paper
2023
Sub-seasonal to seasonal
Climate analysis & attribution New model or method
Transformer / ViTSeasonal to annualNanjing University of Information Science & Technology (NUIST) Journal
★ Highlight 2023-03-10
064
Mar 2023
Lu Zhou, Rong-Hua Zhang
Science Advances
Nanjing University of Information Science & Technology (NUIST)
3D-Geoformer, a self-attention transformer, predicts Niño3.4 SST anomalies 18 months ahead with high correlation skill for ENSO.
1
Sub-seasonal to seasonal★ Highlight
A self-attention–based neural network for three-dimensional multivariate modeling and its skillful ENSO predictions
Why it matters. Transformer-based 3D model gives skillful ENSO forecasts to 18 months; a solid, widely cited step for ML seasonal prediction.
In one sentence. 3D-Geoformer, a self-attention transformer, predicts Niño3.4 SST anomalies 18 months ahead with high correlation skill for ENSO.
Labels
- Topic
- Sub-seasonal to seasonal, Climate analysis & attribution
- Type
- New model or method
- Method
- Transformer / ViT
- Time scale
- Seasonal to annual
- Domain
- Global
- Data
- ERA5 / reanalysis
- Status
- Research
- Identifiers
- DOI:10.1126/sciadv.adf2827 OpenAlex:W4323637041
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.