antarctic-ice-shelf-utac

GenesisAeon Package 120 — real Antarctic ice shelf mass budget, a real Holocene meltwater-release feedback mechanism, and a real, sign-flipped regional asymmetry under CMIP6 projections. Deliberately has no UTAC/CREP/AFET bridge — see DISCLAIMER.md.

For a plain-language explanation of the same topic (German, no jargon, written for general audiences), see WHITEPAPER.md.

Where this package came from

Built alongside el-nino-amplification-utac (P118) and marine-heatwave-utac (P119) from a single broad DeepResearch pass (2026-08-31). cascading-tipping-utac (P87) only references WAIS as one node in a tipping-element network, not a standalone quantified treatment — this package fills that gap.

What's real here

• Davison et al. (2023, Science Advances) — of 162 Antarctic ice shelves studied 1997-2021, 71 lost mass (48 lost more than 30% of their initial mass), 29 gained, 62 showed no significant change. Total loss: 8.3 trillion tonnes. Basal melting was the dominant driver (68%). West Antarctica predominantly lost mass; East Antarctica predominantly stayed stable or gained.
• Nature Geoscience (2025) — sediment-record and ocean-model evidence for a real, self-reinforcing meltwater-release feedback that drove an ice-shelf collapse ~9000 years ago: meltwater strengthens stratification, which enhances warm deep-water inflow, which accelerates melting, which releases more meltwater.
• Muilwijk et al. (2026, The Cryosphere) — 10 CMIP6 models under a standardized meltwater perturbation show opposite-sign regional responses: East Antarctica's feedback amplifies warming (+0.16 to +0.26 degC median, up to +0.82 degC by 2100 under SSP5-8.5), while the Amundsen/Bellingshausen sectors of West Antarctica show cooling or suppressed warming (-0.5 to -1.5 degC).

Deliberately not one-sided

do_east_and_west_responses_share_the_same_sign() returns False — the CMIP6 regional finding directly complicates a naive "meltwater feedback always accelerates melting everywhere" reading of the Holocene mechanism documented alongside it. Both are real findings, kept explicit rather than smoothed into one narrative.

v2.0.0 (2026-09-03): six more real mechanisms

Following a Johann brainstorm prompted by a coral-reef article, six independent research passes each verified a real, citable mechanism:

• Subglacial hydrology — Wilson et al. (2025, Nat. Commun.): 85 newly detected active subglacial lakes (known inventory +58%), median drainage 2.2yr / recharge 3.5yr, 6 within 8km of the grounding zone. Dow et al. (2022, Nat. Geosci.): a 460km high-pressure subglacial river system. Physically distinct from the ice-shelf basal melting above — this is water beneath the grounded ice.
• Southern Annular Mode (SAM) — the real Antarctic analog to the Arctic polar vortex, but the "instability exports cold air" analogy does not transfer: the 2019 vortex weakening produced hot, dry conditions in Australia instead (Lim et al. 2019), a real contributor to the Black Summer bushfires. SAM does drive real, regionally asymmetric basal melt (+40 Gt/yr per +1 SD, opposite-signed in Amundsen vs. Bellingshausen/W. Pacific — Verfaillie et al. 2022), and SAM+ENSO are the two dominant decadal drivers of Antarctic ice-mass variability (King et al. 2023).
• Sea-ice albedo feedback — Riihelä et al. (2021, Nat. Geosci.): a real, quantified feedback that flipped sign, from −0.06±0.02 W/m²/decade (1992-2015 expansion era) to +0.26 W/m² (2016-2018 sea- ice-loss reversal). Explicitly not exposed-continental-rock albedo — that mechanism was searched for and not found as a quantified Antarctica-specific result; don't conflate the two.
• Seismic-ice interactions — real, tidally-modulated basal stick-slip icequakes on the Whillans Ice Plain (~20,000 events on record, Wiens et al. 2008); permafrost-thaw debris flows documented at Potter Peninsula since 1956, though without a quantified trend like the (non-Antarctic) Alpine literature; the 1998 Balleny Islands Mw8.1 earthquake, GIA-triggered — but GIA itself is a stabilizing feedback on West Antarctic ice, not an amplifying one.
• Black carbon deposition — a different geography from black-carbon-albedo-utac (P105, Tibetan Plateau) — no overlap. Cordero et al. (2022, Nat. Commun.): local station/tourism emissions dominate on average. Magalhães et al. (2024, Sci. Adv.): extreme wildfire events (2019-2020 Australian megafires) can temporarily override that local baseline — both kept, not smoothed.
• ENSO teleconnection — Paolo et al. (2018, Nat. Geosci.): during strong El Niño, Amundsen Sea shelves gain surface mass from snowfall but lose ~5x more via basal melt — "the ocean ends up winning." Ayabilah et al. (2026, The Cryosphere): on ENSO timescales, West Antarctica gains mass during El Niño — the opposite short-term sign of this package's own core Davison et al. 2023 long-term trend. Two different timescales, not a contradiction — kept as an explicit honesty-check nuance.

Installation

pip install antarctic-ice-shelf-utac

Quick start

from antarctic_ice_shelf_utac import (
    pct_shelves_losing_mass,
    is_loss_uniform_east_vs_west,
    do_east_and_west_responses_share_the_same_sign,
)

pct_shelves_losing_mass()  # 43.8
is_loss_uniform_east_vs_west()  # False
do_east_and_west_responses_share_the_same_sign()  # False -- the honesty check

License

Code: MIT. Documentation/data notes: see DISCLAIMER.md.

Citation

See CITATION.cff.