Research map / Time series / Long-term forecasting
ARROW: An Adaptive Rollout and Routing Method for Global Weather Forecasting
Jindong Tian, Yifei Ding, Ronghui Xu, Hao Miao, Chenjuan Guo, Bin Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add Ring Positional Encoding (vs learnable 2D PE) (ablation) | ARROW (MIFM pretrain) · T2m/U10/V10 72h | RMSE: T2m 1.12 -> 1.09; U10 1.76 -> 1.71 | Improved |
| add RPE (vs 2D PE) (ablation) | ARROW (MIFM pretrain) · T2m 72h | ACC: 0.91 -> 0.91 | No clear effect |
| add S&P MoE (vs plain FFN) (ablation) | ARROW (MIFM pretrain) · T2m/U10/V10 72h | RMSE: T2m 1.13 -> 1.09; U10 1.77 -> 1.71 | Improved |
| add aux-loss1 (interval specialization) (ablation) | ARROW (MIFM pretrain) · T2m/U10/V10 72h | RMSE: T2m 1.15 -> 1.09; V10 1.92 -> 1.77 | Improved |
| add aux-loss2 (load balancing) (ablation) | ARROW (MIFM pretrain) · T2m/U10/V10 72h | RMSE: T2m 1.12 -> 1.09; V10 1.85 -> 1.77 | Improved |
| use adaptive rollout instead of random or naive (ablation) | ARROW (MIFM pretrain) · T2m/T850 138h | error | Improved |
| use greedy rollout instead of naive 6h rollout (ablation) | ARROW (MIFM pretrain) · T2m/T850 138h | error | Hurt |
| ARROW vs best data-driven baseline | ARROW (MIFM pretrain) · WeatherBench1 6 variables, 5-day | RMSE and ACC: 9.3% RMSE, 10% ACC average improvement | Improved |
| ARROW vs IFS | ARROW (MIFM pretrain) · WeatherBench1 T850 5-day | RMSE: IFS 2.04 vs ARROW 2.06 | Mixed |
Closest papers by meaning
- Integrating Multi-Source Data for Long Sequence Precipitation Forecasting (AAAI 2025 · similarity 0.86)
- Learning with Calibration: Exploring Test-Time Computing of Spatio-Temporal Forecasting (NeurIPS 2025 · similarity 0.86)
- Tyche: One Step Flow for Efficient Probabilistic Weather Forecasting (NeurIPS 2026 · similarity 0.86)
- OmniCast: A Masked Latent Diffusion Model for Weather Forecasting Across Time Scales (NeurIPS 2025 · similarity 0.85)
- Continuous Ensemble Weather Forecasting with Diffusion models (ICLR 2025 · similarity 0.85)
- On the Integration of Spatial-Temporal Knowledge: A Lightweight Approach to Atmospheric Time Series Forecasting (NeurIPS 2025 · similarity 0.85)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
