Published in a Q1 Journal

Congratulations to Dr. Le Duc Trong, Dr. Nguyen Thi Nhat Thanh, the GEOI research group (University of Engineering and Technology, VNU), and the partner research group at Hanoi University of Science and Technology on the successful publication of their scientific paper titled: “From reanalysis to climatology: deep learning reconstruction of tropical cyclogenesis in the western North Pacific”. The paper was published in Geoscientific Model Development – a prestigious journal under Copernicus Publications, ranked Q1 with an Impact Factor of 4.9.

The paper proposes a novel approach for the early forecasting of tropical cyclogenesis using Artificial Intelligence, a highly urgent topic in the context of current complex climate change. The research is not only of high academic value but also carries profound practical significance. It aims to contribute a support tool for disaster forecasting and prevention for government agencies at all levels, hoping to share the hardships of the people in Central Vietnam, a region that frequently suffers damage from annual storms and floods.

By applying the ResNet-18 deep learning architecture—a popular architecture in the field of Computer Vision—the study demonstrated the capability of this deep learning model in the early forecasting of tropical cyclogenesis with a lead time of up to 2 days. Furthermore, the forecasting domain was expanded from the East Sea to cover the vast spatial region of the western North Pacific.

Notably, the research successfully reconstructed the monthly distribution map of tropical cyclogenesis locations and frequencies for the period from 2017 to 2022. This result provides a useful database to analyze the specific impacts of climate change on the tropical cyclogenesis process, laying the groundwork for long-term studies on climate trends in the coming decades.