Preprint · 2026 · Open access

A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations

Chaoran Zhou, Zhuoyang Cui, Hang Chen, Xin Zhang

Research Square · Published 24 September 2026 · DOI 10.21203/rs.3.rs-10892753/v1

Open PDFView at the publisherCite this paper

In brief

“A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations” is a 2026 preprint by Chaoran Zhou and 3 co-authors, published in Research Square. OpenAlex counts 0 citations as of 20 November 2026. It is open access under a CC BY licence: the full text is free to read at Research Square.

Citations
0
References
36
Authors
4

Abstract

Typhoons often trigger compound disasters such as torrential rain, urban flooding, and mountain torrents. Accurate prediction of their tracks is therefore critical for safeguarding life and property. However, typhoon tracks exhibit strong nonlinearity, temporal variability, and uncertainty, making multi-step typhoon track prediction vulnerable to error accumulation and track deviation. To address these challenges, this paper proposes TRGF-KG (Trajectory--Relation Graph Fusion with Knowledge Graph), a multi-step typhoon track prediction method that integrates spatiotemporal trajectory dynamics with prior structural relations. The typhoon knowledge graph is formulated as a relational structural prior that organizes continuous trajectory records into structured relational representations, thereby explicitly preserving event affiliation, temporal ordering, and relational dependencies that are difficult to represent from continuous trajectory sequences alone. A dual-branch architecture is employed: a Transformer learns block-level dynamic temporal representations, while an R-GCN learns relational structural prior representations from the typhoon knowledge graph. During collaborative prediction, the two representations are aligned through time-index-based cross-modal collaborative alignment and selectively integrated through feature-wise gated residual fusion. This design preserves the temporal branch as the prediction backbone while adaptively incorporating aligned relational structural priors to mitigate long-horizon error accumulation and spatial track deviation. Experiments on the Chinese Coastal Typhoon Track Dataset show that TRGF-KG outperforms comparison models in MAE, RMSE, $R^2$, and ADE, demonstrating the effectiveness of integrating spatiotemporal trajectory dynamics and prior structural relations in improving the accuracy and long-horizon stability of multi-step typhoon track prediction.

Cite this paper

Formatted from the catalogue record. Check names, capitals and page numbers against the paper before you submit.

APA 7

Zhou, C., Cui, Z., Chen, H., & Zhang, X. (2026). A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations. Research Square. https://doi.org/10.21203/rs.3.rs-10892753/v1

MLA 9

Zhou, Chaoran, et al. “A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations.” Research Square, 2026. https://doi.org/10.21203/rs.3.rs-10892753/v1.

Chicago

Zhou, Chaoran, Zhuoyang Cui, Hang Chen, and Xin Zhang. 2026. “A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations.” Research Square. https://doi.org/10.21203/rs.3.rs-10892753/v1.

Harvard

Zhou, C. et al. (2026) ‘A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations’, Research Square. doi:10.21203/rs.3.rs-10892753/v1.

IEEE

C. Zhou, Z. Cui, H. Chen, and X. Zhang, “A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations,” Research Square, 2026, doi: 10.21203/rs.3.rs-10892753/v1.

BibTeX

@misc{zhou2026multi,
  title     = {{A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations}},
  author    = {Zhou, Chaoran and Cui, Zhuoyang and Chen, Hang and Zhang, Xin},
  howpublished = {Research Square},
  year      = {2026},
  doi       = {10.21203/rs.3.rs-10892753/v1},
  url       = {https://doi.org/10.21203/rs.3.rs-10892753/v1}
}

Download for a reference manager: BibTeX (.bib) · RIS for Zotero, Mendeley and EndNote (.ris)

Topic and keywords

Physical Sciences › Earth and Planetary Sciences › Atmospheric Science › Tropical and Extratropical Cyclones Research

  • typhoon track prediction
  • multi-step forecasting
  • knowledge graphs
  • transformer models
  • relational graph convolutional network
  • cross-modal representation alignment

Work with this paper

Most-cited papers in Earth and Planetary Sciences

All highly cited Earth and Planetary Sciences papers

Questions about this paper

How do I cite this paper in APA style?

Zhou, C., Cui, Z., Chen, H., & Zhang, X. (2026). A Multi-Step Typhoon Track Prediction Method Integrating Spatiotemporal Trajectory Dynamics and Prior Structural Relations. Research Square. https://doi.org/10.21203/rs.3.rs-10892753/v1

How many times has this paper been cited?

0 times, according to OpenAlex on 20 November 2026. Google Scholar and Scopus count different sets of sources, so their numbers differ.

Is this paper free to read?

Yes. It is open access at Research Square: sign in and open the PDF from this page.

Who wrote this paper?

Chaoran Zhou, Zhuoyang Cui, Hang Chen, Xin Zhang.

What is the DOI of this paper?

10.21203/rs.3.rs-10892753/v1, permanently at https://doi.org/10.21203/rs.3.rs-10892753/v1. Sign in and open it from this page.

Bibliographic data from OpenAlex (CC0), last checked 20 November 2026. ResearchGuru does not host this paper; once you are signed in, the links open it at the publisher or the repository that does.