Satellite-based Analysis of Ocean-Surface Stress across the Ice-free and Ice-covered Polar Oceans

Abstract. Ocean-surface stress is a critical driver of polar sea ice dynamics, air-sea interactions, and ocean circulation. This work provides a daily analysis of ocean-surface stress on 25-km Equal-Area Scalable Earth (EASE) Grids across the ice-free and ice-covered regions of the polar oceans (2011–2018 for Arctic, 2013–2018 for Antarctic), covering latitudes north of 60° N in the Arctic and south of 50° S in the Antarctic and Southern Ocean. Ocean-surface stress is calculated using a bulk parameterization approach that combines ocean-surface winds, ice motion vectors, and sea surface height (SSH) data from multiple satellite platforms. The analysis captures significant spatial and temporal variability in ocean-surface wind stress and the resultant wind-driven Ekman transport, while providing enhanced spatiotemporal resolution. Two sensitivity analyses are conducted to address key sources of uncertainty. The first addresses the fine-scale variability in SSH fields, which was mitigated using a 150-km Gaussian filter to smooth three-day SSH datasets and enhance compatibility with the other monthly product, followed by linear interpolation to achieve daily resolution. The second investigates uncertainty in the ice-water drag coefficient, which revealed that variations in the coefficient have a proportional influence on the computed ocean-surface stress under the tested conditions. These uncertainties are most pronounced during winter, with median values reaching 20 % in the Arctic and 40 % in the Southern Ocean. Validation efforts utilized Ice-Tethered Profiler velocity records, revealing moderate correlations (r = 0.6–0.8) at monthly timescales, effectively capturing low-frequency signals but with small northward biases. Satellite-derived velocity fields, including both Ekman and geostrophic components, explain 40–50 % of the total variance. The unexplained variance reflects unresolved processes, such as mesoscale dynamics and other unparameterized factors. This dataset is publicly available at https://doi.org/10.5281/zenodo.14750492 (Liu & Yu, 2024).