In August and September 1991, thickness, structure, and properties of level multiyear ice were studied at 66 locations in the Eurasian sector of the Arctic Ocean. The mean ice thickness was 2.86 m, with 0.31 m of freeboard (including 0.08 m deteriorated ice of mean density 370 kg m−3). On the basis of the study of ice cores, 61% of the ice cover consisted of undeformed columnar ice, the remaining 39% consisted of a mixture of ice types including frazil (18%) and deformed ice (9%). Through microstructural studies, six main classes of pores could be identified. The mean density of the ice cover increased from 720 kg m−3 at the top to >880 kg m−3 below 0.4 m depth. Sea ice salinities (mean value 2.1‰) correlate with ice thickness. On average, salinity profiles exhibit a linear increase from values close to 0‰ at the top to 2‰ at 1 m depth, with less steep salinity gradients below. Sampling from different depths within the ice cover indicates that the brine in summer sea ice is strongly stratified. The influence of meltwater percolation is evident, with salinities around 5‰ and pH values <8 at the top and >15‰ and >8 at greater depths. Brine volumes (ranging from 80 to 150‰) are controlled by the thermodynamic equilibrium between the solid and liquid phases. Gas volumes decrease from >200‰ at the top to <50‰ below 1 m depth. Pore microstructure is highly variable even on small scales. Salinity and other ice properties do not vary to a large degree between different regions. The evolution of level multiyear sea ice is discussed with particular reference to “hidden” occurrence of deformed ice and the importance of ablation processes.

We studied small‐scale (50 m to 5 km) sea ice deformation from ship radar images recorded during the N‐ICE2015 campaign. The campaign consisted of four consecutive drifting ice stations (Floes 1–4) north of Svalbard, with a total duration of nearly 5 months. Deformation was calculated using 5 different time intervals from 10 min to 24 h, and the deformation rate was found to depend strongly on the time scale. Floes 1–3 had a mean deformation rate within the range of 0.06–0.07 h−1 with the interval of 10 min, and 0.03–0.04 h−1 with the interval of 1 h. Floe 4 represented marginal ice zone (MIZ) with very high deformation rate, 0.14/0.08 h−1 with the interval of 10 min/1 h. Deep in the ice pack, high deformation rates occurred only with high wind and drift speed, while in MIZ they were found also during calm conditions. The deformation rates were found to follow power law scaling with respect to length and time scale even on this small scale and in small domain (15 km × 15 km). The length scale dependence of deformation rate depends on the time scale: the power law scaling exponent β of the whole study period decreases from 0.82 to 0.52 with the time interval increasing from 10 min to 24 h. Ship radar images reveal the importance of the deformation history of the ice pack, since the deformation events were initialized along the lines of previous damages.

No abstract

No abstract

Abstract. Rafting and pressure ridging are important processes in the deformation of sea ice that occur when two ice sheets are pushed together. In this study a two-dimensional computer model of the rafting and ridging process is used to simulate a situation in which two identical ice sheets are pushed together at constant speed. Each model ice sheet is composed of two thicknesses of ice. The ratio of the thicknesses is varied to obtain degrees of inhomogeneity. The accuracy of the simulations is assessed by comparison with a series of similar physical experiments performed in a refrigerated basin. Following this comparison, the computer model is used to perform an extensive series of simulations to explore the effect of the thickness and the thickness inhomogeneity of the model ice sheets on the likelihood of occurrence of ridging and rafting. During the simulations the energy consumption and forces are explicitly calculated. The energy consumed during the simulations is used to demonstrate the smooth transition between ridging and rafting that occurs when the homogeneity of the sheets is varied. IntroductionThe between the unit weight and buoyancy of either sheet and the length of the overlap between the sheets. In simulations of ridging between relatively thin first-year ice and a multiyear floe, the lead ice overrode the floe [Hopkins, 1994[Hopkins, , 1998]. Because of the great buoyancy of the floe the lead ice sheet was lifted from the water, the rafting force rose rapidly, and, consequently, the sheets were unable to progress more than 10 or 20 m. In addition, the large degree of curvature required of the lead ice sheet in mounting the floe facilitated buckling, which also helped to terminate the rafting event. In contrast, in rafting between sheets of roughly equal thickness, the unit buoyancy of the submerged ice sheet is much less and curvature is less. Therefore the rafting force increases more slowly, and rafting progresses much farther. Weeks and Anderson [1958] describe a rafting episode 600 m in length.In this study we consider rafting and ridging between two identical sheets. We began by trying to simulate ridging between identical sheets of uniform thickness. It quickly became apparent that rafting was the preferred mode between uniform sheets. Even simulations with rather thick sheets (0.5-0.9 m), which began by creating rubble, tended to raft as one sheet was lifted by the rubble. Concurrent ice basin model tests by Tuhkuri and Lensu [1998], in which two identical ice sheets were pushed together, also showed that uniform ice sheets tend to raft. Our attention moved to thickness inhomogeneity as the factor that determines the relative likelihood of ridging and rafting between identical ice sheets. In this study we push together two identical model ice sheets each composed of two thicknesses, as shown in Figure 1. In nature an ice sheet composed of two thicknesses might be created from the breaking, dilating, and refreezing of a uniform sheet. Two pressure ridges formed from composite ice shee...