EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

LSTMAtU-Net: A Precipitation Nowcasting Model Based on ECSA Module

Huantong Geng, Xiaoyan Ge iD, Boyang Xie, Jinzhong Min, Xiaoran Zhuang

Sensors · 21.06.2023 · 23 (13) · 5785

DOI: 10.3390/s23135785 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 21.06.2023. Verfasst von Huantong Geng und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer, Nanjing University of Information Science and Technology, Nanjing 210044, China. Hauptthemen laut MEDLINE: Meteorology, Radar. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 43 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Key Research Development Plan, Beijing foundation of NJIAS.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Precipitation nowcasting refers to the use of specific meteorological elements to predict precipitation in the next 0–2 h. Existing methods use radar echo maps and the Z–R relationship to directly predict future rainfall rates through deep learning methods, which are not physically constrained, but suffer from severe loss of predicted image details. This paper proposes a new model framework to effectively solve this problem, namely LSTMAtU-Net. It is based on the U-Net architecture, equipped with a Convolutional LSTM (ConvLSTM) unit with the vertical flow direction and depthwise-separable convolution, and we propose a new component, the Efficient Channel and Space Attention (ECSA) module. The ConvLSTM unit with the vertical flow direction memorizes temporal changes by extracting features from different levels of the convolutional layers, while the ECSA module innovatively integrates different structural information of each layer of U-Net into the channelwise attention mechanism to learn channel and spatial information, thereby enhancing attention to the details of precipitation images. The experimental results showed that the performance of the model on the test dataset was better than other examined models and improved the accuracy of medium- and high-intensity precipitation nowcasting.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

MeteorologyRadarImage Processing, Computer-Assisted

Stichwörter der Autor:innen

Convolutional LSTM (ConvLSTM)Efficient Channel and Space Attention (ECSA)U-Net architectureprecipitation nowcasting

Autor:innen und Institutionen

  • School of Computer, Nanjing University of Information Science and Technology, Nanjing 210044, China.
  • Jiangsu Open University, Nanjing 210036, China.
  • School of Meteorology, Nanjing University of Information Science and Technology, Nanjing 210044, China.
  • Jiangsu Meteorological Observatory, Nanjing 210008, China.

ORCID-Profile: Xiaoyan Ge

Förderung

  • National Key Research Development Plan
  • Beijing foundation of NJIAS

Zitieren

APA

Geng, H., Ge, X., Xie, B., Min, J., & Zhuang, X. (2023). LSTMAtU-Net: A Precipitation Nowcasting Model Based on ECSA Module. Sensors, 23(13), 5785. https://doi.org/10.3390/s23135785

Vancouver

Geng H, Ge X, Xie B, Min J, Zhuang X. LSTMAtU-Net: A Precipitation Nowcasting Model Based on ECSA Module. Sensors. 2023;23(13):5785. doi:10.3390/s23135785

BibTeX
@article{Geng2023,
  author = {Geng, Huantong and Ge, Xiaoyan and Xie, Boyang and Min, Jinzhong and Zhuang, Xiaoran},
  title = {LSTMAtU-Net: A Precipitation Nowcasting Model Based on ECSA Module},
  journal = {Sensors},
  year = {2023},
  volume = {23},
  number = {13},
  pages = {5785},
  doi = {10.3390/s23135785}
}
Kostenlos starten →