Olga Pavlova

Female polar bears, Ursus maritimus, on the Barents Sea drift ice: walking the treadmill

Abstract For animals in dynamic habitats, the contribution of passive (i.e. by wind or current) and active (movements by the animals themselves) displacement determines whether their space use reflects physical or adaptive behavioural processes. Polar bears in the Barents Sea undertake extensive annual migrations in a habitat that is highly dynamic because of continuous sea ice drift. Using combined information from satellite telemetry, satellite images and atmospheric pressure recordings, we estimated the contribution of sea ice drift and movements in the monthly net displacement of female polar bears. We found that movements, and thus behavioural processes, were dominant. Net displacement was directed northwards during summer ice retreat and southwards during winter ice advance. Conversely, movements were directed northwards counteracting a continuous southward drift. Acting as a treadmill, ice drift probably increased the energetic cost of migrations relative to that expected from observed net displacement distances; this suggests that pelagic and adjacent near-shore bears, on stable land-fast ice, have different energy costs. Little concordance between ice drift rates and net displacement and movement rates suggest that polar bears do not adjust their displacement relative to attractive areas with fixed locations, but rather adjust their movements to local habitat suitability. Furthermore, selective use of less dynamic drift ice when with cubs-of-the-year, and use of terrestrial denning areas, appear to be behavioural adaptations to the dynamics of the Barents Sea drift ice. Hence, understanding the behaviour and ecology of animals inhabiting dynamic habitats necessitates incorporation of both dynamic and static habitat variables. Copyright 2003 Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour.

Assessment of potential transport of pollutants into the Barents Sea via sea ice—an observational approach

The present estimates of ice drift in the Arctic include utilization of satellite imagery data (special sensor microwave/imager) and a reconstruction of air pressure for the period 1899-1998. A significant part of the sea ice in the Arctic Ocean has its origin in the Kara Sea and melts in the Greenland and the Barents Sea (BS). Consequently there may be a particular risk of pollutants in the Kara Sea entering the food webs of the Greenland and BS. The ice export from the Kara Sea between 1988 and 1994 was about 208,000 km2 (154 km3) per year. The import of ice into the BS was during the same period 161,000 km2 (183 km3) per year while the ice drift through the Fram Strait into the Greenland Sea was 583,000 km2 (1859 km3) per year. Ice which formed adjacent to the Ob and Yenisey rivers in early January, drifted into the BS within two years (with a probability of about 50%.

Modelling snow ice and superimposed ice on landfast sea ice in Kongsfjorden, Svalbard

Snow ice and superimposed ice formation on landfast sea ice in a Svalbard fjord, Kongsfjorden, was investigated with a high-resolution thermodynamic snow and sea-ice model, applying meteorological weather station data as external forcing. The model shows that sea-ice formation occurs both at the ice bottom and at the snow/ice interface. Modelling results indicated that the total snow ice and superimposed ice, which formed at the snow/ice interface, was about 14 cm during the simulation period, accounting for about 15% of the total ice mass and 35% of the total ice growth. Introducing a time-dependent snow density improved the modelled results, and a time-dependent oceanic heat flux parameterization yielded reasonable ice growth at the ice bottom. Model results suggest that weather conditions, in particular air temperature and precipitation, as well as snow thermal properties and surface albedo are the most critical factors for the development of snow ice and superimposed ice in Kongsfjorden. While both warming air and higher precipitation led to increased snow ice and superimposed ice forming in Kongsfjorden in the model runs, the processes were more sensitive to precipitation than to air temperature.

Time variability in the annual cycle of sea ice thickness in the Transpolar Drift

The annual cycle of modal and mean sea ice thickness was derived from upward looking sonar ice thickness observations (1990–2011) in Fram Strait. The average annual peak-to-trough amplitude of the mode of 0.54 m is superimposed on interannual variability with peak-to-trough amplitudes of 0.73 m on time scales of 6–8 years, which again is superimposed on a long-term trend of −0.55 m/decade over the observation period. The long-term trend is stronger for April than for August, the average months of maximum and minimum modal thickness. As a result, the annual peak-to-trough modal thickness amplitude was reduced by 30% between the 1990s and the 2000s. The average annual peak-to-trough amplitude of the mean ice thickness of 1.20 m is also superimposed on interannual variability, with as much as 0.97 m thickness change over only 3 years. These two modes of variability are superimposed on a long-term trend of −0.35 m/decade through the entire data set. In contrast to the modal thickness, the long-term trend is weaker for the average month of maximum mean thickness (June), than for the average month of minimum (September). Therefore, the annual peak-to-trough amplitude of the mean ice thickness increased by 14% between the 1990s and the 2000s.

Convective Instability in the Ice Edge Area of the Barents Sea

The ice edge of the Barents Sea east of Svalbard is an area where the warm, salty water of the North Atlantic (AtW) interacts with cold, less dense, saline Arctic water (ArW) and the water produced by melting ice (MIW). Many of the CTD profiles (CTD stands for Conductivity-Temperature-Depth) obtained in this region by Norwegian Polar Institute expeditions in 1999 and 2007 contain layers that are quasi-homogeneous in temperature, salinity and density between the depths of 5-7 m to 100-150 m. It is shown that these features are formed by convective instability due to double-diffusion, which can occur where there are positive vertical gradients of both temperature and salinity, as is observed in this region. The rate of development and the thickness of the gradient layer depend on vertical temperature and salinity drops in the zone of interaction of AtW with ArW and MIW. They correspond well, characterized by a correlation coefficient of 0.96.