Antarctic ice melt triggers further melting: Evidence for cascading feedbacks 9,000 years ago / News & Topics / National Institute of Polar Research

Antarctic ice melt triggers further melting: Evidence for cascading feedbacks 9,000 years ago

Integration of proxy records with ocean–climate modeling reveals that early Holocene ice-shelf retreat in East Antarctica was driven by oceanic forcing enhanced by meltwater discharge from neighboring regions.

November 7, 2025

An international team led by Japan’s National Institute of Polar Research discovered that large-scale East Antarctic ice loss around 9,000 years ago was amplified by a “cascading positive feedback” between meltwater and ocean circulation. Sediment records and ocean–climate modeling show that meltwater from other Antarctic regions strengthened warm deep-water inflow, accelerating ice-shelf collapse and inland thinning. The finding highlights how Antarctic ice melt can self-reinforce under ongoing global warming.

Marine sediment coring from the icebreaker Shirase during the 61st Japanese Antarctic Research Expedition (2019–2020).
A large cylindrical coring device is deployed vertically into the seafloor and recovered to obtain sediment cores. Credit: National Institute of Advanced Industrial Science and Technology (AIST)

A recent study published in Nature Geoscience has revealed that the substantial retreat of the East Antarctic Ice Sheet (EAIS) approximately 9,000 years ago was driven by a self-reinforcing feedback loop between ice melt and ocean circulation. The research team, led by Professor Yusuke Suganuma from the National Institute of Polar Research (NIPR) and the Graduate University for Advanced Studies (SOKENDAI), found that the inflow of warm deep water into coastal East Antarctica caused the collapse of ice shelves, which in turn accelerated inland ice loss. The discovery indicates that Antarctic ice retreat is not merely a regional phenomenon; rather, it has the potential to propagate across multiple sectors through oceanic connections, thereby amplifying the overall magnitude of ice loss. This phenomenon, in which meltwater from one region accelerates melting processes in other regions, is referred to as a "cascading positive feedback". This feedback loop may be a crucial factor in comprehending the instability of Antarctic ice sheets, both in the past and in the present.

Uncovering the mechanism of past ice-sheet collapse

The present study aims to elucidate the mechanism of past ice-sheet collapse.
The East Antarctic Ice Sheet, which contains more than half of the world's freshwater, is currently undergoing mass loss in certain coastal regions. It is imperative to comprehend the manner in which substantial ice sheets have responded to prior periods of climate warming, as this furnishes indispensable insights into their prospective stability in the context of the present global warming phenomenon. In order to investigate this phenomenon, the research team analysed marine sediment cores collected from Lützow-Holm Bay, which is located along the Sôya Coast in the vicinity of Japan's Syowa Station. These cores were analysed in conjunction with geomorphological and geological surveys conducted across Dronning Maud Land. These sediments were obtained through multiple Japanese Antarctic Research Expeditions (JARE) between 1980 and 2023, including recent sampling from the icebreaker Shirase. Utilising a range of analytical techniques, including sedimentological, micropaleontological, and geochemical analyses, the researchers were able to reconstruct past environmental changes in the bay and its environs. This reconstruction was facilitated by the use of measurements of beryllium isotope ratios (10Be/9Be). The results of the study indicated that approximately 9,000 years ago, the presence of warm Circumpolar Deep Water (CDW) intensified within the bay, resulting in the collapse of floating ice shelves. As the shelves disintegrated, they lost their buttressing effect, causing inland ice to flow more rapidly towards the ocean.

Schematic illustration of the mechanism of East Antarctic Ice Sheet melting in Dronning Maud Land.
Around 9,000 years ago, at the time when the regional sea level reached its peak, warm deep water flowed into the area, leading to the collapse of the ice shelf. As a result, ice discharge accelerated, and the Antarctic Ice Sheet retreated substantially from the coast toward the inland. Credit : Yusuke Suganuma (NIPR)

Modeling reveals a self-reinforcing feedback

In order to comprehend the factors that precipitated the augmented influx of warm deep water, the research team conducted climate and high-resolution ocean model simulations. The models demonstrated that meltwater discharged from other Antarctic regions, including the Ross Ice Shelf, disseminated throughout the Southern Ocean. This process resulted in the freshening of the surface layer and the intensification of vertical stratification. This phenomenon has been shown to suppress the upward mixing of cold surface water, thereby facilitating the intrusion of warm deep water towards the continental shelf of East Antarctica. This process resulted in the establishment of a positive feedback loop, whereby meltwater led to stronger stratification, which in turn led to enhanced deep-water inflow, thereby further facilitating ice melt. The existence of such a cascading mechanism suggests that melting in one Antarctic sector may trigger or accelerate melting in others through oceanic teleconnections.

A warning from the past

The present study provides one of the clearest pieces of evidence to date that the Antarctic Ice Sheet may be susceptible to self-reinforcing, widespread melting in response to natural climate warming. Despite the occurrence of the event during the early Holocene epoch, a period characterised by elevated global temperatures relative to the last glacial period, the physical mechanisms elucidated by this study bear a direct relevance to the contemporary global warming phenomenon. Current observations indicate that parts of the West Antarctic Ice Sheet, such as the Thwaites and Pine Island glaciers, are already undergoing rapid retreat driven by warm deep-water intrusion. In the event of analogous cascading feedbacks being in operation in the present day, there is the possibility of regional melting spreading and accelerating the overall ice-sheet loss, thus contributing to accelerated global sea-level rise.

International collaboration and significance

The research involved collaboration among more than 30 institutions, including the National Institute of Polar Research (NIPR), the Geological Survey of Japan (AIST), the Japan Agency for Marine–Earth Science and Technology (JAMSTEC), the University of Tokyo, Kochi University, Hokkaido University, and international partners from New Zealand, Spain, and others. This interdisciplinary endeavour entailed the integration of geological field surveys, marine sediment analyses, cosmogenic nuclide exposure dating, and coupled climate–ocean modelling to achieve a comprehensive reconstruction of the ice-sheet–ocean system in East Antarctica.

Professor Suganuma concludes: This study provides essential data and modelling evidence that will facilitate more accurate predictions of future Antarctic ice-sheet behaviour. The cascading feedbacks identified in this study serve to underscore the notion that minor regional alterations can potentially engender global ramifications.

Original Article

Journal: Nature Geoscience
Title: Antarctic ice shelf collapse in Holocene driven by meltwater release feedbacks
Authors:
Yusuke Suganuma^1,2, Takuya Itaki³, Yuki Haneda³, Kazuya Kusahara⁴, Takashi Obase^5,4, Takeshige Ishiwa^1,2, Takayuki Omori⁶, Minoru Ikehara⁷, Rob McKay⁸, Osamu Seki⁹, Daisuke Hirano^1,2, Masakazu Fujii^1,2, Yuji Kato⁷, Atsuko Amano³, Yuki Tokuda¹⁰, Hokuto Iwatani¹¹, Yoshiaki Suzuki³, Motohiro Hirabayashi¹, Hiroyuki Matsuzaki¹², Takeyasu Yamagata¹², Masao Iwai⁷, Kota Katsuki¹³, Francisco J. Jimenez-Espejo^4,14, Hiroki Matsui¹⁵, Koji Seike³, Moto Kawamata^2,16, Naohisa Nishida¹⁷, Masato Ito¹, Shin Sugiyama⁸, Jun'ichi Okuno^1,2,18, Takanobu Sawagaki¹⁹, Ayako Abe-Ouchi^1,5, Shigeru Aoki⁸, Hideki Miura^1,20

1 National Institute of Polar Research,
2 The Graduate University for Advanced Studies (SOKENDAI)
3 Geological Survey of Japan, The National Institute of Advanced Industrial Science and Technology
4 Japan Agency for Marine-Earth Science and Technology
5 Atmosphere Ocean Research Institute, University of Tokyo
6 The University Museum, University of Tokyo
7 Marine Core Research Institute, Kochi University
8 Antarctic Research Centre, Victoria University of Wellington
9 Institute of Low Temperature Science, Hokkaido University
10 Faculty of Environmental Studies, Tottori University of Environmental Studies
11 Faculty of Science, Yamaguchi University, Yamaguchi
12 Micro Analysis Laboratory, Tandem Accelerator (MALT), The University of Tokyo
13 Estuary Research Center, Shimane University
14 Instituto Andaluz de Ciencias de la Tierra, Consejo Superior de Investigaciones Científicas
15 Graduate School of International Resource Sciences, Akita University
16 Civil Engineering Research Institute for Cold Region, Public Works Research Institute
17 Department of Environmental Sciences, Tokyo Gakugei University
18 Joint Support-Center for Data Science Research, Research Organization of Information and Systems
19 Faculty of Social Sciences, Hosei University
20 Faculty of Economics, Aomori Public University

DOI: 10.1038/s41561-025-01829-7
URL: https://www.nature.com/articles/s41561-025-01829-7

Funding

The study was supported by the JSPS Kakenhi (JP16H05739; JP17H06321; JP17H06318; JP19H00728; JP21KK0246; JP24H00026; JP24H02344), NIPR through Advanced Project (KP-1), and GRAntarctic. A part of this study was performed under the cooperative research program of Marine Core Research Institute (MaCRI), Kochi University (No. 20A040, 20B037, 22A007, 22B006).