MFD-DF: A PM2.5 Concentration Prediction Method Based on Multimodal Feature Decomposition and Dynamic Fusion
Atmosphere · 18.06.2026 · 17 (6) · 616
DOI: 10.3390/atmos17060616 ↗Kurz gefasst
Fachartikel in der Zeitschrift Atmosphere, online erschienen am 18.06.2026. Verfasst von Chen Song und 8 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Accurate air pollutant concentration prediction is crucial for public health and sustainable urban development. Existing methods predominantly rely on single-modal data, resulting in inadequate representation of pollutant spatiotemporal evolution, poor prediction accuracy, and limited generalization capabilities. To address these challenges, this research proposes a novel PM2.5 prediction framework termed MFD-DF that integrates ground-station time series and satellite remote sensing images. In feature extraction, learnable decomposition and deformable convolution are introduced, and a Cross-Modal Slot Attention module explicitly decomposes features to resolve information blurring. Subsequently, a dynamic cross-modal alignment mechanism is designed alongside a learnable Time-Expansion Network (TEN) to ensure fine-grained interaction. Furthermore, a local-global attention feature fusion mechanism is proposed to optimize data integration efficacy. Experimental results demonstrate that in single-step PM2.5 prediction tasks, the proposed MFD-DF achieves significant improvements of approximately 10–20% in MAE, RMSE, and MAPE compared to state-of-the-art baselines. In multi-step PM2.5 prediction, it effectively alleviates the error accumulation problem in long-sequence forecasting, demonstrating superior robustness and accuracy.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Song, C., Long, Q., Su, Z., Jiang, Y., Wan, L., Zhang, X., Lv, T., Hao, W., & Shi, Z. (2026). MFD-DF: A PM2.5 Concentration Prediction Method Based on Multimodal Feature Decomposition and Dynamic Fusion. Atmosphere, 17(6), 616. https://doi.org/10.3390/atmos17060616
Vancouver
Song C, Long Q, Su Z, Jiang Y, Wan L, Zhang X, et al. MFD-DF: A PM2.5 Concentration Prediction Method Based on Multimodal Feature Decomposition and Dynamic Fusion. Atmosphere. 2026;17(6):616. doi:10.3390/atmos17060616
BibTeX
@article{Song2026,
author = {Song, Chen and Long, Quanbo and Su, Zhaobo and Jiang, Yanchao and Wan, Li and Zhang, Xiankun and Lv, Tiantian and Hao, Wenhu and Shi, Zuxuan},
title = {MFD-DF: A PM2.5 Concentration Prediction Method Based on Multimodal Feature Decomposition and Dynamic Fusion},
journal = {Atmosphere},
year = {2026},
volume = {17},
number = {6},
pages = {616},
doi = {10.3390/atmos17060616}
}
EUVIMED