Hierarchical U-net with re-parameterization technique for spatio-temporal weather forecasting

doi:10.1007/s10994-023-06445-3

CORC > 自动化研究所 > 中国科学院自动化研究所 > 中科院工业视觉智能装备工程实验室

	Hierarchical U-net with re-parameterization technique for spatio-temporal weather forecasting
	Xu, Baowen 1,2; Wang, Xuelei 1; Li, Jingwei 1,2; Liu, Chengbao 1
刊名	MACHINE LEARNING
	2024-01-12
页码	19
关键词	Spatio-temporal weather forecasting U-Net Re-parameterization
ISSN号	0885-6125
DOI	10.1007/s10994-023-06445-3
通讯作者	Xu, Baowen(xubaowen2021@ia.ac.cn)
英文摘要	Due to the considerable computational demands of physics-based numerical weather prediction, especially when modeling fine-grained spatio-temporal atmospheric phenomena, deep learning methods offer an advantageous approach by leveraging specialized computing devices to accelerate training and significantly reduce computational costs. Consequently, the application of deep learning methods has presented a novel solution in the field of weather forecasting. In this context, we introduce a groundbreaking deep learning-based weather prediction architecture known as Hierarchical U-Net (HU-Net) with re-parameterization techniques. The HU-Net comprises two essential components: a feature extraction module and a U-Net module with re-parameterization techniques. The feature extraction module consists of two branches. First, the global pattern extraction employs adaptive Fourier neural operators and self-attention, well-known for capturing long-term dependencies in the data. Second, the local pattern extraction utilizes convolution operations as fundamental building blocks, highly proficient in modeling local correlations. Moreover, a feature fusion block dynamically combines dual-scale information. The U-Net module adopts RepBlock with re-parameterization techniques as the fundamental building block, enabling efficient and rapid inference. In extensive experiments carried out on the large-scale weather benchmark dataset WeatherBench at a resolution of 1.40625 degrees, the results demonstrate that our proposed HU-Net outperforms other baseline models in both prediction accuracy and inference time.
资助项目	Key Technologies Research and Development Program
WOS关键词	PREDICTION
WOS研究方向	Computer Science
语种	英语
出版者	SPRINGER
WOS记录号	WOS:001141343500001
资助机构	Key Technologies Research and Development Program
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/54801]
专题	中科院工业视觉智能装备工程实验室
通讯作者	Xu, Baowen
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China
推荐引用方式 GB/T 7714	Xu, Baowen,Wang, Xuelei,Li, Jingwei,et al. Hierarchical U-net with re-parameterization technique for spatio-temporal weather forecasting[J]. MACHINE LEARNING,2024:19.
APA	Xu, Baowen,Wang, Xuelei,Li, Jingwei,&Liu, Chengbao.(2024).Hierarchical U-net with re-parameterization technique for spatio-temporal weather forecasting.MACHINE LEARNING,19.
MLA	Xu, Baowen,et al."Hierarchical U-net with re-parameterization technique for spatio-temporal weather forecasting".MACHINE LEARNING (2024):19.