基于多模态数据融合的改进中尺度涡检测模型-海洋预报

基于多模态数据融合的改进中尺度涡检测模型

单位：1. 中国石油大学(华东)海洋空间与信息学院, 山东青岛 266400;
2. 中国石油大学(华东)青岛软件学院, 计算机科学与技术学院, 山东青岛 266400

分类号：P731.2

出版年·卷·期（页码）：2024·41·第二期（53-62）

摘要：

提出一种基于多模态数据融合的改进中尺度涡检测模型。该模型以海平面高度数据为基础,首次将融合表层海温数据扩展为融合多深度层海温数据;将海温数据的深度层作为通道,嵌入通道注意力机制,使得模型能够关注于海水温度数据中最具有区分度的深度层;模型在编码及解码过程中采用残差学习单元,在加深网络深度的同时,更好地拟合激活函数,缓解训练问题,以提高模型的检测准确率。以中国南海部分海域为例开展实验验证,结果表明该中尺度涡检测模型准确率达到93.62 %,模型具备有效性和可靠性。

In this paper, an improved mesoscale eddy detection model based on multimodal data fusion is proposed. On the basis of sea level height data, the model extends the sea surface temperature fusion method into multi-depth ocean temperature fusion method for the first time. Taking the depth layer of ocean temperature data as a channel with channel attention mechanism, the model can focus on the depth layer with the most distinguishing degree in ocean temperature data. Residual learning unit is used in the encoding and decoding processes to improve the detection accuracy of the model, which not only deepen the depth of the network, but also better fit the activation function and alleviate the training problem. Utilizing the model in the South China Sea mesoscale eddy detection shows that the accuracy rate of the mesoscale eddy detection reaches 93.62 %. The effective and reliable results suggest that the model can provide a new and reasonable idea for the research of mesoscale eddy detection.

参考文献：

[1] D'ALIMONTE D. Detection of mesoscale eddy-related structures through Iso-SST patterns[J]. IEEE Geoscience and Remote Sensing Letters, 2009, 6(2):189-193.
[2] CHEN G, EZRATY R. Non-tidal aliasing in seasonal sea-level variability and annual Rossby waves as observed by satellite altimetry[J]. Annales Geophysicae, 1997, 15(11):1478-1488.
[3] 白志鹏,韩君,郭贤鹏,等.基于CORA2再分析数据的南海中尺度涡时空分布特征初步研究[J].海洋预报, 2020, 37(2):73-83.BAI Z P, HAN J, GUO X P, et al. Spatial and temporal distribution characteristics of mesoscale eddies in the South China Sea based on the CORA2 reanalysis data[J]. Marine Forecasts, 2020, 37(2):73-83.
[4] WYRTKI K, MAGAARD L, HAGER J. Eddy energy in the oceans[J]. Journal of Geophysical Research, 1976, 81(15):2641-2646.
[5] OKUBO A. Horizontal dispersion of floatable particles in the vicinity of velocity singularities such as convergences[J]. Deep Sea Research and Oceanographic Abstracts, 1970, 17(3):445-454.
[6] WEISS J. The dynamics of enstrophy transfer in two-dimensional hydrodynamics[J]. Physica D:Nonlinear Phenomena, 1991, 48(2-3):273-294.
[7] SADARJOEN I A, POST F H. Detection, quantification, and tracking of vortices using streamline geometry[J]. Computers&Graphics, 2000, 24(3):333-341.
[8] NENCIOLI F, DONG C M, DICKEY T, et al. A vector geometrybased eddy detection algorithm and its application to a highresolution numerical model product and high-frequency radar surface velocities in the southern California bight[J]. Journal of Atmospheric and Oceanic Technology, 2010, 27(3):564-579.
[9] 杜艳玲,刘倩倩,王丽丽,等.融合多尺度旋转锚机制的海洋中尺度涡自动检测[J].中国图象图形学报, 2022, 27(10):3092-3101.DU Y L, LIU Q Q, WANG L L, et al. Multi-scale rotating anchor mechanism based automatic detection of ocean mesoscale eddy[J].Journal of Image and Graphics, 2022, 27(10):3092-3101.
[10] XU G J, CHENG C, YANG W X, et al. Oceanic eddy identification using an AI scheme[J]. Remote Sensing, 2019, 11(11):1349.
[11] 董子意,杜震洪,吴森森,等.基于改进U-Net网络的海洋中尺度涡自动检测模型[J].海洋学报, 2022, 44(2):123-131.DONG Z Y, DU Z H, WU S S, et al. An automatic marine mesoscale eddy detection model based on improved U-Net network[J]. Haiyang Xuebao, 2022, 44(2):123-131.
[12] MOSCHOS E, SCHWANDER O, STEGNER A, et al. Deep-SSTeddies:a deep learning framework to detect oceanic eddies in sea surface temperature images[C]//Proceedings of 2020 IEEE International Conference on Acoustics, Speech and Signal Processing. Barcelona, Spain:IEEE, 2020.
[13] MOSCHOS E, KUGUSHEVA A, COSTE P, et al. Computer vision for ocean eddy detection in infrared imagery[C]//Proceedings of 2023 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV). Waikoloa, HI, USA:IEEE, 2023:6384-6393.
[14] BAI X, WANG C B, LI C H. A streampath-based RCNN approach to ocean eddy detection[J]. IEEE Access, 2019, 7:106336-106345.
[15] LIU Y J, ZHENG Q A, LI X F. Detection and analysis of mesoscale eddies based on deep learning[M]//LI X F, WANG F.Artificial Intelligence Oceanography. Singapore:Springer, 2023.
[16] FAN Z L, ZHONG G Q, WEI H X, et al. EDNet:a mesoscale eddy detection network with multi-modal data[C]//Proceedings of 2020 International Joint Conference on Neural Networks (IJCNN). Glasgow, UK:IEEE, 2020.
[17] LGUENSAT R, SUN M, FABLET R, et al. EddyNet:a deep neural network for pixel-wise classification of oceanic eddies[C]//IGARSS 2018-2018 IEEE International Geoscience and Remote Sensing Symposium. Valencia, Spain:IEEE, 2018.
[18] WANG Q L, WU B G, ZHU P F, et al. ECA-Net:efficient channel attention for deep convolutional neural networks[C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). Seattle, WA, USA:IEEE, 2020.
[19] RONNEBERGER O, FISCHER P, BROX T. U-Net:convolutional networks for biomedical image segmentation[C]//Proceedings of the 18th Medical Image Computing and ComputerAssisted Intervention. Munich, Germany:Springer, 2015.
[20] 谢旭丹,王静,储小青,等.南海中尺度涡温盐异常三维结构[J].海洋学报, 2018, 40(4):1-14.XIE X D, WANG J, CHU X Q, et al. Three-dimensional thermohaline anomaly structures of mesoscale eddies in the South China Sea[J]. Haiyang Xuebao, 2018, 40(4):1-14.
[21] 赵福,张蕴斐,朱学明,等.冬季台湾西南海域一对冷、暖中尺度涡的同化模拟研究[J].海洋预报, 2017, 34(5):1-15.ZHAO F, ZHANG Y F, ZHU X M, et al. An assimilative numerical study of the paired cold and warm mesoscale eddies during winter in the Southwest of Taiwan[J]. Marine Forecasts,2017, 34(5):1-15.
[22] HU J, SHEN L, ALBANIE S, et al. Squeeze-and-excitation networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8):2011-2023.
[23] HE K M, ZHANG X Y, REN S Q, et al. Identity mappings in deep residual networks[C]//Proceedings of the 14th European Conference on Computer Vision-ECCV 2016. Amsterdam, The Netherlands:Springer, 2016.
[24] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, USA:IEEE, 2016.
[25] MILLETARI F, NAVAB N, AHMADI S A. V-Net:fully convolutional neural networks for volumetric medical image segmentation[C]//Proceedings of the 4th International Conference on 3D Vision. Stanford, CA, USA:IEEE, 2016.
[26] MASON E, PASCUAL A, MCWILLIAMS J C. A new sea surface height-based code for oceanic mesoscale eddy tracking[J].Journal of Atmospheric and Oceanic Technology, 2014, 31(5):1181-1188.
[27] 郑全安,谢玲玲,郑志文,等.南海中尺度涡研究进展[J].海洋科学进展, 2017, 35(2):131-158.ZHENG Q A, XIE L L, ZHENG Z W, et al. Progress in research of mesoscale eddies in the South China Sea[J]. Advances in Marine Science, 2017, 35(2):131-158.
[28] 王萌,张艳伟,刘志飞,等.南海北部中尺度涡的时空分布特征:基于卫星高度计资料的统计分析[J].地球科学进展, 2019, 34(10):1069-1080.WANG M, ZHANG Y W, LIU Z F, et al. Temporal and spatial characteristics of mesoscale eddies in the northern South China Sea:statistics analysis based on altimeter data[J]. Advances in Earth Science, 2019, 34(10):1069-1080.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】