Several studies have suggested that the carbon sink in the Southern Oceanthe ocean's strongest region for uptake of anthropogenic CO 2 -has weakened in recent decades. Here, we demonstrate on the basis of multi-decadal analyses 1

Recent salinity changes in the Southern Ocean are among the most prominent signals of climate change in the global ocean, yet their underlying causes have not been firmly established. Here we propose that trends in the northward transport of Antarctic sea ice are a major contributor to these changes. Using satellite observations supplemented by sea-ice reconstructions, we estimate that wind-driven northward freshwater transport by sea ice increased by 20 ± 10 per cent between 1982 and 2008. The strongest and most robust increase occurred in the Pacific sector, coinciding with the largest observed salinity changes. We estimate that the additional freshwater for the entire northern sea-ice edge entails a freshening rate of -0.02 ± 0.01 grams per kilogram per decade in the surface and intermediate waters of the open ocean, similar to the observed freshening. The enhanced rejection of salt near the coast of Antarctica associated with stronger sea-ice export counteracts the freshening of both continental shelf and newly formed bottom waters due to increases in glacial meltwater. Although the data sources underlying our results have substantial uncertainties, regional analyses and independent data from an atmospheric reanalysis support our conclusions. Our finding that northward sea-ice freshwater transport is also a key determinant of the mean salinity distribution in the Southern Ocean further underpins the importance of the sea-ice-induced freshwater flux. Through its influence on the density structure of the ocean, this process has critical consequences for the global climate by affecting the exchange of heat, carbon and nutrients between the deep ocean and surface waters.

[1] We describe the evolution of Arctic sea ice as modeled by the Max Planck Institute for Meteorology's Earth System Model (MPI-ESM). The modeled spatial distribution and interannual variability of the sea-ice cover agree well with satellite observations and are improved relative to the model's predecessor ECHAM5/MPIOM. An evaluation of modeled sea-ice coverage based on sea-ice area gives, however, conflicting results compared to an evaluation based on sea-ice extent and is additionally hindered by uncertainties in the observational record. Simulated trends in sea-ice coverage for the satellite period range from more strongly negative than observed to positive. The observed evolution of Arctic sea ice is incompatible with modeled internal variability and probably caused by external forcing. Simulated drift patterns agree well with observations, but simulated drift speed is generally too high. Simulated sea-ice volume agrees well with volume estimates of the PIOMAS reanalysis for the past few years. However, a preceding Arctic wide decrease in sea-ice volume starts much earlier in MPI-ESM than in PIOMAS. Analyzing this behavior in MPI-ESM's ocean model MPIOM, we find that the modeled volume trend depends crucially on the specific choice of atmospheric reanalysis forcing, which casts some doubt on the reliability of estimates of volume trends. In our CMIP5 scenario simulations, we find a substantial delay in sea-ice response to increasing CO 2 concentration; a seasonally ice-free Arctic can result for a CO 2 concentration of around 500 ppm. Simulated winter sea-ice coverage drops rapidly to near ice-free conditions once the mean Arctic winter temperature exceeds 25 C. , 5, 173-194,

Abstract. In contrast to the Arctic, where total sea ice extent (SIE) has been decreasing for the last three decades, Antarctic SIE has shown a small, but significant, increase during the same time period. However, in 2016, an unusually early onset of the melt season was observed; the maximum Antarctic SIE was already reached as early as August rather than the end of September, and was followed by a rapid decrease. The decay was particularly strong in November, when Antarctic SIE exhibited a negative anomaly (compared to the 1979–2015 average) of approximately 2 million km2. ECMWF Interim reanalysis data showed that the early onset of the melt and the rapid decrease in sea ice area (SIA) and SIE were associated with atmospheric flow patterns related to a positive zonal wave number three (ZW3) index, i.e., synoptic situations leading to strong meridional flow and anomalously strong southward heat advection in the regions of strongest sea ice decline. A persistently positive ZW3 index from May to August suggests that SIE decrease was preconditioned by SIA decrease. In particular, in the first third of November northerly flow conditions in the Weddell Sea and the Western Pacific triggered accelerated sea ice decay, which was continued in the following weeks due to positive feedback effects, leading to the unusually low November SIE. In 2016, the monthly mean Southern Annular Mode (SAM) index reached its second lowest November value since the beginning of the satellite observations. A better spatial and temporal coverage of reliable ice thickness data is needed to assess the change in ice mass rather than ice area.