Available via license: CC BY 4.0
Content may be subject to copyright.
Data Quality of OpenStreetMap for Industrial Sites in the Arctic
Daniel Kwakye1, Sabrina Marx1, Benjamin Herfort1, Moritz Langer3,4, and Sven Lautenbach1,2
1Heidelberg Institute for Geoinformation Technology, Heidelberg, Germany
2GIScience Chair, Institute of Geography, Heidelberg University, Heidelberg, Germany
3Department of Earth Sciences, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands
4Alfred Wegener Institute, Helmholtz Center for Polar and Marine Research, Postdam, Germany
Correspondence: Sven Lautenbach (sven.lautenbach@heigit.org)
Abstract. Climate change is causing rapid warming in
the Arctic region, resulting in the thawing of permafrost.
This has substantial environmental implications, such as
the release and mobilisation of contaminants from past
and present industrial activities. However, freely accessi-
ble public geographical information is scarce on industrial
sites and activities in much of the Arctic, which makes sci-
entific research such as impact assessment difficult. Open-
StreetMap (OSM) can be a valuable resource for iden-
tifying and assessing industrial sites for contamination.
However, OSM data quality is not uniform across regions
necessitating our evaluation of its reliability for identify-
ing industrial sites and contamination hotspots in the ar-
eas most susceptible to permafrost thawing. Therefore, we
examined in our study the object and attribute complete-
ness as well as the currentness of OSM data on industrial
sites. Our study focused on the regions defined by the pres-
ence of either discontinuous or continuous permafrost lo-
cated in Canada, the USA, Denmark, Russia, and Norway,
as these regions are expected to show strongest impacts
of rising temperatures with respect to industrial pollution.
The highest object completeness and currentness were ob-
tained in Denmark (99% and 48% respectively). Russia
had the lowest completeness (68%) and Canada had the
lowest currentness (30%). Despite the promising average
completeness of 86% and the average currentness of 35%,
only 5.6% of industrial sites mapped in OSM contained in-
formation on the type of industry. This finding highlights
the need for efforts to enhance attribute completeness gaps
to maximize the use of OSM data in comprehensive envi-
ronmental analyses.
Keywords. OpenStreetMap, Intrinsic Quality Assess-
ment, Permafrost, Volunteered Geographic Information,
Ohsome Quality API
1 Introduction
The Arctic region is experiencing unprecedented rapid
warming due to climate change, with scientists estimating
rates that are up to four times faster than the global aver-
age (Rantanen et al., 2022). One consequence of this rapid
warming is thawing of permafrost which covers vast areas
of the Arctic. Permafrost is defined as ground that is frozen
for at least two consecutive years. The permafrost has
served as a natural barrier that prevents the spread of pol-
lutants (Miner et al., 2021) as well as a stable and depend-
able foundation for buildings and infrastructure (Langer
et al., 2023). Past and current industrial activities in the
Arctic have resulted in the accumulation of hazardous sub-
stances in the permafrost region. The thawing of the per-
mafrost potentially releases these accumulated hazardous
substances into the ecosystem (Vonk et al., 2015). In addi-
tion, the instability of the frozen ground due to the thaw-
ing increases the risk of industrial containment structures
collapsing (Langer et al., 2023) and further polluting the
environment. The spread of these contaminations through
the ecosystems poses a severe risk to communities within
and outside of the Arctic region. Consequently, the Arc-
tic region has become a focal point for extensive research,
including the development of comprehensive risk assess-
ment frameworks (Hjort et al., 2018) and climate mod-
els for its rapidly changing conditions, which are causing
various issues, including the legacy pollution. Despite the
growing attention, there is the challenge of the scarcity
of freely accessible public geographical data on indus-
trial sites and activities in the region. In response, many
researchers (Liu et al., 2023; Lloyd-Jones et al., 2023;
Xu et al., 2022) have turned to OpenStreetMap (OSM)
(Bartsch et al., 2020), a crowdsourced geographic database
aimed at mapping the whole world.
While OSM data can be a valuable tool for Arctic research,
particularly for monitoring and assessing industrial sites
and activities in the region, its reliability and accuracy
need to be assessed. OSM data can provide useful infor-
1 of 7
AGILE: GIScience Series, 5, 34, 2024. https://doi.org/10.5194/agile-giss-5-34-2024
Proceedings of the 27th AGILE Conference on Geographic Information Science, 4–7 Sept. 2024.
Editors: Alison Heppenstall, Mingshu Wang, Urska Demsar, Rob Lemmens, and Jing Yao.
This contribution underwent peer review based on a full paper submission.
© Author(s) 2024. This work is distributed under the Creative Commons Attribution 4.0 License.
mation on the location, extent, and characteristics of in-
dustrial sites along with local knowledge. However, like
any Volunteered Geographic Information (VGI) which is
mainly generated by non-professionals, OSM carries no
assurance of quality (Goodchild and Li, 2012). Accuracy
assessments of OSM data in the Arctic regions are largely
missing (Xu et al., 2022). The quality of OSM data varies
across different regions (Raifer et al., 2019), with sub-
stantial heterogeneity in completeness and other aspects
of data quality between and within countries (Barrington-
Leigh and Millard-Ball, 2017). Factors such as the gad-
gets used for the mapping, and the skills and motivation of
contributors (Barron et al., 2014; Sehra et al., 2017) add to
potential inaccuracies. Moreover, the "Any tags you like"
policy by OSM which permits map editors to establish
their keys and values for tags of entities they are editing
also introduces inconsistencies in the data (Mooney and
Corcoran, 2012). Addressing the question of reliability be-
comes paramount, particularly when using OSM data in
identifying industrial sites and contamination hotspots for
critical applications such as risk assessment frameworks in
the Arctic.
We therefore intrinsically assessed the quality of OSM
data across our study region to address these concerns.
This intrinsic approach considers the temporal evolution of
the OSM data within the study area. We aim to spot where
extra mapping efforts are required and to give us an idea
of the difference in OSM data quality across the countries.
This assessment will provide insights into the reliability
of OSM data and its fitness for the purpose of identifying
industrial sites and contamination in the Arctic.
2 OSM Data Quality Assessment
Defining and evaluating data quality is a complex task as
distinct quality criteria and considerations are required for
different uses of data (Reda et al., 2023). There are several
approaches for the assessment of OSM data quality. A dis-
tinction is made between two different methods (Brück-
ner et al., 2021). Extrinsic methods are based on a com-
parison of OSM data with reference data sources. These
approaches are limited by the requirement of external
datasets which are not always available due to licensing re-
strictions or costs (Barron et al., 2014). Intrinsic methods
(Barron et al., 2014; Ballatore and Zipf, 2015; Antoniou
and Skopeliti, 2015; Barrington-Leigh and Millard-Ball,
2017; Brückner et al., 2021) provide a way to evaluate the
fitness for purpose of OSM data by considering just the
history of the OSM data itself. With the benefit of the en-
tire editing history of each edit to the OSM database, it
serves as a great source for quality assessment (Minghini
and Frassinelli, 2019). This has motivated the implemen-
tation of several software applications based on the analy-
sis of the OSM history data that provides a way to access
the intrinsic quality measures for OSM objects. OSHDB
(Raifer et al., 2019), a framework for spatio-temporal anal-
ysis of OSM history data, is one realisation of such soft-
ware optimized for working with OSM history data that
makes it easy to assess the intrinsic quality measures of
OSM data.
There is a limited amount of research on the OSM data
quality in the Arctic region. To the best of our knowl-
edge, only two studies have been conducted to assess the
quality of OSM data in this area. Both studies utilized ex-
trinsic methods to evaluate the completeness of OSM data
by comparing it to the SACHI (Sentinel-1/2 derived Arc-
tic Coastal Human Impact) dataset. In a study conducted
by Bartsch et al. (2021), the completeness of OSM data
regarding infrastructure in the Arctic region was exam-
ined. The study found that, depending on the specific re-
gion, the SACHI dataset contained 8%–48% more infor-
mation (in terms of human presence) compared to OSM.
Another study conducted by Langer et al. (2023) com-
bined OSM data with the Atlas of Population, Society
and Economy in the Arctic (APSEA) dataset to assess the
logical consistency with the SACHI dataset. The evalua-
tion revealed that OSM-APSEA had a significantly lower
number of industrial sites compared to SACHI, with ap-
proximately 40% missing. While both studies contribute
to the general assessment of OSM data quality in the Arc-
tic, the spatial scope of the SACHI dataset used as the ref-
erence dataset is limited to 100 km inland (Langer et al.,
2023) and focuses on the Arctic coastal areas. Addition-
ally, Bartsch et al. (2021) analyzed only building footprints
in the OSM database while Langer et al. (2023) consid-
ered only the OSM tags ‘landuse’=’industrial’ and ‘build-
ing’=’industrial’.
For these reasons, we extended the OSM tags considered
and performed an intrinsic quality assessment of industrial
objects in OSM considering object completeness, current-
ness and attribute completeness.
3 Methods and Data
3.1 Data
We defined the spatial extent of our analysis based on
the presence of either discontinuous or continuous per-
mafrost in five Arctic countries, namely Canada, the
USA (Alaska), Denmark (Greenland), Russia, and Nor-
way (Svalbard). The extents were generalised to approxi-
mate the areas with permafrost coverage greater than 50%.
Figure 1shows a map of our study area.
In an exploratory analysis, the study examined features
within the OSM database that represent industrial activ-
ity. Table 1presents the various tags that were considered
most relevant for identifying industrial sites.
2 of 7
AGILE: GIScience Series, 5, 34, 2024 | https://doi.org/10.5194/agile-giss-5-34-2024
Table 1. Tags for describing industrial-related features in OSM used for the study. Source https://taginfo.openstreetmap.org/
Tag Description
landuse=industrial Used for areas with predominantly workshops, factories, warehouses, etc
building=industrial Used for buildings constructed for some manufacturing process
industrial=* Used to describe the type of industry or the industrial object
man_made=works For mapping an industrial production plant or factory
landuse=quarry For mapping open pit mines (area for surface extraction of mineral resources)
man_made=storage_tank For mapping containers that hold liquids or compressed gases
man_made=flare For mapping towers constructed to burn off excess gases
Figure 1. Map of the study area using Permafrost extent data
from Obu et al. (2018) Source:https://apgc.awi.de/dataset/pex.
3.2 Methods
We used the Ohsome API and Ohsome Quality API
(OQAPI) in a Jupyter Notebook to analyze the underlying
rich data source of the OpenStreetMap History Database
(OSHDB) (Raifer et al., 2019). The Ohsome API provides
endpoints for aggregated statistics on the contributions,
users and OSM elements as well as endpoints for extract-
ing them. We filtered the data according to the required
features using two Ohsome API endpoints. The elements
extraction endpoint was used to extract the map features of
interest in the study areas for the attribute analysis. Addi-
tionally, the ‘elements’ aggregation endpoint was used to
compute the count of the map features of interest within
our study regions.
OQAPI was used to compute intrinsic quality indicator es-
timates for OSM data in the study areas. This involved as-
sessing the intrinsic completeness and currentness of the
map features of interest in the OSM database using a satu-
ration curve approach. The theory behind how OQAPI cal-
culates the completeness quality estimates is thoroughly
explained in Brückner et al. (2021). The general idea is
that, given a substantial level of activity from contributors
over a period, contributions eventually converge with the
actual number of real-world objects in the given area. Sat-
uration curves can be used to estimate the level of satura-
tion, which serves as an estimate for the true number of
objects in a given region. Saturation refers to the point at
which the growth or contribution rate levels off or stabi-
lizes.
OQAPI retrieves the aggregated monthly contributions to
the specified topic (a set of features whose completeness
we are interested in computing, see table 1) for the pe-
riod from 2008-01-01 to the latest available snapshot of
the OSM database. Six different curve fitting models (lo-
gistic and non-logistics curves) are applied to the retrieved
data. The best curve is selected based on the one that gives
the minimum Mean Absolute Error (MAE). The satura-
tion within the last 3 years is then calculated based on the
selected curve.
The currentness indicator is computed by binning edits
into three classes: up-to-date (0-3 years), in-between (3-8
years), and out-of-date (over 8 years) based on the times-
tamps for the edits. The percentage of objects that have
received updates within the feature class in the last three
years is used as the score to determine the currentness of
a particular feature within the dataset. This indicator gives
an estimation of the up-to-dateness of features in OSM
We computed the attribute completeness by assessing
the user-entered information of specific attributes. As the
type of industry is relevant for predicting how indus-
trial sites will be affected by permafrost thaw, we fo-
cused on the completeness of the attributes ‘industrial’
and ‘plant:source’ (on the two major OSM tags lan-
duse=industrial and building=industrial). The percentage
of objects within an OSM feature class that have values
populated for the field of interest is used as the measure.
Equation 1is the formula used for the calculation.
Attribute Completeness =F
G×100 (1)
Fis the number of objects within the feature with a value
present for the field of interest
Gis the total number of objects within the feature
3.3 Software and data availability
The analysis was performed in a Python
Jupyter Notebook. All source code, prepro-
cessed data and results can be found at https:
//gitlab.gistools.geog.uni-heidelberg.de/giscience/
big-data/ohsome/ohsome-api-analysis-examples/
osm-data-quality-in-the-arctic.
3 of 7
AGILE: GIScience Series, 5, 34, 2024 | https://doi.org/10.5194/agile-giss-5-34-2024
OSM data were extracted using the Ohsome API
at https://api.ohsome.org. Intrinsic completeness
and currentness quality estimates were computed
using a local instance of OQAPI pulled from
https://github.com/GIScience/ohsome-quality-api.git.
We created a topic by editing the ’preset.yaml’ file
https://github.com/GIScience/ohsome-quality-api/blob/
main/ohsome_quality_api/topics/presets.yaml to suit the
context of this research by specifying the map features
to be used for the estimation and the aggregation type.
The edited ’presets.yaml’ file is provided in our working
folder to be used to replace the original version that comes
with cloning the OQAPI repository.
The data for the permafrost extent was acquired from Obu
et al. (2018) at https://apgc.awi.de/dataset/pex.
4 Results
In our study, we analysed the completeness, current-
ness and attribute completeness of industrial sites in the
OSM database. We found a total of 16,351 objects in the
database that were likely associated with industrial activi-
ties. Figure 2shows the distribution of the sites across the
countries. The number of objects associated with each tag
is captured in table 2.
Table 2. Count of all the features with the specified tags in the
OSM database for the study area. Some features may have more
than one tag.
Tag Number of Features
landuse=industrial 6357
building=industrial 3830
man_made=works 483
industrial=* 1118
landuse=quarry 1682
man_made=storage_tank 3705
man_made=flare 463
Figure 2. The distribution and density of industrial sites and re-
lated features within the study area.
The average intrinsic completeness for the study region
was approximately 86%, with a range from 68% to 99%.
Denmark and Norway exhibited high estimated complete-
ness, with values exceeding 97%. However, the attribute
completeness was much lower, with an average value of
only 5.6%. The highest attribute completeness value of
26.6% was obtained for the attribute ’industrial’ for ob-
jects with the main tag ’landuse’=’industrial’ in the USA,
while Norway had the lowest value of 7.5%. For the main
tag ‘building’=’industrial’, the attribute ‘industrial’, which
provides information on the type of industry was miss-
ing for both Denmark and the USA. Norway, Canada, and
Russia all had attribute completeness values for the field
‘industrial’ of the main tag ‘building’=’industrial’ that
were less than 1%. The analysis for the field ’plant:source’
showed that Norway had the highest attribute complete-
ness of 13.3% for the main tag ’landuse’=’industrial’. On
the other hand, Canada had the highest value of 5% for the
main tag ’building’=’industrial’. Table 3shows the results
of the attribute completeness analysis.
Even though the results in the intrinsic completeness of the
data showed heterogeneity across the countries, the up-to-
dateness of the map data, was relatively consistent across
the four countries. Denmark stood out with a high current-
ness score of 48%. The average currentness obtained for
the study region was 35%. Figure 3shows the results of the
computed OSM data quality elements for the countries.
5 Discussions
Comparing our findings with a previous extrinsic study by
Bartsch et al. (2021) on the completeness of OSM building
footprints in the Arctic region, our estimated completeness
was approximately +8% more. This relatively high esti-
mate can be accounted for by a mixture of factors. Firstly,
the average currentness estimate of 35% suggests approx-
imately one-third of the features were updated in the last
three years. Recent updates lead to improved complete-
ness. Denmark with the highest intrinsic completeness re-
ceived the highest editing updates within the last three
years with almost half of the features considered to be cur-
rent. This explains why the completeness of Denmark im-
proved significantly compared to the previous study. Sec-
ondly, the limitation of the intrinsic completeness method-
ology. The completeness computation uses the best fitting
curve from a set of limited growth curves. As described
in Brückner et al. (2021), non-logistic curves can lead to
lower completeness estimates than sigmoid curves. Den-
mark’s intensive mapping activity in the last 3 years can be
described best by a sigmoid curve, which may have bene-
fited the completeness value.
The attribute completeness analysis revealed that sec-
ondary fields were barely populated by mappers. The is-
sue of low attribute completeness is one major challenge
with OSM data globally. The average attribute complete-
ness of 5.6% hinders the potential of maximising the use
of the data for comprehensive advanced spatial analysis.
For instance, Langer et al. (2023) could not classify the
4 of 7
AGILE: GIScience Series, 5, 34, 2024 | https://doi.org/10.5194/agile-giss-5-34-2024
Table 3. Attribute completeness analysis of tags ’landuse’=’industrial’ and ’building’=’industrial’ in the study area.
Country Tag Number of Objects Field/Attribute Values Present % completeness average
USA
landuse=industrial 263 industrial 70 26.62%
7.33%
plant:source 5 1.90%
building=industrial 125 industrial field absent 0%
plant:source 1 0.80%
Canada
landuse=industrial 380 industrial 37 9.74%
5.49%
plant:source 25 6.58%
building=industrial 160 industrial 1 0.63%
plant:source 8 5.00%
Russia
landuse=industrial 5579 industrial 981 17.58%
4.66%
plant:source 35 0.63%
building=industrial 3124 industrial 8 0.26%
plant:source 6 0.19%
Denmark
landuse=industrial 120 Industrial 9 7.50%
4.95%
plant:source 14 11.67%
building=industrial 313 Industrial field absent 0.00%
plant:source 2 0.64%
Norway
landuse=industrial 15 Industrial 1 6.67%
5.69%
plant:source 2 13.33%
building=industrial 108 Industrial 1 0.93%
plant:source 2 1.85%
Figure 3. The computed quality estimates for the countries.
majority of features (65% of mapped industrial sites) in
their combined APSEA and OSM database due to miss-
ing tags. These results indicate the need to complete infor-
mation about most of the objects in the OSM database to
improve the quality and usability of the data for industrial
site analysis.
5.1 Conclusion and Outlook
We found variations in OSM data quality across the re-
gion. The promising estimated mean completeness level
of 86% found in the study shows the great potential to use
OSM data in identifying industrial sites to support Arctic
research. However, the low mean attribute completeness
of 5.6% found is a major concern. This low attribute ac-
curacy poses a challenge in the use of OSM data for tasks
like classifying industries and conducting other advanced
spatial analyses. This highlights the need for intensified
efforts to complete information about most of the objects
analysed. This can be achieved through regular organisa-
tion of social events such as mapping parties and map-
athons in the Arctic region (Hristova et al., 2021; Schott
et al., 2021).
Additionally, critical applications like the development of
risk assessment frameworks require high positional accu-
racy and detailed information. Consequently, in terms of
5 of 7
AGILE: GIScience Series, 5, 34, 2024 | https://doi.org/10.5194/agile-giss-5-34-2024
future research, we intend to apply extrinsic approaches to
check the geometric accuracies of the OSM data. In addi-
tion, we would examine the variability of the OSM data
quality within the countries.
References
Antoniou, V. and Skopeliti, A.: Measures AND Indicators OF
VGI Quality: An Overview, ISPRS Annals of the Photogram-
metry, Remote Sensing and Spatial Information Sciences, II-
3/W5, 345–351, https://doi.org/10.5194/isprsannals-II-3-W5-
345-2015, 2015.
Ballatore, A. and Zipf, A.: A Conceptual Quality Frame-
work for Volunteered Geographic Information, in: Spatial In-
formation Theory, edited by Fabrikant, S. I., Raubal, M.,
Bertolotto, M., Davies, C., Freundschuh, S., and Bell, S., vol.
9368, pp. 89–107, Springer International Publishing, Cham,
https://doi.org/10.1007/978-3-319-23374-1_5, 2015.
Barrington-Leigh, C. and Millard-Ball, A.: The world’s user-
generated road map is more than 80
Barron, C., Neis, P., and Zipf, A.: A Comprehensive Framework
for Intrinsic OpenStreetMap Quality Analysis, Transactions in
GIS, 18, 877–895, https://doi.org/10.1111/tgis.12073, 2014.
Bartsch, A., Pointner, G., Ingeman-Nielsen, T., and Lu, W.: To-
wards Circumpolar Mapping of Arctic Settlements and Infras-
tructure Based on Sentinel-1 and Sentinel-2, Remote Sensing,
12, 2368, https://doi.org/10.3390/rs12152368, 2020.
Bartsch, A., Pointner, G., Nitze, I., Efimova, A., Jakober, D., Ley,
S., Högström, E., Grosse, G., and Schweitzer, P.: Expand-
ing infrastructure and growing anthropogenic impacts along
Arctic coasts, Environmental Research Letters, 16, 115 013,
https://doi.org/10.1088/1748-9326/ac3176, 2021.
Brückner, J., Schott, M., Zipf, A., and Lautenbach, S.: As-
sessing shop completeness in OpenStreetMap for two fed-
eral states in Germany, AGILE: GIScience Series, 2, 1–7,
https://doi.org/10.5194/agile-giss-2-20-2021, 2021.
Goodchild, M. F. and Li, L.: Assuring the quality of volun-
teered geographic information, Spatial Statistics, 1, 110–120,
https://doi.org/10.1016/j.spasta.2012.03.002, 2012.
Hjort, J., Karjalainen, O., Aalto, J., Westermann, S., Ro-
manovsky, V. E., Nelson, F. E., Etzelmüller, B., and Lu-
oto, M.: Degrading permafrost puts Arctic infrastructure
at risk by mid-century, Nature Communications, 9, 5147,
https://doi.org/10.1038/s41467-018-07557-4, 2018.
Hristova, D., Quattrone, G., Mashhadi, A., and Capra, L.:
The Life of the Party: Impact of Social Mapping in
OpenStreetMap, Proceedings of the International AAAI
Conference on Web and Social Media, 7, 234–243,
https://doi.org/10.1609/icwsm.v7i1.14416, 2021.
Langer, M., Von Deimling, T. S., Westermann, S., Rolph,
R., Rutte, R., Antonova, S., Rachold, V., Schultz, M.,
Oehme, A., and Grosse, G.: Thawing permafrost poses en-
vironmental threat to thousands of sites with legacy in-
dustrial contamination, Nature Communications, 14, 1721,
https://doi.org/10.1038/s41467-023-37276-4, 2023.
Liu, Z., Yang, J., and Huang, X.: Landsat-derived impervi-
ous surface area expansion in the Arctic from 1985 to
2021, Science of The Total Environment, 905, 166 966,
https://doi.org/10.1016/j.scitotenv.2023.166966, 2023.
Lloyd-Jones, T., Dick, J. J., Lane, T. P., Cunningham, E. M., and
Kiriakoulakis, K.: Occurrence and sources of microplastics
on Arctic beaches: Svalbard, Marine Pollution Bulletin, 196,
115 586, https://doi.org/10.1016/j.marpolbul.2023.115586,
2023.
Miner, K. R., D’Andrilli, J., Mackelprang, R., Edwards,
A., Malaska, M. J., Waldrop, M. P., and Miller, C. E.:
Emergent biogeochemical risks from Arctic permafrost
degradation, Nature Climate Change, 11, 809–819,
https://doi.org/10.1038/s41558-021-01162-y, 2021.
Minghini, M. and Frassinelli, F.: OpenStreetMap history
for intrinsic quality assessment: Is OSM up-to-date?,
Open Geospatial Data, Software and Standards, 4, 9,
https://doi.org/10.1186/s40965-019-0067-x, 2019.
Mooney, P. and Corcoran, P.: The Annotation Process
in OpenStreetMap, Transactions in GIS, 16, 561–579,
https://doi.org/10.1111/j.1467-9671.2012.01306.x, 2012.
Obu, J., Westermann, S., Kääb, A., and Bartsch, A.: Ground Tem-
perature Map, 2000-2016, Northern Hemisphere Permafrost,
https://doi.org/10.1594/PANGAEA.888600, 2018.
Raifer, M., Troilo, R., Kowatsch, F., Auer, M., Loos, L., Marx,
S., Przybill, K., Fendrich, S., Mocnik, F.-B., and Zipf, A.:
OSHDB: a framework for spatio-temporal analysis of Open-
StreetMap history data, Open Geospatial Data, Software and
Standards, 4, 3, https://doi.org/10.1186/s40965-019-0061-3,
2019.
Rantanen, M., Karpechko, A. Y., Lipponen, A., Nordling, K.,
Hyvärinen, O., Ruosteenoja, K., Vihma, T., and Laaksonen,
A.: The Arctic has warmed nearly four times faster than the
globe since 1979, Communications Earth Environment, 3,
168, https://doi.org/10.1038/s43247-022-00498-3, 2022.
Reda, O., Benabdellah, N. C., and Zellou, A.: A sys-
tematic literature review on data quality assessment,
Bulletin of Electrical Engineering and Informatics, 12,
https://doi.org/10.11591/eei.v12i6.5667, 2023.
Schott, M., Grinberger, A. Y., Lautenbach, S., and
Zipf, A.: The Impact of Community Happenings in
OpenStreetMap—Establishing a Framework for On-
line Community Member Activity Analyses, ISPRS
International Journal of Geo-Information, 10, 164,
https://doi.org/10.3390/ijgi10030164, 2021.
Sehra, S., Singh, J., and Rai, H.: Assessing OpenStreetMap
Data Using Intrinsic Quality Indicators: An Extension to
the QGIS Processing Toolbox, Future Internet, 9, 15,
https://doi.org/10.3390/fi9020015, 2017.
Vonk, J. E., Tank, S. E., Bowden, W. B., Laurion, I., Vincent,
W. F., Alekseychik, P., Amyot, M., Billet, M. F., Canário,
J., Cory, R. M., Deshpande, B. N., Helbig, M., Jammet, M.,
Karlsson, J., Larouche, J., MacMillan, G., Rautio, M., Wal-
ter Anthony, K. M., and Wickland, K. P.: Reviews and synthe-
ses: Effects of permafrost thaw on Arctic aquatic ecosystems,
Biogeosciences, 12, 7129–7167, https://doi.org/10.5194/bg-
12-7129-2015, 2015.
Xu, X., Liu, C., Liu, C., Hui, F., Cheng, X., and Huang,
H.: Fine-resolution mapping of the circumpolar Arctic Man-
made impervious areas (CAMI) using sentinels, Open-
6 of 7
AGILE: GIScience Series, 5, 34, 2024 | https://doi.org/10.5194/agile-giss-5-34-2024
