Fram Centre

Climate-Ecological Observatory for Arctic Tundra—Status 2020

August 25, 2020
The Arctic tundra is challenged by climate change — more so than most other ecosystems on Earth. The rapid shifts to new climate regimes may give rise to new ecosystems with unknown properties. These dramatic changes call for ecosystem-based monitoring of climate impacts on Arctic food webs.

By: Eeva Soininen and Rolf A Ims // UiT The Arctic University of Norway, Åshild Ø Pedersen // Norwegian Polar Institute and Audun Stien // NINA – Norwegian Institute for Nature Research



Top photo:

Full-scale automatic weather stations are core infrastructure in COAT’s climate monitoring network. The weather stations are “hot-spots” for potential co-location and expansion of measurements to cover a wider range of variables related to both the biosphere and cryosphere within COAT and the Svalbard Integrated Arctic Earth Observing System (SIOS). Data from the weather stations can be downloaded from Photo: Ketil Isaksen

The Climate-Ecological Observatory for Arctic Tundra (COAT) is a response from five Fram Centre institutions to international calls to establish scientifically robust observation systems that enable real-time detection, documentation, and understanding of climate impacts on Arctic tundra ecosystems.

Ecosystem-based monitoring and research

COAT is a system for long-term research with a focus on the Low-Arctic Varanger Peninsula and High-Arctic Svalbard. It combines state-of-the-art climate-ecological research with management. The focal COAT regions provide contrasts in system complexity, climate, and management regimes. COAT builds on and expands the ongoing research and long-term monitoring in both regions.

Food web models and modules

COAT aims to establish causal relations between components of the food webs that are important to ecosystem functioning and management, and how climate and management actions impact these relations. The likely paths for such causal relations are expressed in terms of conceptual “climate and management impact path models”.

These models encompass tightly linked clusters (termed modules) of organism groups that are expected to be directly or indirectly impacted by the same set of climate and management drivers. The purpose of the conceptual models is to form a framework for data-driven causal analyses and predictions of climatic effects, and further infer how management could be effective in mitigating predicted unwanted effects.



Drones are an important infrastructure to monitor landscape disturbances caused by geese. Photo: Virve Ravolainen



Monitoring, analyses, adaptive updates

The COAT science plan (download from describes the overall approach, background knowledge, climate impact path models, and the monitoring design. The study methods include both ground observations, automatic data recording, and remote sensing. The plan also describes the adaptive monitoring approach of COAT: How new knowledge, technology, science questions, and management intervention will be incorporated into models and monitoring designs in an iterative manner – a process in which stakeholders and management authorities may be engaged (see example XXX). As part of adaptive monitoring, COAT also develops new monitoring technologies and new data analyses and modeling tools.

COAT Infrastructure in the field

Further reading

Ims RA, Killengreen ST, Ehrich D, Flagstad Ø, Hamel S, Henden J-A, Jensvoll I, Yoccoz NG (2017) Ecosystem drivers of an Arctic fox population at the western fringe of the Eurasian Arctic. Polar Research 36: DOI:10.1080/17518369.2017.1323621

In 2016, COAT started to implement research infrastructure related to data capture (i.e. gathering information related to both food webs and climate), field logistics and data management solutions.

To cover the range of existing variation in climatic and management contexts, COAT data sampling systems are geographically distributed. The first five COAT weather stations were set up in 2018-2019, in inland regions of Svalbard and across one coast-to-inland gradient at Varanger. Five more will be set up in Svalbard in 2020-2021.

Other types of infrastructure that have been established are herbivore exclosures, networks of cameras traps and acoustic sensors, telemetry equipment, drones, and networks of small instruments that log climate parameters. These are distributed at spatial and temporal scales appropriate for estimating the weather patterns and ecological interactions of interest in the COAT modules.

Field logistics is essential for the large COAT field crews that operate in remote tundra areas in Varanger Peninsula and Svalbard. As part of the infrastructure project, COAT Varanger has established local storage facilities in Vadsø, acquired transport units (snowmobiles, ATVs, cars), and in fall 2019 established a permanent field station. COAT Svalbard field operations are organised under the umbrella of Norwegian Polar Institute logistics.

COAT Digital Infrastructure

COAT partners

Norwegian Institute for Nature Research, Norwegian Meteorological Institute, Norwegian Polar Institute, University Centre in Svalbard and UiT The Arctic University of Norway

A data management system is a crucial part of the COAT infrastructure. The COAT data portal will gather all primary data from COAT, providing open access to external users. Work with the COAT data portal has advanced to the testing of the first version. Concurrently, the COAT team is working with establishing data format standards, organising and documenting datasets, and developing transparent and reproducible data pipelines for activities ranging from taking field notes to monitoring state variables. An open version of the data portal will be released in 2020 with access through the COAT web pages.


The food webs in Svalbard and on the Varanger Peninsula are represented by five and six monitoring modules, respectively. Examples of linkages within the modules include predator-prey and plant-herbivore interactions or competitive interactions. The monitoring targets in each module are listed within the boxes and the target that gives each module its name is written in bold.


Spring foraging (grubbing) by pink-footed geese has changed the composition, structure, and function of the vegetation community. Photo: Cornelia Jaspers


Pink-footed goose population. The pink-footed goose population has increased substantially during the last decades. Graph modified from the AEWA international single species management plan.



The Varanger Arctic fox module – an example of the COAT adaptive monitoring


The Arctic fox is the only mammalian predator endemic to the terrestrial Arctic. Over the last century, Arctic fox populations have declined steeply in the southernmost parts of their range, including the Varanger Peninsula.

The conceptual model for the Arctic fox module includes the two climate impacts paths likely involved in the population decline. Both are due to icier snowpack in warmer winters increasing mortality in herbivores. Path 1 describes how a decrease in the abundance of key prey (lemming) implies lost opportunities for Arctic fox reproduction. 

Path 2 describes how abundant reindeer carrion subsidizes an increase in the population of red fox (key natural enemy), which ultimately implies the competitive exclusion of the Arctic fox. Analyses of 15 years of monitoring data have provided evidence for both of these paths (see further reading).

The conceptual model originally included two potential management intervention paths (see Path 3 involves reindeer management to reduce the amount of reindeer carrion by reducing the size of herds, while path 4 involves culling the population of red fox. Culling of red fox was implemented in 2005. Although it had some positive effect on the use of the area by Arctic fox, this management intervention was not sufficient to rescue the Arctic fox population, which was estimated at only five individuals in 2016 (Ims et al 2017).

The Arctic fox module has a reference group (consisting of researchers, stakeholders and management authorities) that advised the Norwegian Environment Agency in 2017 to implement two additional management interventions. Consequently, during 2018 and 2019 a total of 53 captive-bred Arctic foxes were released (path 5) and supplementary food (path 6) was provided at Arctic fox breeding dens.


This article was originally published in the Fram Forum



The Barents Observer Newsletter

After confirming you're a real person, you can write your email below and we include you to the subscription list.

Privacy policy