TSSM: Triaxial State Space Model for Global Station Weather Forecasting with Temporal-Variable-Historical Modeling

Short summary

The paper proposes TSSM, a Triaxial State Space Model for global station weather forecasting that incorporates period-aligned historical data to capture long-term weather patterns beyond standard lookback windows. TSSM achieves SOTA on the Weather-5K dataset with 10% accuracy gains and 61% improvement on extreme event metrics. It retains over 90% performance under up to 80% missing observations, compared to less than 43% for baselines.

• •Proposes TSSM with temporal-variable-historical scanning to capture long-term periodic weather patterns
• •Achieves SOTA on Weather-5K with 10% accuracy and 61% extreme event metric gains
• •Retains >90% performance under 80% missing observations vs <43% for baselines

Generated with AI, which can make mistakes.