Antarctic sea ice extent has changed substantially over the satellite observation period since 1979. Total sea ice extent experienced a small but significant linear increase until 2016 when it plummeted from record highs in winter to record lows in summer. Up until recently, models have failed to capture the fluctuations of the highly variable sea ice extent. Fogt et al. (2022) and Maierhofer et al. (2024) created observation-based reconstruction ensembles since 1905 for seasonal and monthly data, respectively allowing the opportunity to put the contemporary variability of Antarctic sea ice into a historical context. This study compared the two datasets with a series of metrics. Each metric revealed the similarities and differences between the datasets and provided additional insight into pre-satellite sea ice behaviors, including extremes and temporal variability. In general, the datasets align better in sectors and seasons during sea ice maximums, with correlations up to 0.94. Additionally, the Maierhofer et al. (2024) dataset revealed interannual and decadal variability similar to that recorded in observations. Overall, both datasets suggest a general decrease in Antarctic sea ice in most sectors during the 20th century, except the Amundsen-Bellingshausen sector. However, there are distinctive differences between the two datasets which could be due to the different methods used to create the reconstructions, the initial resolution of the reconstructions, and the methods used to calculate the anomalies for the Fogt et al. (2022) data.

This study examines the influence of an atmospheric circulation pattern, known as the zonal wave three (ZW3), in terms of the sea ice’s seasons from 1979-2009 in order to better understand the response of the sea ice. An index to represent the amplitude of the ZW3 was calculated using zonal anomalies of 850 hPa geopotential heights taken from the ERA-Interim data set. Statistical analysis showed sea ice concentrations, taken from the Hadley sea ice and sea surface temperature data set, to be significantly dependent upon the ZW3 in parts of the Ross Sea, the ice edge in the Amundsen-Bellingshausen Seas and off the Amery ice shelf during the ice advance season. The results suggest that the ZW3 plays a role in the occurrence of the observed sea ice trends in the Ross Sea, Amundsen-Bellingshausen Seas, Weddell Sea and off the Amery ice shelf regions during the ice advance season, the critical period for sea ice growth.

Open water and thin ice areas, known as coastal polynyas, form along the Antarctic coastline and allow continued interaction between the ocean and atmosphere throughout the sea ice advance season. Coastal polynyas are the most productive locations of sea ice formation, Antarctic bottom water formation, and biological activity in the Southern Ocean. Changes in these elements are greatly controlled by polynya area variability. To carry out an in-depth study of polynya area variability, a 26-year 25-polynya daily area dataset was created and analyzed. The long term trend and the daily, monthly, seasonal, and annual variations are separated to analyze the multi-temporal variability of the polynyas and investigate their individual and regional responses to prominent large-scale atmospheric circulation patterns. Results indicate that most polynya variability occurs at the daily scale, followed by monthly and seasonal variations. Very little variability occurs interannually. Thus, studies done at the annual scale mask most of the polynya activity. Only five of the polynyas have long term trends, which are all non-linear and arise from abrupt changes in the icescape. Three of the significant trends occur within the top four most significant regions of sea ice and bottom water formation. Long term changes in polynya area cause long term changes in the overall productivity of the Southern Ocean. The Southern Annular Mode (SAM), El Nino-Southern Oscillation (ENSO), and the Amundsen Sea Low (ASL) significantly contribute to individual and regional coastal polynya variability. Influence from the SAM and the ASL is primarily driven at the monthly and seasonal scales. Influence from ENSO is driven at the annual scale. Using Pearson correlations, principal component analysis, gaussian mixture models, and hierarchical agglomerative clustering, six regional polynya groups are delineated based on the strength and direction of inter-polynya co-variability. The mean polynya variability within each region is significantly correlated, which is driven at the seasonal scale. While the SAM, ENSO, and ASL are not the primary drivers of regional polynya group delineations, they are significantly influential in the mean variability of each group.

General increases in Antarctic sea ice coverage occur primarily in the Ross Sea. This study investigates the Ross Sea Polynya's relationship with the Ross Sea ice areal coverage. A unique, relatively long term Ross Sea Polynya area dataset was created through the application of the Polynya Signature Simulation Method (PSSM) onto Special Sensor Microwave Imager (SSM/I) data inputs. Bivariate regression analyses were used to determine the relationships, at the 95% confidence level, between Ross Sea Polynya and ice areal trends, annual seasonalities, and anomalies at the full temporal scale as well as the monthly level. Polynya and sea ice have significant positive relationships in the late austral summer and early spring (February to March), and a significant negative relationship in the late austral winter (August). The areal anomalies only had a significant relationship in February, while the trends were not correlated at any time.