Research map / Time series / Long-term forecasting
SatSolarCast: A Flexible Framework for Multimodal Solar Irradiance Forecasting via Memory-Alignment Learning
Kuai Dai, Hui Su, Chengxing Zhai, Huiwei Lin, Mingliang Bai
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 the recall phase of memory alignment training (ablation) | LSTM+Sat · 3-hour forecasting ablation | R2: 0.656 -> 0.726 | Improved |
| add the recall phase of memory alignment training (ablation) | LSTM+Sat · 3-hour forecasting ablation | RMSE: 178.21 -> 152.18 | Improved |
| add the recall phase of memory alignment training (ablation) | PatchTST+Sat · 3-hour forecasting ablation | R2: 0.690 -> 0.707 | Improved |
| add the global spatial feature extractor (ablation) | LSTM+Sat · 3-hour forecasting ablation | R2: 0.702 -> 0.726 | Improved |
| add the global spatial feature extractor (ablation) | PatchTST+Sat · 3-hour forecasting ablation | R2: 0.645 -> 0.707 | Improved |
| use spatiotemporal attention fusion instead of simple pooled addition (ablation) | LSTM+Sat · 3-hour forecasting ablation | RMSE: 153.97 -> 152.18 | Improved |
| use spatiotemporal attention fusion instead of simple pooled addition (ablation) | PatchTST+Sat · 3-hour forecasting ablation | R2: 0.703 -> 0.707 | Improved |
| add SatSolarCast satellite memory module | PatchTST · SolarCube 3-hour forecasting (4 stations) | R2 and RMSE: 24.04% R2 increase and 17.57% RMSE reduction | Improved |
| add SatSolarCast satellite memory module | PatchTST · SolarCube 6-hour forecasting (4 stations) | R2 and RMSE: 21.92% R2 and 12.07% RMSE | Improved |
| add SatSolarCast satellite memory module | Transformer · SolarCube 3-hour forecasting (psu) | R2: 0.603 -> 0.716 | Improved |
| add SatSolarCast satellite memory module | LSTM · SolarCube 3-hour forecasting (psu) | RMSE: 180.73 -> 147.17 | Improved |
| use LSTM-att+Sat instead of the ViT-based CrossVIVIT multimodal model | CrossVIVIT · SolarCube 3-hour forecasting (avg of 4 stations) | R2: 0.693 -> 0.727 | Improved |
Closest papers by meaning
- SolarCube: An Integrative Benchmark Dataset Harnessing Satellite and In-situ Observations for Large-scale Solar Energy Forecasting (NeurIPS 2024 · similarity 0.85)
- Learning with Calibration: Exploring Test-Time Computing of Spatio-Temporal Forecasting (NeurIPS 2025 · similarity 0.84)
- DAWP: A framework for global observation forecasting via Data Assimilation and Weather Prediction in satellite observation space (NeurIPS 2025 · similarity 0.83)
- On the Integration of Spatial-Temporal Knowledge: A Lightweight Approach to Atmospheric Time Series Forecasting (NeurIPS 2025 · similarity 0.82)
- Solar Power Generation Forecasting via Multimodal Feature Fusion (Student Abstract) (AAAI 2024 · similarity 0.82)
- SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery (AAAI 2025 · similarity 0.81)
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.
