Progress has been made in the inversion reconstruction of marine three-dimensional warm salt fields

Recently, Yin Baoshu, a researcher at the Institute of Oceanology, Chinese Academy of Sciences, and his team has made new progress in the inversion reconstruction of marine three-dimensional temperature and salt fields, and the research results have been published in the international academic journal “Frontier Science”.

Schematic diagram of the three-dimensional thermosalt field inversion reconstruction model in the tropical Indian Ocean, courtesy of the Institute of Oceanography

The three-dimensional marine thermosalt field is the basis for the study of ocean dynamics, reflecting the density distribution and movement of water in the ocean. Accurate estimation of the ocean’s temperature and salinity structure can help reveal the ocean’s circulation system, water mass formation, and movement paths, and thus understand ocean dynamics processes and their impact on climate change, marine ecosystems, and global circulation. However, due to the limitations of observation technology, it is still difficult to achieve direct observation of temperature structures with three-dimensional high spatiotemporal resolution, and satellite remote sensing can provide a variety of marine dynamic environmental parameters with high spatiotemporal coverage. Therefore, how to use high-resolution satellite remote sensing data combined with Argo and other observation data to invert the key dynamic environmental parameter fields in the ocean has become one of the important contents of physical oceanographic research. 

The reporter learned that based on multi-source satellite remote sensing surface data (including sea surface temperature, sea surface salinity, sea surface height and sea surface wind field, etc.) and Argo measured data, Yin Baoshu’s research team innovatively proposed a new model based on Convolutional Block Attention Module-Convolutional Neural Network (CBAM-CNN), which can simultaneously invert and reconstruct the three-dimensional warm salt field of the tropical Indian Ocean.

The results show that the CBAM-CNN model is significantly superior to the traditional convolutional neural network (CNN) model in estimating the structure of the warm salt field in the tropical Indian Ocean, and has excellent performance.

In addition, the research team confirmed the accuracy of the CBAM-CNN model in estimating ocean warm salt fields at different depths by comparing with Argo observation data, and demonstrated the effectiveness of the model in capturing observation features using sea surface data.

The study also confirmed that the CBAM-CNN model showed good adaptability in seasonal applications. The results of this research will provide important support for our in-depth understanding of ocean dynamics, promote research on marine environmental change, and respond to global climate change. 

The research was jointly funded by the Strategic Leading Science and Technology Project of the Chinese Academy of Sciences and the National Natural Science Foundation of China. (Source: China Science News, Liao Yang, Wang Min)

Related paper information:

Source link

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button