Multi-Generator Tropical Cyclone Forecasting Based on Cross-Modal Fusion

13 Pages Posted: 8 Jun 2024

See all articles by Qian Liu

Qian Liu

Nanjing University of Posts and Telecommunications

Hu Sun

Nanjing University of Posts and Telecommunications

Yaocheng Gui

Nanjing University of Posts and Telecommunications

Guilan Dai

Tsinghua University

Guoqiang Zhou

Nanjing University of Posts and Telecommunications

Abstract

Accurate prediction of the trajectory and intensity of tropical cyclones (TCs) is crucial for disaster mitigation and prevention. Traditional TC forecasting methods suffer from high research costs, subjectivity, and low prediction accuracy, which pose challenges in meeting the demand for real-time weather forecasting. Recent advancements primarily rely on deep learning methods, which have shown some improvement in prediction performance. However, they have not fully capitalized on the correlation between features of different modalities, thus limiting the potential of multimodal feature fusion. To address this gap in current research, this paper introduces a novel multimodal feature fusion module integrated into a multi-branching model framework based on a Generative Adversarial Network with multiple generators. This framework aims to predict TC information and multiple potential motion trends at future time points. Additionally, a feature fusion loss function is devised to evaluate the effectiveness of the feature fusion module. To evaluate the effectiveness of the proposed model, extensive experiments are conducted on the CMA-BST dataset. Experimental findings reveal that the model's prediction performance surpasses that of existing deep learning methods and often outperforms the official predictor China Central Meteorological Observatory (CMO) across various metrics.

Keywords: Tropical cyclone prediction, cross-modal fusion, multiple generators, Attention mechanism

Suggested Citation

Liu, Qian and Sun, Hu and Gui, Yaocheng and Dai, Guilan and Zhou, Guoqiang, Multi-Generator Tropical Cyclone Forecasting Based on Cross-Modal Fusion. Available at SSRN: https://ssrn.com/abstract=4858194 or http://dx.doi.org/10.2139/ssrn.4858194

Qian Liu

Nanjing University of Posts and Telecommunications ( email )

China

Hu Sun

Nanjing University of Posts and Telecommunications ( email )

China

Yaocheng Gui

Nanjing University of Posts and Telecommunications ( email )

China

Guilan Dai

Tsinghua University ( email )

Beijing, 100084
China

Guoqiang Zhou (Contact Author)

Nanjing University of Posts and Telecommunications ( email )

China

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
36
Abstract Views
145
PlumX Metrics