Answers are grounded in this paper and cite their sources.
Ocean heat transport drives approximately half of Antarctic ice mass loss, yet in situ ocean observations near glacier ice fronts remain exceedingly scarce given the challenges of data collection in this remote environment. Remotely sensed sea surface temperatures (SST) offer a powerful alternative, but no reliable product resolves the fine spatial resolution needed to observe partially ice-covered coastal Antarctic seas. Here, we develop and validate an efficient, open-source, cloud-native single-channel SST algorithm for Landsat 8, designed for Antarctic conditions and transferable to previous Landsat sensors carrying a thermal infrared band. Cross-calibration with MODIS SSTs yields a root mean square error of 0.26 ∘ C. For comparison, we apply the existing Landsat Surface Temperature algorithm to the same scenes with an assumed ocean emissivity, since that product is not generated south of 60 ∘ S. This approach yields an RMSE ∼ 30% larger than our algorithm, consistent with the reduced accuracy expected for land-heritage algorithms over water. Independent validation using 6 Argo-float-derived and 81 seal-tag-derived measurements shows strong agreement, with an RMSE of 0.22 ∘ C for SST retrievals spanning from −1.9 to 1.0 ∘ C. Given the dearth of Argo observations, we find that seal-borne sensors represent an untapped resource for validating high-resolution polar SST, with no degradation in quality even with larger location uncertainties. Using this product, we present the first SST retrieval of a sub-kilometer Antarctic meltwater plume. This new Antarctic SST product provides a high-resolution thermal record extending back to 2013, opening new opportunities to study polar ice-ocean interactions.
No takes yet. Share an insight, caveat, or question.
Snow et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: