Bathymetric Prediction Using Multisource Gravity Data Derived From a Parallel Linked BP Neural Network. Issue 11 (11th November 2022)

• Record Type: Journal Article
• Authors: Sun, Heyuan; Feng, Yikai; Fu, Yanguang; Sun, Weikang; Peng, Cong; Zhou, Xinghua; Zhou, Dongxu
• Is Part Of: Journal of geophysical research. Volume 127:Issue 11(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-11-11
• Subjects: bathymetric prediction -- data fusion -- neural network; Geomagnetism -- Periodicals; Geochemistry -- Periodicals; Geophysics -- Periodicals; Earth sciences -- Periodicals; 551.1
• DOI: 10.1029/2022JB024428
• Languages: English
• ISSNs: 2169-9313

Abstract

Abstract: Gravity anomalies (GAs) and vertical gravity gradient anomalies (VGGs) are regularly used to predict bathymetry. However, few studies have explored the combination of the GAs and VGGs to predict bathymetry. We introduce the back propagation (BP) neural network into bathymetric prediction field and propose a method to predict depth from GAs and VGGs. The method was tested in the Mariana Trench region, and a neural network bathymetry model was constructed using feature data obtained from GAs and VGGs as the input of the BP neural network. Additionally, single‐beam sounding data were used as label data. By comparing the neural network and the gravity‐geologic method (GGM), the neural network was found to provide better performance with an accuracy improvement of 19%. The root‐mean‐square of the absolute difference between the neural network bathymetry model and the single‐beam sounding data was 72.40 m, with a relative accuracy of 1.71%. Approximately 50% of the differences were distributed within ±20 m, and 90% were distributed within ±100 m. The neural network bathymetry model was also compared with the GGM bathymetry model for different depths and topographies, and the results verified the feasibility and effectiveness of the BP neural network method. Plain Language Summary: The gravity data currently used to predict bathymetry include gravity anomalies and vertical gravity gradient anomalies. We propose a new method that introduces the back propagation neural network into bathymetric prediction, based on two kinds of gravity data, which are used to predict depth. The results show that the depths predicted by the neural network are closer to the shipborne depths than are those estimated with the current method. We also test the adaptability of the neural network method for different depths and topographies (seamount, trench, and plain). Finally, we find that the BP neural network method is competent for various inversion tasks. Key Points: A new method is proposed to predict depth using gravity data derived from a back propagation neural network Bathymetric prediction using gravity anomalies and vertical gravity gradient anomalies The BP neural network method displays outstanding performance compared to the gravity‐geologic method for different depths and topographies