This work asks how far in advance a climate tipping event can, in principle,
be predicted once the internal variability of the Earth system is taken into
account, rather than treating tipping as driven purely by external forcing.
A data assimilation approach is used to estimate the predictability horizon
of tipping in a simplified Earth system setting, connecting the state
estimation problem to the risk assessment of abrupt climate change.
Published as Kubo, A., & Sawada, Y. (2025), Geophysical Research Letters,
52, e2024GL113146. Earlier versions of this work were presented at the
Paleoclimate Modelers’ Workshop 2024 (Tokyo), JpGU Meeting 2024 (Makuhari,
Oral), the 10th International Symposium on Data Assimilation (Kobe), and
AGU Fall Meeting 2024 (Washington, D.C., Poster).