of encoding spatial vector data (points, lines, polygons) in computers. The sf package implements simple features in R, and has roughly the same capacity for spatial vector data as packages sp, rgeos, and rgdal. We describe the need for this package, its place in the R package ecosystem, and its potential to connect R to other computer systems. We illustrate this with examples of its use.
Article
Vessel traffic has been increasing rapidly in the Arctic, and within the Canadian Arctic, tourist vessels are the fastest growing maritime sector. Vessel traffic can cause a variety of impacts on whales, including ship strikes and acoustic disturbance. Here, the overlap between tourist vessels (e.g., pleasure craft/yachts and passenger vessels/cruise ships) and whale concentration areas is assessed within the Inuvialuit Settlement Region of the western Canadian Arctic. Different management measures which could be used to reduce impacts on whales are also assessed. Passenger vessels have had a relatively constant overlap with whale concentration areas through time, whereas pleasure craft have had a recent and rapid increase. Passenger vessels may have a greater impact on whales, compared to pleasure craft, since they are larger and travel faster. Excluding vessels from the two marine protected areas in the region would have no impact on whales within concentration areas, since vessels would likely just be displaced to adjacent areas with similar whale concentrations. Restricting vessels to the Canadian government's proposed low-impact corridor may reduce impact slightly, but creating a corridor completely outside of the known whale area could more significantly reduce the potential impact of vessels on whales in those areas. Restricting vessel speed within whale areas would also reduce the impact of passenger vessels, but would not likely reduce the impact of pleasure craft. Overall, a combination of management measures may be the best way to reduce impacts on whales in concentration areas.
Article
Ecotourism is growing rapidly in biodiversity hotspots because of its promise to achieve both economic growth and environmental conservation. We reviewed the literature for empirical evidence that ecotourism protects forests. Our conclusions are at once both sobering and encouraging. Ecotourism, as it is typically practiced, leads to deforestation. However, when accompanied by conservation mechanisms (e.g. protected area, Payment for Ecosystem Services, monitoring/enforcement), ecotourism can protect forests. Ecotourism sometimes leads to forest regeneration in agrarian landscapes, but trade-offs, for example old-growth deforestation or water pollution, may occur. From a methodological perspective, we found a dearth (only 17) of articles that empirically analyzed ecotourism impacts on forests, and no studies that used counterfactual impact evaluation approaches. We conclude that there is an insufficient evidence base for inferring effects of ecotourism on forests, and we identify research priorities to build knowledge about how, when, and where to implement ecotourism.