Jump to content Jump to footer

News detail view

MOSAiC - Linking scientific results (Part 2)

Ocean Science Arctic

Arctic atmosphere, ice and ocean results from MOSAiC published. What’s happening in the Arctic Ocean under the sea ice: seasonal changes, regionality, eddies and mixing throughout a full year during MOSAiC.

Many have seen pictures of the Arctic icescape in the sun, or under the moon during the dark winter months north of the polar circle. Yet, have you ever wondered what it all looks like underneath the surface? Scientists from around the globe took part in the MOSAiC drift expedition by deploying autonomous devices or manually taking measurements in the field for an entire year (see the drift route in Figure 1). You can see some of their equipment and a glimpse under the ice in Figure 2. The team of more than 50 scientists investigated how warm or cold the ocean was under the sea ice, how isolated the lower parts of the water column were from the sea ice and atmosphere, and how currents moved throughout the ocean.

They assessed the effects of fast and slow ice drift, which are caused by changing winds near the surface and produce “turbulent mixing” in the water. They measured the structure of whirls referred to as eddies, only a few kilometres across, using manual profiles near RV Polarstern and autonomous instruments in a 50-km radius around the ship. In winter, they observed the changes in the water layer under the ice, with temperatures close to freezing. In summer, they studied shallow lenses of meltwater that formed after leads had opened in the sea ice.

But why did they do all this? Currently, the Arctic Ocean is less observed than many other ocean regions in the world. Part of the reason is the difficulty in accessing the ice-covered part, especially in winter; other factors include the lack of regional satellite observations of the ocean, which can only, if at all, be taken from the few leads in the sea ice. And even autonomous instruments have to be adapted to this harsh environment. Yet, the Arctic is now undergoing the most extreme climate change in the world, warming at up to three times the global rate, as can be seen not only in the atmosphere and sea ice, but also in the ocean.

A first overview

A summary of the work done by the physical oceanography team, or “OCEAN Team”, has recently been published in the journal Elementa (Rabe et al., 2022).  Like the rest of MOSAiC, the OCEAN section has delivered the most comprehensive set of physical oceanography measurements ever taken on and around a drifting ice floe in the ice-covered Arctic Ocean (Figures 1 and 2). The team’s observational programme was aimed at assessing the basic state of the ocean by measuring e.g. the temperature and salinity throughout the entire water column on a weekly basis, and from the upper water column down to the warm Atlantic water on a daily basis. “It was challenging, since there were at most five people onboard at a time to carry out these measurements. This included making holes in the ice so that we could lower the profiler even in the middle of winter, but also maintaining the equipment located far from the ship, collecting countless water samples, etc.,” says Benjamin Rabe, co-cruise leader of Leg 2 and co-leader of the OCEAN Team, summarising the work done during the expedition.

Velocity and mixing

During MOSAiC, we also measured the horizontal velocity in the ocean for the top 500 m of the water column. Ideally, we would have measured the velocity all the way to the seafloor, but our equipment wasn’t capable of doing so. The measurements confirmed the impression of a generally quiet Arctic Ocean, with horizontal velocity usually below 5 cm/s. By comparison, the Gulf Stream can reach nearly 2 m/s near the surface and slows down to 40 cm/s as it widens to the north. However, we did observe these velocities increasing in parts of the Eurasian Basin north of Fram Strait and in the Central Arctic in late summer (Figure 1), likely due to looser ice cover in the respective regions. These observations are currently being analysed to study the role of internal waves in initiating mixing, enhanced wind and ice motion, and the effects of decreased stratification and tides north of Svalbard in the Nansen Basin, close to the inflow region of the warm Atlantic water, along with the variability of these processes. Said processes are important to understand because of their role in the diminishing Arctic sea-ice cover and changes in upper-ocean dynamics that can affect not only the physics, but also aspects like gas exchange between the ocean and atmosphere, or the ecosystem.

Temperature, salinity and circulation

Observations of temperature and salinity can give us an indication of the different water layers with origins inside and outside the Arctic. The vertical slices (“sections”) of the top 200 m of the ocean along the drift trajectory from October 2019 to June 2020 (Figure 3; x-axis is the time, y-axis is the depth) show us how close the upper-ocean mixed layer (dark grey area in Figure 3) was to the underlying layer near the beginning, north of the Laptev Sea. This underlying layer (light grey in Figure 3) marks a strong salinity gradient referred to as the “lower halocline”, which acts as a barrier between the warm Atlantic water (warmer than 0° C, black area in Figure 3) and the mixed layer directly below the ice, which is the connection between the surface and deeper waters. More simply put, we could see the “Atlantification” of the Arctic – that is, the replacement of the halocline with Atlantic water – with our own eyes when we were there. “This Atlantification affects how deep the upper ocean mixes and, thus, how heat content and nutrients come to the surface,” adds Benjamin Rabe.

The circulation of the different types of water shown in Figure 3 can further be inferred from the observed temperature and salinity along the drift, in relation to historical observations dating all the way back to the first scientific drift in the Arctic by Nansen in 1893–1896, along with samples of dissolved gases that we collected and that act as a tracer for the water ... but that’s a story for another publication (several are in preparation, in fact).

Summer melting

In early summer, small-scale measurements of salinity with a mobile instrument mounted on a fishing rod detected a shallow meltwater lens close to the surface and only a few metres thick (Figure 4). Such lenses are crucial to understanding and modelling sea-ice melting, as they temporarily separate the ice from the ocean below. This lens disappeared again quickly after a storm passed. This small-scale, short-lived phenomenon had been hypothesized before, but we were one of the first to study its development (and film it!) first-hand. “This was such a fantastic coincidence! We had planned from the beginning to study the effects of ice opening on the ocean, but we never thought we’d be so lucky and see this thin meltwater layer so easily,” adds Céline Heuzé, the other co-leader of the OCEAN Team.

Summary and outlook

The measurements taken by the OCEAN Team were coordinated with those of several other teams, who focused on the sea ice and snow, the atmosphere, biogeochemical processes and the ecosystem, respectively. This multidisciplinary “Earth system” approach is necessary if we want to fully understand all the processes at play, since they are all interconnected.  Further coordination with satellite-based measurements and numerical modelling allows us to analyse our observations more comprehensively. “Thus, our ongoing and future study of the MOSAiC observations will help us understand the Arctic Ocean in the regional and global climate system, and to improve numerical models for predicting changes in the region and beyond,” concludes Heuzé, who is still amazed by this enormous group effort and its fantastic outcomes.


Rabe, B, Heuzé, C, Regnery, J, Aksenov, Y, Allerholt, J, Athanase, M, Bai, Y, Basque, C, Bauch, D, Baumann, TM, Chen, D, Cole, ST, Craw, L, Davies, A, Damm, E, Dethloff, K, Divine, DV, Doglioni, F, Ebert, F, Fang, Y-C, Fer, I, Fong, AA, Gradinger, R, Granskog, MA, Graupner, R, Haas, C, He, H, He, Y, Hoppmann, M, Janout, M, Kadko, D, Kanzow, T, Karam, S, Kawaguchi, Y, Koenig, Z, Kong, B, Krishfield, RA, Krumpen, T, Kuhlmey, D, Kuznetsov, I, Lan, M, Laukert, G. Lei, R, Li, T, Torres-Valde  ́s, S, Lin, L, Lin, L, Liu, H, Liu, N, Loose, B, Ma, X, MacKay, R, Mallet, M, Mallett, RDC, Maslowski, W, Mertens, C, Mohrholz, V, Muilwijk, M, Nicolaus, M, O’Brien, JK, Perovich, D, Ren, J, Rex, M, Ribeiro, N, Rinke, A, Schaffer, J, Schuffenhauer, I, Schulz, K, Shupe, MD, Shaw, W, Sokolov, V, Sommerfeld, A, Spreen, G, Stanton, T, Stephens, M, Su, J, Sukhikh, N, Sundfjord, A, Thomisch, K, Tippenhauer, S, Toole, JM, Vredenborg, M, Walter, M, Wang, H, Wang, L, Wang, Y, Wendisch, M, Zhao, J, Zhou, M, Zhu, J. 2022. Overview of the MOSAiC expedition: Physical oceanography. Elementa: Science of the Anthropocene 10(1). DOI: doi.org/10.1525/elementa.2021.00062 .



Contact us via E-Mail or our contact form.

Pictures showing various instruments and installations used for ocean measurements during the expedition.

Figure 2: Pictures showing various instruments and installations used for ocean measurements during the expedition. (a) shows a tent with a buoyant and reinforced floor that was used during winter and into early summer on the ice floe a few hundred metres from RV Polarstern, seen here from outside during the Polar Night. (b) offers a view from inside the tent as an instrument used to measure water properties such as temperature and salinity and collect water samples (“rosette”) is being lowered into the water through a hole in the ice. In late summer, a mobile tent (c) was used to carry out similar work with all but the heaviest equipment. A larger version of the rosette was deployed from on board the ship through a hole in the ice, with a hood around the instrument while not in the water to protect it from the sub-freezing temperatures prevalent during the first half of the expedition (d). Photo credits: Ying-Chih Fang (a), Esther Horvath (b), Mario Hoppmann (c) and Janin Schaffer (d). Figure modified from Rabe et al. (2022; Figures 5 and 6).