Documented attemptclimate science · ocean-atmosphere variability and seasonal forecasting
EarthLink attempts an eight-month Atlantic Niño forecast
EarthLink proposed an Atlantic wind-thermocline precursor and improved on persistence and ENSO-only comparators, but its hindcasts missed the requested temporal-correlation target above 0.5.
Editorial context
What happened
The prompt asked the agent to exceed a temporal-correlation skill of 0.5 at an eight-month lead, identify useful precursors, build multiple forecasting models, and explain the physical mechanism. The attempt is useful precisely because its outcome is not binary. It missed the stated threshold while still generating a physically consistent candidate pathway and outperforming simpler comparators in retrospective analysis.
Keep in view
The numerical target was not met, the mechanism remains an author-proposed hypothesis, and the exact model used for this production run is not sufficiently clear from the public materials.
Chronology
Timeline
Exact research request, analyses, and target-missed result publicly disclosedpreprint · day date
Trust boundaries
Validation
Historical hindcasts reached temporal correlations of roughly 0.36 to 0.46, above comparison methods but below the requested threshold.
Authors and experts assessed the proposed mechanism; no prospective forecast or independent replication was identified.
System
AI and tools
- EarthLink
- Exact production backbone undisclosed
People
Human contributors
- Zijie Guo
- Jiong Wang
- Fenghua Ling
- Lei Bai
- EarthLink collaborators
Organizations
Affiliations
- OpenEarthLab
Primary materials