Reconstruction of subsurface ocean state variables using Convolutional Neural Networks with combined satellite and in situ data

Frontiers in Marine Science

Marine Science

Bruno Buongiorno Nardelli¹  Filipe Rodrigues²  Michael St. John³  Patrizio Mariani⁴  Anshul Chauhan⁴  Philip A. H. Smith⁴  Asbjørn Christensen⁴  Kristian Aa. Sørensen⁵ 

[1] Consiglio Nazionale delle Ricerche, Istituto di Scienze Marine (CNR-ISMAR), Naples, Italy;Machine Learning for Smart Mobility Group, Department of Technology, Management and Economics, Technical University of Denmark (DTU), Kongens Lyngby, Denmark;National Institute of Aquatic Resources, Section for Marine Living Resources, Technical University of Denmark (DTU), Lyngby, Denmark;National Institute of Aquatic Resources, Section for Oceans and Arctic, Technical University of Denmark (DTU), Lyngby, Denmark;National Space Institute of Denmark, Center for Security, Technical University of Denmark (DTU), Kongens Lyngby, Denmark;

关键词: 3D reconstruction;    remote sensing;    Convolutional Neural Networks;    sea surface temperature;    sea surface salinity;    earth observation;    hydrography;    ARGO;   

DOI  :  10.3389/fmars.2023.1218514

 received in 2023-05-07, accepted in 2023-08-09,  发布年份 2023

来源: Frontiers

【 摘 要 】

Subsurface ocean measurements are extremely sparse and irregularly distributed, narrowing our ability to describe deep ocean processes and thus also limiting our understanding of the role of ocean and marine ecosystems in the Earth system. To overcome these observational limitations, neural networks combining remotely-sensed surface measurements and in situ vertical profiles are increasingly being used to retrieve high-quality three-dimensional estimates of the ocean state. This study proposes a convolutional neural network (CNN) architecture for the reconstruction of vertical profiles of temperature and salinity starting from surface observation-based data. The model is trained on satellite and in situ data collected between 2005 and 2020 in the Atlantic Ocean. Rather than using spatially gridded in situ observations, we use directly measured vertical profiles. Different combinations of surface variables are analyzed and compared in order to determine the most effective inputs for the CNN. Furthermore, the relative importance of each of these variables in the vertical reconstruction is assessed using Shapley values, originally developed in the framework of cooperative game theory. The model performance is shown to be superior to current state-of-the-art methods and the same approach can easily be extended to other basins or to the global ocean.

【 授权许可】

Unknown   
Copyright © 2023 Smith, Sørensen, Buongiorno Nardelli, Chauhan, Christensen, St. John, Rodrigues and Mariani