carbon-sinks-utac

GenesisAeon Package 59 — combines two real, independent 2025/2026 corrections to global natural carbon sink capacity, plus a SPECULATIVE UTAC/CREP sink-stress bridge (this package's own, not established science — see DISCLAIMER.md).

What's real here

• Global Carbon Budget 2025 (ESSD 18, 3211, 2026): 2024 totals — 11.6 GtC/yr anthropogenic emissions, 3.4 GtC/yr ocean sink, 1.9 GtC/yr land sink, 7.9 GtC/yr atmospheric growth (with the GCB's own disclosed ~1.6 GtC/yr budget imbalance).
• Ocean sink correction (2025, real, upward): "Asymmetric bubble-mediated gas transfer enhances global ocean CO2 uptake", Nature Communications 16, 10595 (2025) — Plymouth Marine Laboratory & GEOMAR. Finds ocean CO2 uptake ~15% higher (average 1991-2020) than conventional estimates once bubble-transfer asymmetry is accounted for.
• Land sink correction (2026, real, downward): O'Sullivan, M. et al., "An improved approach to estimate the natural land carbon sink", npj Climate and Atmospheric Science 9, 29 (2026). Finds the 2015-2024 decadal land sink is ~20% smaller than previously estimated (3.00 → 2.42 PgC/yr), because older models assumed fixed pre-industrial land cover.

Why both corrections, not just one

A package reporting only the negative land-sink correction would itself be an unbalanced, overly pessimistic picture — exactly the kind of "compounding crisis" overstatement this ecosystem's citation-accuracy discipline exists to catch. sink_revision.py computes the net effect of both real corrections together. The honest finding: crediting the real ocean-sink upgrade does not flip the overall picture positive — the land-sink correction is larger, so the net revision to estimated natural sink capacity is still a reduction, just a smaller one than looking at the land correction alone would suggest.

What's SPECULATIVE

utac_bridge.py maps a "sink-efficiency deficit" (how far current sink efficiency sits below a reference value) onto a UTAC steepness beta, using the same K / driver convention used elsewhere in the ecosystem. Unlike amoc-utac or amazon-utac, there is no established bistable tipping-point threshold for "carbon sink collapse" in the cited literature — this bridge is this package's own untested framework choice, not a re-derivation of a validated result.

Quickstart

pip install carbon-sinks-utac

from carbon_sinks_utac import CarbonSinksUTAC, compute_net_sink_revision

print(compute_net_sink_revision().summary())

sys = CarbonSinksUTAC()
result = sys.run_cycle()
print(result["net_revision_summary"])

Development

pip install -e ".[dev]"
pre-commit install
ruff check src tests
mypy src
pytest

Citation

See CITATION.cff and .zenodo.json.