Inferring the transition type from early‑warning signals in real‑world observables

Ascertain what type of transition—full AMOC collapse to a shallow, salinity‑driven state, partial collapse localized to a specific deep‑water formation region, or a change in spatio‑temporal flow mode—is indicated by early‑warning signals detected in real‑world AMOC‑related observables.

Background

The paper’s conceptual model shows that different observables can exhibit diverse and even non‑monotonic early‑warning behaviors, and that EWS generally signal local loss of stability without uniquely determining the post‑bifurcation state.

Given overlapping impacts of multiple potential tipping points on common observables (e.g., AMOC strength), interpreting EWS to determine whether a forthcoming transition is partial or full remains an unresolved task.