Meteorology-Conditioned High-Resolution Vegetation Forecasting: A Hierarchical Multi-Modal Fusion Network
Remote Sensing · 22.05.2026 · 18 (11) · 1684
DOI: 10.3390/rs18111684 ↗Kurz gefasst
Fachartikel in der Zeitschrift Remote Sensing, online erschienen am 22.05.2026. Verfasst von Zhihang Yi und 6 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Joint Funds of the National Natural Science Foundation of China, Climate Change Special Project of China Meteorological Administration, Natural Science Foundation of Ningxia.
Abstract (Originalsprache)
Predicting high-resolution Normalized Difference Vegetation Index (NDVI) in mountainous ecosystems is challenging due to topographic complexity and climate heterogeneity. Existing methods often struggle to balance fine-grained spatial patterns with multi-scale meteorological drivers. This paper introduces the Hierarchical Multi-Modal Fusion Network (HMMFN), which employs a conditioned reconstruction strategy to decouple spatial learning from environmental forcing. The architecture utilizes a dual-stream encoder to process NDVI imagery and meteorological data in parallel. A Transformer module captures long-term temporal dependencies, while a multi-level fusion decoder integrates climate semantics with local vegetation details. The model is optimized using a hybrid loss function that combines Mean Squared Error and Structural Similarity Index Measure to ensure both numerical precision and spatial fidelity. Evaluated in the Liupan Mountains, HMMFN consistently outperforms baseline models across multiple lead times. For prediction horizons ranging from one to five months, the model maintains high accuracy with R2 values between 0.9123 (1-month horizon) and 0.8195 (5-month horizon), achieving a 10.1% and 3.6% reduction in RMSE compared to the optimal baseline model, respectively. These results demonstrate that HMMFN effectively preserves fine-scale spatial structures while maintaining accurate temporal trends across various time steps.
Förderung
- Joint Funds of the National Natural Science Foundation of China
- Climate Change Special Project of China Meteorological Administration
- Natural Science Foundation of Ningxia
- Open Research Project of the Key Laboratory of Monitoring, Early Warning and Risk Management for Specialty Agricultural Meteorological Disasters in Arid Region…
- Graduate Education Quality Improvement Project of North Minzu University
Zitieren
APA
Yi, Z., Yang, J., Wang, H., Kang, X., Zhang, S., Zhu, X., & Han, Y. (2026). Meteorology-Conditioned High-Resolution Vegetation Forecasting: A Hierarchical Multi-Modal Fusion Network. Remote Sensing, 18(11), 1684. https://doi.org/10.3390/rs18111684
Vancouver
Yi Z, Yang J, Wang H, Kang X, Zhang S, Zhu X, et al. Meteorology-Conditioned High-Resolution Vegetation Forecasting: A Hierarchical Multi-Modal Fusion Network. Remote Sensing. 2026;18(11):1684. doi:10.3390/rs18111684
BibTeX
@article{Yi2026,
author = {Yi, Zhihang and Yang, Jianling and Wang, Hairong and Kang, Xiong and Zhang, Suzhao and Zhu, Xiaowei and Han, Yingjuan},
title = {Meteorology-Conditioned High-Resolution Vegetation Forecasting: A Hierarchical Multi-Modal Fusion Network},
journal = {Remote Sensing},
year = {2026},
volume = {18},
number = {11},
pages = {1684},
doi = {10.3390/rs18111684}
}
EUVIMED