Research map / Time series / Causal and generative time series
Sub-Seasonal Climate Forecasting via Machine Learning: Challenges, Analysis, and Advances
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| Use target-date covariates rather than temporally local training data | XGBoost - one day · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.1965(0.04) -> 0.3044(0.03) | Helped |
| Add soil moisture covariates (ablation) | XGBoost - one day · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2685(0.03) -> 0.3044(0.03) | Helped |
| Add NAO and all Niño indices (ablation) | XGBoost - one day · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2081(0.03) -> 0.3044(0.03) | Helped |
| Add soil moisture covariates (ablation) | Lasso - one day · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2638(0.04) -> 0.2499(0.04) | Hurt |
| Add NAO and all Niño indices (ablation) | Lasso - one day · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.1956(0.04) -> 0.2499(0.04) | Helped |
| Add soil moisture covariates (ablation) | Encoder (LSTM)-Decoder (FNN) · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2157 (0.04) -> 0.2616 (0.04}) | Helped |
| Add NAO and all Niño indices (ablation) | Encoder (LSTM)-Decoder (FNN) · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2236 (0.04) -> 0.2616 (0.04}) | Helped |
| Use target-date covariates only instead of a longer feature sequence (ablation) | XGBoost · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2080(0.03) -> 0.3044(0.03) | Helped |
| Use target-date covariates only instead of a longer feature sequence (ablation) | Lasso · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2160(0.04) -> 0.2499(0.04) | Helped |
| Replace the LSTM decoder with an FNN decoder (ablation) | Encoder (LSTM)-Decoder · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.0740(0.03) -> 0.2616 (0.04}) | Helped |
| Give the FNN decoder every encoder output step instead of only the last step (ablation) | Encoder (LSTM)-Decoder (FNN) · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.1614 (0.05) -> 0.2616 (0.04}) | Helped |
| Add an autoregressive temperature component (ablation) | Encoder (LSTM)-Decoder (FNN) · US mainland tmp2m, 2017-2018 | spatial cosine similarity, mean: 0.2616 (0.04}) -> 0.1733 (0.04) | Hurt |
Closest papers by meaning
- DroughtSet: Understanding Drought Through Spatial-Temporal Learning (AAAI 2025 · similarity 0.88)
- ChaosBench: A Multi-Channel, Physics-Based Benchmark for Subseasonal-to-Seasonal Climate Prediction (NeurIPS 2024 · similarity 0.86)
- OmniCast: A Masked Latent Diffusion Model for Weather Forecasting Across Time Scales (NeurIPS 2025 · similarity 0.86)
- HR-Extreme: A High-Resolution Dataset for Extreme Weather Forecasting (ICLR 2025 · similarity 0.85)
- SEMPO: Lightweight Foundation Models for Time Series Forecasting (NeurIPS 2025 · similarity 0.85)
- CirT: Global Subseasonal-to-Seasonal Forecasting with Geometry-inspired Transformer (ICLR 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.