n Antarctica, areas of ice that have been compacted over time and are exposed by clearing of snow or through melt are called Blue Ice Areas (BIAs). Because of its compressed air-free nature, blue ice is smooth and has very low reflectance in longer wavelengths, distinguishing it from snow and other ice types. BIAs can be stable and longlived with some seasonal variation, mapping the Blue Ice over time helps us to understand the ice behaviour in a given region.

Digital Earth Antarctica is building a platform of analysis ready Sentinel-1 SAR and Landsat/Sentinel-2 optical data across Antarctica. Combining SAR and optical data, we can corroborate detections from one method with the other, and the SAR allows us to understand the blue ice behaviour when optical data is not available.

Blue Ice can be detected in optical data using the Normalised Difference Blue Ice Index, the Normalised Difference Snow Index and by thresholding values in the Infrared bands, also appearing blue in true colour images. Blue Ice appears smooth and dark in SAR backscatter imagery, with values falling between wet or dry snow, but can also look similar to melt, so combining the SAR with the optical improves our detection. We identified BIAs using our prototype data collections and consequently developed an automated BIA mapping algorithm.

The Amery Ice Shelf is the third largest ice shelf in Antarctica, and drains a large region of the East Antarctic Ice Sheet. Despite cold oceanic conditions within the ice shelf cavity, the deep draft and large area of the ice shelf mean that it produces a significant volume of meltwater, which in turn affects surrounding processes such as sea ice production and dense shelf water formation. The processes governing ice-ocean interactions under the Amery ice shelf have the potential to significantly affect both ice sheet dynamics, and regional ocean processes.
We present data from an in situ autonomous radar system (ApRES), which measured ocean-driven melt rates underneath the Amery Ice Shelf over a three-year period from 2015-2018. Measured melt rates show strong seasonality, with higher melt rates during austral summer, and lower melt rates during winter, although the precise timing of minimum and maximum melt varies interannually. We use a high-resolution regional ocean model to investigate the cause of this seasonal variability. Results show a seasonal shift in circulation pattern in the ice shelf cavity, which leads to cooler water temperatures and lower current speeds at the instrument site during winter. The timing of this shift is affected by local polynya activity, confirming the deep connection between different components of the ice-ocean-atmosphere system in this region.

Detection of glacier change, and early precursors of such change, is a vital step towards understanding and anticipating Antarctic tipping points. Passive seismic methods are a powerful observational technique as they allow hidden or transient processes to be recorded from the surface. Extracting and interpreting seismic signals of interest (such as those arising from subglacial lake drainage, calving, and stick-slip motion) therefore presents an opportunity for continuous glacier monitoring. Unsupervised classification methods applied to continuous seismic recordings are ideal for initial reconnaissance tasks by finding groups of signals of similar character, and extracting trends from large volumes of data.

In this work, we perform hierarchical clustering on the deep scattering spectra of seismic data from Totten Glacier, East Antarctica. We infer the evolution of glacier processes in the context of its environment (e.g. wind and tidal conditions), and separate transient event classes. Change detection is of particular interest at large outlets such as Totten Glacier, where identified stick-slip motion, potentially transient subglacial hydrology, and other subglacial processes are important for modulating ice flow.

Antarctica’s accelerating ice discharge is increasing the rate of global sea level rise, offset to some extent by enhanced snowfall over East Antarctica. Decomposing total mass balance estimates into surface mass balance and ice discharge components shows that reduced coastal snowfall in Wilkes Land is also contributing to mass loss. Modeling temporal noise with a Generalized Gauss-Markov model, rather than assuming white noise, results in far fewer significant accelerations in total mass balance across nearly all of Antarctica. Ice discharge has accelerated in the major West Antarctic outlet glaciers, yet there is no significant acceleration in Antarctic-wide total mass balance. A temporary snowfall-driven mass increase occurred from 2021 to 2024, interrupting long-term Antarctic mass loss. GRACE-FO observations since 2024 suggest that the Antarctic Ice Sheet is now in a state of approximate mass balance.

Accurate time-dependent models of the Antarctic ice sheet are central to global predictions of future climate and sea-level change. The ACCESS-NRI Ice Sheet Modelling team has developed an integrated, open-source workflow that makes running the Ice-sheet and Sea-level System Model (ISSM) simpler and more accessible for the Australian cryospheric research community. We present four connected components: pyISSM, a Python interface for easy setup of ISSM models, job submission to the National Computational Infrastructure (NCI) Gadi supercomputer, and visualisation of outputs, all from an NCI Australian Research Environment (ARE) Jupyter notebook; ACCESS-ISSM, an open-access bespoke ISSM build optimised for stable, performant runs on Gadi; the ACCESS Cryosphere Community Datapool (CCD), with its companion Python package ccdtools, a curated, searchable repository of datasets and model configurations commonly used to parameterise and evaluate ice sheet models and broader cryospheric research; and ACCESS-AIS3, an open, community-driven, whole-Antarctic ISSM configuration available to all researchers, forming the basis of ice sheet coupling efforts within the ACCESS climate model suite. Together this workflow lets researchers easily configure and share model setups, run experiments on national infrastructure, and assess results within a single coherent, reproducible environment, both lowering technical overhead for new users while facilitating robust science. As all components are open-source, we warmly invite the community to adopt it, test it, contribute code, and suggest and submit datasets to the CCD.