AUTHOR=Chai Xingyu , Li Jincai , Zhao Jun , Hu Yifan , Zhao Xiaofeng 

TITLE=NWPP-EDH: Numerical weather prediction products evaporation duct regional prediction model

JOURNAL=Frontiers in Earth Science

VOLUME=Volume 10 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2022.1031186

DOI=10.3389/feart.2022.1031186

ISSN=2296-6463

ABSTRACT=The evaporation duct is a special atmospheric stratification that can affect the propagation path of electromagnetic waves at sea. Hence, it plays a critical role in determining the efficiency of the radio communication systems. The traditional theoretical models of the evaporation duct often have limited accuracy. The actual observational data from voyages and stations are insufficient, and the existing data-driven evaporation duct height prediction models can only predict a particular point or route but cannot reproduce the regional distribution of the evaporation duct. To address these issues, we used the numerical weather prediction products (NWPP) data and proposed a regional prediction model called the NWPP model to predict evaporation duct height based on a convolutional neural network. The fitting ability of the NWPP model was tested. Its accuracy was compared with that of the Babin–Young–Carton model, Musson–Gauthier–Bruth model, and the classical Naval Postgraduate School model; compared to these models, the NWPP model showed reductions of 71.8%, 87%, and 60.9%, respectively, in the root mean square error. Thus, we proved that the NWPP evaporation duct regional prediction model provide better accuracy than the traditional theoretical models.