Deutsch
 
Datenschutzhinweis Impressum
  DetailsucheBrowse

Datensatz

DATENSATZ AKTIONENEXPORT

Freigegeben

Buchkapitel

Exploring Geometric Deep Learning for Precipitation Nowcasting

Urheber*innen

Zhao,  Shan
External Organizations;

Saha,  Sudipan
External Organizations;

Xiong,  Zhitong
External Organizations;

/persons/resource/Niklas.Boers

Boers,  Niklas
Potsdam Institute for Climate Impact Research;

Zhu,  Xia Xiang
External Organizations;

Externe Ressourcen
Es sind keine externen Ressourcen hinterlegt
Volltexte (frei zugänglich)
Es sind keine frei zugänglichen Volltexte in PIKpublic verfügbar
Ergänzendes Material (frei zugänglich)
Es sind keine frei zugänglichen Ergänzenden Materialien verfügbar
Zitation

Zhao, S., Saha, S., Xiong, Z., Boers, N., Zhu, X. X. (2023): Exploring Geometric Deep Learning for Precipitation Nowcasting. - In: IGARSS 2023 - 2023 IEEE International Geoscience and Remote Sensing Symposium: Proceedings, New York : Institute of Electrical and Electronics Engineers.
https://doi.org/10.1109/IGARSS52108.2023.10282387


Zitierlink: https://publications.pik-potsdam.de/pubman/item/item_29839
Zusammenfassung
Precipitation nowcasting (up to a few hours) remains a challenge due to the highly complex local interactions that need to be captured accurately. Convolutional Neural Networks rely on convolutional kernels convolving with grid data and the extracted features are trapped by limited receptive field, typically expressed in excessively smooth output compared to ground truth. Thus they lack the capacity to model complex spatial relationships among the grids. Geometric deep learning aims to generalize neural network models to non-Euclidean domains. Such models are more flexible in defining nodes and edges and can effectively capture dynamic spatial relationship among geographical grids. Motivated by this, we explore a geometric deep learning-based temporal Graph Convolutional Network (GCN) for precipitation nowcasting. The adjacency matrix that simulates the interactions among grid cells is learned automatically by minimizing the L1 loss between prediction and ground truth pixel value during the training procedure. Then, the spatial relationship is refined by GCN layers while the temporal information is extracted by 1D convolution with various kernel lengths. The neighboring information is fed as auxiliary input layers to improve the final result. We test the model on sequences of radar reflectivity maps over the Trento/Italy area. The results show that GCNs improves the effectiveness of modeling the local details of the cloud profile as well as the prediction accuracy by achieving decreased error measures.