Since 2016, successive record lows in Antarctic sea ice extent have raised concerns that the Southern Ocean sea ice system may be undergoing a significant shift. While atmospheric forcing has historically driven interannual sea ice variability, recent extreme ice loss appears increasingly linked to anomalous heat within the polar Southern Ocean mixed layer. However, the source of this heat remains uncertain.

This study investigates three potential pathways for upper-ocean warming associated with Antarctic sea ice decline: vertical entrainment of Circumpolar Deep Water, lateral heat advection, and enhanced solar absorption through the ice–albedo feedback. It is motivated in particular by the anomalously low winter sea ice observed in the Weddell Sea in 2023, with a focus on identifying the sources of upper-ocean heat contributing to this event. Using a combination of reanalysis products and gridded observational datasets, I will present initial results that quantify the relative importance of these mechanisms and identify the dominant drivers of recent upper-ocean warming and sea ice loss. By constraining the sources of mixed-layer heat, this work aims to improve understanding of the processes driving recent Antarctic sea ice decline and their implications for future variability and change.

ACCESS-OM3 is being extended to include three-way coupling between ocean circulation, sea-ice evolution, and surface waves for the forthcoming global 100 km MOM6-CICE6-WAVEWATCH III (MCW) configuration. This work describes recent developments in the fully coupled MCW system, implemented through NUOPC. The coupling enables three-way interaction among MOM6, CICE6, and WW3, exchanging key prognostic and diagnostic fields including ocean currents, mixed-layer depth, sea-ice concentration, ice thickness, floe-size diameter, and wave spectral information. A key new development is two-way CICE6-WW3 coupling for floe-size distribution: the WW3 surface elevation spectrum drives wave-induced fracture in CICE6 and updates floe sizes, while the evolving CICE6 sea-ice concentration, thickness, and floe-size distribution constrain wave attenuation in WW3. The coupled configuration also represents wave-driven ocean mixing in MOM6 through Langmuir-enhanced turbulence. Evaluation against WHACS shows lower significant-wave-height bias with ERA5 forcing than JRA55-do, while near-island positive biases point to limitations from diffusive propagation and unresolved obstacles. Mean wave period remains negatively biased, consistent with under-represented swell. The 100 km MCW configuration improves Antarctic sea-ice extent relative to earlier tested wave-mixing options and reproduces realistic temporal variability in wave penetration into sea ice. These results inform final tuning and validation of the ACCESS-OM3 100 km MCW release, enabling future studies of wave-ice-ocean feedbacks, marginal ice zone processes, and the role of wave-induced mixing in Southern Ocean dynamics.

Antarctic GPS time series are used to inform geophysical models notably those of Glacial Isostatic Adjustment (GIA), ice-mass change, and Earth rheology. Their interpretation is impacted by several sources of uncertainty, among which Surface Mass Balance (SMB) loading corrections and terrestrial reference frame realisation are particularly important. In this study, we explore both aspects by first evaluating a range of SMB products, including RACMO2.3p2 (27km), statistically downscaled RACMO2.3p2 (2km), MAR (35km), GEMB (10km), HIRHAM5 (12.5km) and MERRA2 (12.5km), by comparing them with estimated GPS-derived vertical bedrock displacements. We then quantify the impact of different reference-frame realisation strategies on Antarctic GPS timeseries, including the use of IGS20, JTRF2020 as well as alternative frame realisation approaches.

Results show that, for most Antarctic GPS sites considered, all considered SMB models reduce long-period (>1.5 years) variance on average, but performance varies across Antarctic regions and GPS sites. Among the models, RACMO variants perform the best, with RACMO_2DS and RACMO_11 showing the largest variance reduction. Preliminary analysis of different frame realisation strategies suggests a small impact on time series noise properties at low frequencies for Antarctic sites. Some systematic differences are observed and quantified.

The ACCESS-NRI Ice Sheet Modelling team is developing the capability to dynamically couple the Ice-sheet and Sea-level System Model (ISSM) within the ACCESS Earth System Model. This work is being advanced through two complementary efforts: coupling ISSM to the ACCESS-AM3 atmosphere model at Monash University, and to the ACCESS-OM3 ocean model at the University of Tasmania.

Ice sheets are the final major Earth system component yet to be integrated into the ACCESS coupled modelling suite. Their inclusion is essential for representing Antarctic contributions to sea-level rise, ocean circulation, and climate feedbacks. However, unlike ocean-atmosphere or land-atmosphere interfaces, ice-sheet boundaries move through grounding-line migration, calving-front retreat, surface ablation, and basal melting, changing how the ice sheet interacts with the atmosphere and ocean.

Using the ACCESS NUOPC coupling framework, we are building flexible infrastructure to exchange fields between ISSM and existing ACCESS components while accounting for evolving ice-sheet geometry. Initial development focuses on idealised ice-ocean and ice-atmosphere configurations, providing controlled testbeds for coupling strategies, moving boundaries, timestep coordination, and flux exchange. These experiments will guide the transition to realistic Antarctic simulations using the ACCESS-AIS3 Antarctic ISSM community configuration.

This work establishes the foundation for independent ice-ocean and ice-atmosphere coupling within ACCESS, ultimately leading to a fully coupled Earth system model with dynamically evolving Antarctic and Greenland ice sheets.

Coleoid cephalopods (octopuses, squids, and cuttlefishes) are recognised for their complex nervous systems, behavioural traits, and camouflage capabilities. Genomic research, including genome assemblies, is increasingly revealing the molecular basis and evolutionary history of these traits. In cephalopods, genome assemblies have identified genomic features associated with neural complexity and adaptation. Comparative genomic studies have also revealed genomic variation among species, providing insights into how environmental conditions may have influenced genome evolution. Antarctic octopods have contributed to studies of Southern Ocean and deep-sea evolutionary history, cold adaptation, and Antarctic ice sheet history. Approximately 33 octopod species from six families occur in Antarctic waters, with Megaleledonidae representing the most diverse family and including endemic genera such as Adelieledone that have evolved in situ for more than 30 million years. However, no genome assembly is currently available for an Antarctic octopod. Early results from the genome assembly of Adelieledone adelieana estimate a genome size of ~7 Gb, placing it among the largest octopod genomes assembled to date. Ongoing work aims to generate a chromosome-level assembly and annotation to investigate Antarctic genome evolution, compare Antarctic octopods with temperate and tropical species, support studies of contemporary connectivity and diversity, and provide insights into Antarctic ice sheet dynamics.

A census of the southern elephant seal (Mirounga leonina) subpopulation at Heard Island has been conducted only 13 times since 1948. There had been a 30 year gap in data collection prior to October 2025. We used ground and aerial platforms to survey the island’s entire coastline for breeding female southern elephant seals. After applying correction factors to account for detection biases, we estimate that approximately 17,500 breeding females gave birth on Heard Island during the 2025 breeding season. Analysis of the data between 1948 – 2025 indicates that the population has recovered from a decline that reached a low point in the mid-1980s, followed by a brief period of stability in the early 1990s. However, current population levels remain below those recorded in the early 1950s.

Investigating how plant forms and functions vary across different environments can be informative of how these plants are likely to respond to climate changes. Sub-Antarctic Macquarie Island is experiencing changes in a range of climate variables, however we currently lack a fundamental understanding of the trait-environment relationships (TERs) that shape the island’s various plant communities. Importantly, contextual factors associated with the distinct climate and evolutionary history of the island can result in different selective pressures and trait trade-offs, leading to TERs that differ from other global regions. Using a community-weighted mean approach, we quantified variation in a set of key plant functional traits related to resource economics, competition, and size, across a range of environmental gradients on Macquarie Island. While we found evidence of strong TERs, this varied widely between traits, with leaf dry matter content being the trait most closely associated with environmental conditions. In addition, many of these TERs deviated from what might be expected based on theory and findings from other global regions. Temperature was found to be a key variable driving variation across all traits, however other climate variables such as insolation, as well as several soil variables, were also consequential. By placing these TERs in the context of ongoing climate changes, we develop some predictions of how the Macquarie Island’s vegetation may respond to these changes, including potential constraints.

Environmental DNA (eDNA) analysis holds immense promise for monitoring vast ecosystems like the Southern Ocean. However, the lack of reliable, appropriate DNA reference data required to interpret eDNA remain one of the greatest blockers to the uptake of these approaches generally, and a particular challenge for these remote waters. The National Biodiversity DNA Library (NBDL) is a CSIRO-led initiative to support Australia's environmental monitoring responsibilities. With our partners, the NBDL is generating uniquely comprehensive and authoritative DNA reference sequences derived from expertly identified specimens held in Australia’s research collections. CSIRO has developed a cost-effective highly miniaturised genome-skimming technology ideally suited to collection specimens to build this library, and thousands of mitochondrial genomes are now publicly available through our dedicated portal launched in late 2025. Working with the Minderoo Foundation, Australian Antarctic Division, Tasmanian Museum and Art Gallery, Securing Antarctica’s Environmental Future, and James Cook University, among others to date, the NBDL are now working to generate verified reference data to power the uptake of eDNA technologies in Antarctica and the Southern Ocean. As part of this effort, the NBDL is advocating for community support for more direct pathways for collected specimens to be accessioned into collections to support this work and further research applications into the future.

The Southern Ocean ecosystem has been shaped by nearly two centuries of human exploitation, from industrial whaling and sealing to krill fishing. Currently, many Southern Ocean predator populations are characterized by steep population trajectories emblematic of a system undergoing change. To understand the role of recovering baleen whale populations in Southern Ocean ecosystems, we estimate abundance for fin, humpback and Antarctic minke for the Antarctic Peninsula region. We developed a cost-effective method for collecting distance sampling data from Antarctic tourist vessels over 4 Antarctic summer (Nov-Mar) sampling seasons (2021/2022 - 2024/2025). We estimate summer abundance for fin (28,345, CV = 0.34), humpback (32,990, CV = 0.12) and Antarctic minke (2,024, CV = 0.33) whales which suggested an intrinsic annual abundance increase across the last two decades (since the year 2000) for fin +13.1% p.a. (95% CI: +7.1-19.3% p.a.) and humpback +6.7% p.a.(95% CI: +2.5-11.0% p.a.) whales and an intrinsic annual decrease for Antarctic minke whales -2.7% p.a.(95% CI: -6.3-1.1% p.a.). Spatial predictions of density indicate that fin, humpback, and Antarctic minke whales forage in discrete, largely non-overlapping regions, consistent with body-size-structured habitat partitioning which is potentially the result of competition among whale species. Abundance trends mirror this pattern, with fin whales, the largest and presumably most efficient foragers, showing most rapid increase, humpback whales showing moderate increase, and Antarctic minke whales likely declining.

Climate change is warming the global ocean, and ocean heat is melting marine ice shelves, reducing their buttressing effect on the grounded ice sheet. Understanding ice-shelf fracture mechanisms and stability is therefore essential for improving projections of sea-level rise. Numerical methods have become powerful tools for investigating ice-shelf evolution and calving; however, traditional continuum ice-flow models cannot explicitly capture fracture processes. While discontinuum methods are effective for simulating fracture, they cannot capture the stress evolution associated with ice flow. To overcome these limitations, this study develops a three-dimensional hybrid finite-discrete element method (HFDEM3D) to simulate crack initiation, fracture propagation and stress evolution. A simplified hydrodynamic model is incorporated to represent buoyancy and drag force, and a strength reduction method is developed to simulate degradation, damage, fracture, and calving. Preliminary numerical simulations have been conducted to investigate the possibility of evaluating marine ice shelf stability with a factor of safety linked to fracture and calving behavior under different seawater levels. The simulated stress redistribution, fracture mechanism, calving process, retreat distance, and fragment size are compared with previous studies. Preliminary results indicate that HFDEM3D has potential as an effective tool for analyzing ice-shelf fracture processes and calving behavior.

Thermal infrared images of the cloud-free surface of the Antarctic ice sheet exhibit subtle stream-like features that can be traced over hundreds of kilometres. Visible and infrared bands of Landsat images clearly show the features are related to plumes of drift snow aligned along the direction of the surface wind. So, each satellite image provides an instantaneous wide-area map of the surface wind-field over a large part of the Antarctic continent. The stream features result from the interaction of a surface temperature inversion with spatially varying density and thickness of drift snow in the atmospheric boundary layer. They are superimposed on broader features in the temperature field related to the large-scale topography of the ice sheet. Interaction of the surface wind with meso-scale topography is a key modulator of the snow accumulation distribution.

In this work, the stream-features have been mapped from images acquired over many months and accumulated as time series of wind vectors projected onto a regular spatial grid covering the continent. The wind directions vary systematically, both spatially and temporally. At a grid point the more frequently occurring directions are related to the (cold-air-drainage) katabatic wind flow. Departure from this direction is related to the influence of passing atmospheric synoptic systems. The derived wind-field directions compare well with contemporaneous surface wind observations from Automatic Weather Stations.