AUTHOR=Rong Lintai , Xiong Xuejun , Chen Liang TITLE=An automatic identification algorithm of internal solitary wave for mooring data based on geometric characteristics of the flow field JOURNAL=Frontiers in Marine Science VOLUME=Volume 10 - 2023 YEAR=2023 URL=https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2023.1147268 DOI=10.3389/fmars.2023.1147268 ISSN=2296-7745 ABSTRACT=Internal solitary waves (ISWs) are active in the South China Sea (SCS). With characteristics such as large amplitude, strong current, and concentrated energy, ISW has become one of the major risks that must be considered in offshore engineering and underwater navigation. Therefore, in this study, an automatic identification algorithm of the ISWs was proposed for moorings data based on the geometric characteristics of the flow field. The algorithm involved five constraints, which were applied to all the grid points in the velocity vector field of the current profile, and the points that satisfied all the constraints were considered the centers of the ISW currents, thus the ISWs were successfully identified. Furthermore, the start-stop time of the ISW can be determined by the wavelet transform, and then the characteristic parameters (amplitude, propagation velocity, etc.) of the ISW can be obtained. In addition, the range of detection parameters π‘Ÿ 𝑑 and π‘Ÿ 𝑧 , which constituted a rectangular detection box with the size of 2π‘Ÿ 𝑑 Γ— 2π‘Ÿ 𝑧 , should be set based on the constraints for the detection of all the points in the velocity field. Since the values of these parameters can directly affect the accuracy of the algorithm, the sensitivity experiments were conducted based on the dataset of a mooring station deployed on the west slope of Dongsha Islands in the northern SCS. Subsequently, the optimal parameter combination of π‘Ÿ 𝑑 = 13 and π‘Ÿ 𝑧 = 3 (which constituted a detection box of 13 minΓ—60 m) was obtained. The true positive 批注 [SF1]: Reviewer #3 Comment #1 rate (𝑇𝑃𝑅) of ISWs was 87.2%, and the false negative rate (𝐹𝑁𝑅) was 16.3%. Finally, the algorithm was applied to the characteristic parameters of the ISWs and its reliability was further verified, indicating that the algorithm can effectively detect the ISWs and provide a reference for preventing and avoiding ISWs in ocean engineering.