Multi-timescale variation of temperature fronts in the Yellow Sea during winter and spring and its main controlling factors analysis
-
摘要: 海洋锋面强度变化对陆源物质输运和全球物质循环有重要作用。冬春季节，中国东部陆架区西太平洋边界流分支与沿岸流之间形成了海洋温度锋。为探究冬季风暴和陆架环流双重影响下温度锋面的多时间尺度变化及主控因素，本文以黄海为研究区，分别在年代际尺度和天气尺度，利用信号分解和可解释深度学习方法，研究了低纬度驱动的环流系统和高纬度驱动的冬季风暴对锋面变化的耦合作用。在年代际尺度，通过使用经验正交函数分解和集合经验模态分解的方法，将北黄海的温度变化与黄海暖流强度相联系。研究结果表明，黄海的海表温度经验正交函数（EOF）分解第一模态的空间分布有明显的黄海暖流—沿岸流体系特征；海表温度EOF第一模态时间序列与黄海暖流强度指标的相关性良好，且受低频率厄尔尼诺‒南方涛动信号调控。在天气尺度，对卷积神经网络−长短时记忆网络（CNN-LSTM）模型进行训练并使用可解释性指标进行分析，结果发现无风或弱风条件下，海洋锋面主要由压力梯度力和科里奥利力的地转平衡维持；但在冬季风暴条件下，受开尔文波传播和切变锋破碎的影响，流场的低频波动成为导致锋面强度变化的主因。本文研究结果表明，大数据及机器学习方法是在众多海洋参数间建立联系，并发现一些独特物理海洋过程的重要手段，具有广阔的应用前景。Abstract: Ocean fronts variations in strength are key to the terrestrial material transport and global material cycle. Ocean temperature fronts are formed between the branches of West Pacific Boundary Current and the coastal current during the winter and spring seasons in the eastern shelf of China. In order to investigate the multi-time scale variation and main controlling factors of temperature front over the Yellow Sea under the dual influence of winter storms and shelf circulation, we investigate the coupling of low-latitude driven circulation systems and high-latitude driven winter storms on frontal variability with the methods of signal decomposition and explainable deep learning on the decadal and weather scale. On the decadal scale, empirical orthogonal decomposition and ensemble empirical modal decomposition are used to relate temperature changes in the Yellow Sea to the strength of the Yellow Sea Warm Current. The results indicate that the spatial distribution of first sea surface temperature (SST) EOF mode has obvious characteristics of the Yellow Sea Warm Current-coastal current system; the time series of the first SST EOF mode correlates well with the Yellow Sea Warm Current intensity index and is modulated by the low frequency ENSO signal. On the weather scale, this paper trains CNN-LSTM models and uses interpretability metrics to conduct the research. The results show that, in windless or weak wind conditions, the strength of ocean front is maintained by the combination of pressure gradient forces resulted from sea surface height and Coriolis forces caused by flow field. However, in the storm conditions, influenced by Kelvin Wave propagation and shear front fragmentation, the flow field is responsible for the ocean front variation. The results of this study show that big data and machine learning methods are important means to establish connections between many ocean parameters and discover some unique physical ocean processes, which have broad application prospects.
-
Key words:
- temperature front /
- big data analysis /
- winter storms /
- Yellow Sea Warm Current /
- coastal currents
-
图 1 研究区域冬季（12月至翌年3月）平均海表温度（a）和冬季平均温度梯度（b）
a中箭头代表研究区域冬季平均20 m水深流场
Fig. 1 Mean sea surface temperature (a) and gradient (b) in the study domain during winter (december to march )
The arrows in a represent the mean 20 m water depth flow field during winter
图 2 海表温度经验正交函数（EOF）分解第一模态空间分布（a），以及海表温度EOF分解第一模态（相反数）和黄海暖流强度指标第六、第七本征模态（IMF 6+7）时间序列变化（b）
Fig. 2 The first-mode spatial distribution pattern of sea surface temperature empirical orthogonal function (EOF) decomposition (a); time series of the first mode decomposed by EOF (opposite number) and sum of the sixth and seventh intrinsic mode functions of the Yellow Sea Warm Current intensity indicator (b)
图 3 空间相关性关系
a. 海表温度（SST）与纬向风场；b. SST与经向风场；c. 经向风场与$\Delta $SST；d. SST与表层纬向流场；e. SST与表层经向流场；f. SST与40 m水深纬向流场；g. SST与40 m水深经向流场；h. 经向风场与20 m水深位温变化量
Fig. 3 Spatial correlations
a. SST and zonal wind; b. SST and meridional wind; c. meridional wind field and $\Delta $SST; d. SST and surface zonal flow; e. SST and surface meridional flow; f. SST and zonal flow at 40 m water depth; g. SST and meridional flow at 40 m water depth; h. meridional wind and 20 m water depth potential temperature change
图 4 纬向（a）和经向（b）流场平均积分梯度空间分布
Fig. 4 Zonal (a) and meridional (b) spatial pattern of the flow field integrated gradient
图 5 2019年12月至2020年3月研究区域各要素日均变化
a. 风向序列；b. 鲁北锋面强度；c. 鲁北锋面磷酸盐内外浓度差；d. 辽南、朝鲜半岛表层锋面强度
Fig. 5 Oceanographic daily average of study area during December 2019 to March 2020
a. Wind sequence; b. strength of the Lubei front; c. difference between internal and external phosphate concentrations at the Lubei front; d. strength of surface fronts in Liaonan and Korea Peninsula
图 6 黄海暖流强度与锋面强度散点图（a）及Niño3与黄海暖流强度指标时间序列（b）
Fig. 6 Scatterplot of Yellow Sea Warm Current (YSWC) strength versus frontal strength (a), and the time series of Niño3 and YSWC strength index (b)
图 7 2020年2月13‒20日风序列（a），3条主要锋面强度（b），锋面内3点（P1、P3、P5）海表高度变化（c）及神经网络9个输入变量平均积分梯度序列（d）
Fig. 7 Wind sequence for February 13‒20, 2020 (a), intensity of the three main fronts （b）, sea surface height variation at three points (P1, P3, P5) within the front (c), and neural network mean integral gradient series of nine input variables (d)
图 8 海表高度异常值（a‒f）和积分梯度场（g‒j）
SSH：海表高度；V：经向流速
Fig. 8 Sea surface height anomalies (a‒f), and integrated gradient field (g‒j)
SSH: sea surface heigh; V: meridional velocity
图 9 122°E断面流速（a, b）和37°N断面流速（c, d）
Fig. 9 Flow veloctiy at Section 122°E (a, b) and Section 37°N (c, d)
S2 研究区域37°N断面平均位温（a）,黄海暖流指标EEMD分解结果（仅展示分析使用的6～9本征模态函数，b‒e）
S2 Mean potential temperature at the Section 37°N in the study area (a), EEMD decomposition results for the Yellow Sea Warm Current indicator (only the 6‒9 IMFs used in the analysis are shown, b‒e)
S1 CNN-LSTM模型架构
S1 CNN-LSTM model structure
层序号 神经网络架构 类型 输出参数 参数 1 输入层 85×121×9 核尺寸：（1,1）
外缘填充:02 卷积层 85×121×32 核尺寸：（1,1）
外缘填充:03 卷积层 85×121×64 核尺寸：（3,3）
外缘填充：14 卷积层 85×121×128 核尺寸：（1,1）
外缘填充:05 卷积层 85×121×64 核尺寸：（1,1）
外缘填充:06 卷积层 85×121×32 核尺寸：（1,1）
外缘填充:07 卷积层 85×121×10 核尺寸：（1,1）
外缘填充:08 卷积层 85×121×8 核尺寸：（3,3）
外缘填充：19 卷积层 85×121×3 核尺寸：（1,1）
外缘填充:010 LSTM层 85×121×1 11 全连接层 10285×1 12 全连接层 10285×1 13 全连接层 10285×1 14 全连接层 10285×1 
下载: 导出CSV
S2 CNN-LSTM模型9个输入变量的平均积分梯度
S2 The mean integrated gradient of nine kinds of parameters which is input to the CNN-LSTM model
纬向流速/10−5 经向流速/10−5 海表高度/10−5 合计/10−5 当天 1.683 1.725 1.887 5.295 1天前 2.387 1.154 0.857 4.398 2天前 0.725 2.558 0.857 4.140 合计 4.795 5.437 3.601 13.833
下载: 导出CSV
-
[1] 汤毓祥, 郑义芳. 关于黄、东海海洋锋的研究[J]. 海洋通报, 1990, 9(5): 89−96.Tang Yuxiang, Zheng Yifang. Research on fronts in East China Sea[J]. Marine Science Bulletin, 1990, 9(5): 89−96. [2] 冯士筰, 李凤岐, 李少菁. 海洋科学导论[M]. 北京: 高等教育出版社, 1999.Feng Shizuo, Li Fengqi, Li Shaojing. An Introduction to Marine Science[M]. Beijing: Higher Education Press, 1999. [3] Lohmann R, Belkin I M. Organic pollutants and ocean fronts across the Atlantic Ocean: a review[J]. Progress in Oceanography, 2014, 128: 172−184. doi: 10.1016/j.pocean.2014.08.013 [4] Chen Dake, Liu W T, Tang Wenqing, et al. Air-sea interaction at an oceanic front: implications for frontogenesis and primary production[J]. Geophysical Research Letters, 2003, 30(14): 1745. [5] 宁修仁, 史君贤, 蔡昱明, 等. 长江口和杭州湾海域生物生产力锋面及其生态学效应[J]. 海洋学报, 2004, 26(6): 96−106.Ning Xiuren, Shi Junxian, Cai Yuming, et al. Biological productivity front in the Changjiang Estuary and the Hangzhou Bay and its ecological effects[J]. Haiyang Xuebao, 2004, 26(6): 96−106. [6] 艾乔, 石勇, 高建华, 等. 辽东半岛东岸近海泥区悬沙浓度的时空分布及控制因素分析[J]. 海洋学报, 2019, 41(1): 121−133.Ai Qiao, Shi Yong, Gao Jianhua, et al. Spatio-temporal distribution and control factors of surface suspended sediment concentration in the mud deposition along eastern coast offshore of the Liaodong Peninsula[J]. Haiyang Xuebao, 2019, 41(1): 121−133. [7] Zhong Yi, Qiao Lulu, Song Dehai, et al. Impact of cold water mass on suspended sediment transport in the South Yellow Sea[J]. Marine Geology, 2020, 428: 106244. doi: 10.1016/j.margeo.2020.106244 [8] Zhou Feng, Xue Huijie, Huang Daji, et al. Cross-shelf exchange in the shelf of the East China Sea[J]. Journal of Geophysical Research: Oceans, 2015, 120(3): 1545−1572. doi: 10.1002/2014JC010567 [9] Owen R W. Fronts and eddies in the sea: mechanisms, interactions and biological effects[M]//Longhurst A R. Analysis of Marine Ecosystems. London: Academic Press, 1981: 197−233. [10] 任诗鹤, 王辉, 刘娜. 中国近海海洋锋和锋面预报研究进展[J]. 地球科学进展, 2015, 30(5): 552−563. doi: 10.11867/j.issn.1001-8166.2015.05.0552Ren Shihe, Wang Hui, Liu Na. Review of ocean front in Chinese marginal seas and frontal forecasting[J]. Advances in Earth Science, 2015, 30(5): 552−563. doi: 10.11867/j.issn.1001-8166.2015.05.0552 [11] Shi Y, Gao Jianhua, Sheng Hui, et al. Cross-front sediment transport induced by quick oscillation of the Yellow Sea Warm Current: evidence from the sedimentary record[J]. Geophysical Research Letters, 2019, 46(1): 226−234. doi: 10.1029/2018GL080751 [12] 吴德星, 兰健. 中国东部陆架边缘海海洋物理环境演变及其环境效应[J]. 地球科学进展, 2006, 21(7): 667−672.Wu Dexing, Lan Jian. Marine physical variations in eastern marginal seas of China and their environmental impacts[J]. Advances in Earth Science, 2006, 21(7): 667−672. [13] Ichikawa H, Beardsley R C. The current system in the Yellow and East China Seas[J]. Journal of Oceanography, 2002, 58(1): 77−92. doi: 10.1023/A:1015876701363 [14] 朱伟军, 李莹. 冬季北太平洋风暴轴的年代际变化特征及其可能影响机制[J]. 气象学报, 2010, 68(4): 477−486.Zhu Weijun, Li Ying. Inter-decadal variation characteristics of winter North Pacific storm tracks and its possible influencing mechanism[J]. Acta Meteorologica Sinica, 2010, 68(4): 477−486. [15] Taguchi B, Xie Shangping, Schneider N, et al. Decadal variability of the kuroshio extension: observations and an eddy-resolving model hindcast[J]. Journal of Climate, 2007, 20(11): 2357−2377. doi: 10.1175/JCLI4142.1 [16] Li Chunyan, Nelson J R, Koziana J V. Cross-shelf passage of coastal water transport at the South Atlantic Bight observed with MODIS Ocean Color/SST[J]. Geophysical Research Letters, 2003, 30(5): 1257. [17] Wang Chenghao, Liu Zhiqiang, Harris C K, et al. The impact of winter storms on sediment transport through a narrow strait, Bohai, China[J]. Journal of Geophysical Research: Oceans, 2020, 125(6): e2020JC016069. [18] Wu Xiaodong, Voulgaris G, Kumar N. Shelf cross-shore flows under storm-driven conditions: role of stratification, shoreline orientation, and bathymetry[J]. Journal of Physical Oceanography, 2018, 48(11): 2533−2553. doi: 10.1175/JPO-D-17-0090.1 [19] Lima E, Sun Xin, Dong Junyu, et al. Learning and transferring convolutional neural network knowledge to ocean front recognition[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(3): 354−358. doi: 10.1109/LGRS.2016.2643000 [20] Sun Jianyuan, Zhong Guoqiang, Dong Junyu, et al. Cooperative profit random forests with application in ocean front recognition[J]. IEEE Access, 2017, 5: 1398−1408. doi: 10.1109/ACCESS.2017.2656618 [21] Qiao Baiyou, Wu Zhongqiang, Tang Zhong, et al. Sea surface temperature prediction approach based on 3D CNN and LSTM with attention mechanism[C]//Proceedings of the 23rd International Conference on Advanced Communication Technology (ICACT). PyeongChang: IEEE, 2021. [22] Ham Y G, Kim J H, Luo Jingjia. Deep learning for multi-year ENSO forecasts[J]. Nature, 2019, 573(7775): 568−572. doi: 10.1038/s41586-019-1559-7 [23] Tang Meng, Liu Yimin, Durlofsky L J. A deep-learning-based surrogate model for data assimilation in dynamic subsurface flow problems[J]. Journal of Computational Physics, 2020, 413: 109456. doi: 10.1016/j.jcp.2020.109456 [24] 刘传玉, 王凡. 黄海暖流源区海表面温度锋面的结构及季节内演变[J]. 海洋科学, 2009, 33(7): 87−93.Liu Chuanyu, Wang Fan. Distributions and intra-seasonal evolutions of the sea surface thermal fronts in the Yellow Sea Warm Current origin area[J]. Marine Sciences, 2009, 33(7): 87−93. [25] Poitevin C, Wöppelmann G, Raucoules D, et al. Vertical land motion and relative sea level changes along the coastline of Brest (France) from combined space-borne geodetic methods[J]. Remote Sensing of Environment, 2019, 222: 275−285. doi: 10.1016/j.rse.2018.12.035 [26] Duan Haiqin, Xu Jingping, Wu Xiao, et al. Periodic oscillation of sediment transport influenced by winter synoptic events, Bohai Strait, China[J]. Water, 2020, 12(4): 986. doi: 10.3390/w12040986 [27] Wu Xiao, Wu Hui, Wang Houjie, et al. Novel, repeated surveys reveal new insights on sediment flux through a Narrow Strait, Bohai, China[J]. Journal of Geophysical Research: Oceans, 2019, 124(10): 6927−6941. doi: 10.1029/2019JC015293 [28] Lellouche J M, Greiner E, Le Galloudec O, et al. Recent updates to the Copernicus Marine Service global ocean monitoring and forecasting real-time 1∕12° high-resolution system[J]. Ocean Science, 2018, 14(5): 1093−1126. doi: 10.5194/os-14-1093-2018 [29] Lellouche J M, Le Galloudec O, Drévillon M, et al. Evaluation of global monitoring and forecasting systems at Mercator Océan[J]. Ocean Science, 2013, 9(1): 57−81. doi: 10.5194/os-9-57-2013 [30] Hersbach H, Bell B, Berrisford P, et al. The ERA5 global reanalysis[J]. Quarterly Journal of the Royal Meteorological Society, 2020, 146(730): 1999−2049. doi: 10.1002/qj.3803 [31] Molina M O, Gutiérrez C, Sánchez E. Comparison of ERA5 surface wind speed climatologies over Europe with observations from the HadISD dataset[J]. International Journal of Climatology, 2021, 41(10): 4864−4878. doi: 10.1002/joc.7103 [32] Hannachi A. A primer for EOF analysis of climate data[D]. Reading: Department of Meteorology, University of Reading, 2004. [33] Wu Zhaohua, Huang N E. Ensemble empirical mode decomposition: a noise-assisted data analysis method[J]. Advances in Adaptive Data Analysis, 2009, 1(1): 1−41. doi: 10.1142/S1793536909000047 [34] 周飞燕, 金林鹏, 董军. 卷积神经网络研究综述[J]. 计算机学报, 2017, 40(6): 1229−1251.Zhou Feiyan, Jin Linpeng, Dong Jun. Review of convolutional neural network[J]. Chinese Journal of Computers, 2017, 40(6): 1229−1251. [35] Han Mingxu, Feng Yuan, Zhao Xueli, et al. A convolutional neural network using surface data to predict subsurface temperatures in the Pacific Ocean[J]. IEEE Access, 2019, 7: 172816−172829. doi: 10.1109/ACCESS.2019.2955957 [36] Aires F, Boucher E, Pellet V. Convolutional neural networks for satellite remote sensing at coarse resolution. Application for the SST retrieval using IASI[J]. Remote Sensing of Environment, 2021, 263: 112553. doi: 10.1016/j.rse.2021.112553 [37] Wu Hongcai, Yang Qinli, Liu Jiaming, et al. A spatiotemporal deep fusion model for merging satellite and gauge precipitation in China[J]. Journal of Hydrology, 2020, 584: 124664. doi: 10.1016/j.jhydrol.2020.124664 [38] 周锋, 黄大吉, 万瑞景, 等. 南黄海西北部夏季潮锋的观测和分析[J]. 海洋学报, 2008, 30(3): 9−15.Zhou Feng, Huang Daji, Wan Ruijing, et al. Observations and analysis of tidal fronts in the southwestern Huanghai Sea[J]. Haiyang Xuebao, 2008, 30(3): 9−15. [39] 赵保仁. 黄海冷水团锋面与潮混合[J]. 海洋与湖沼, 1985, 16(6): 451−460.Zhao Baoren. The fronts of the Huanghai Sea cold water mass induced by tidal mixing[J]. Oceanologia et Limnologia Sinica, 1985, 16(6): 451−460. [40] Wang Fan, Liu Chuanyu, Meng Qingjia. Effect of the Yellow Sea Warm Current fronts on the westward shift of the Yellow Sea Warm Tongue in winter[J]. Continental Shelf Research, 2012, 45: 98−107. doi: 10.1016/j.csr.2012.06.005 [41] Xu Xiaomei, Gao Jianhua, Shi Yong, et al. Cross-front transport triggered by winter storms around the Shandong Peninsula, China[J]. Frontiers in Marine Science, 2022, 9: 975504. doi: 10.3389/fmars.2022.975504 [42] 成科扬, 王宁, 师文喜, 等. 深度学习可解释性研究进展[J]. 计算机研究与发展, 2020, 57(6): 1208−1217.Cheng Keyang, Wang Ning, Shi Wenxi, et al. Research advances in the interpretability of deep learning[J]. Journal of Computer Research and Development, 2020, 57(6): 1208−1217. [43] Sundararajan M, Taly A, Yan Qiqi. Axiomatic attribution for deep networks[C]//Proceedings of the 34th International Conference on Machine Learning. Sydney: JMLR, 2017. [44] Mudrakarta P K, Taly A, Sundararajan M, et al. Did the model understand the question?[C]//Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics. Melbourne: ACL, 2018. [45] Smilkov D, Thorat N, Kim B, et al. SmoothGrad: removing noise by adding noise[EB/OL]. （2017‒06‒12）[2022‒07‒17]. https://arxiv.org/abs/1706.03825. [46] Shi Yong, Xu Xiaomei, Sheng Hui, et al. Neglected role of continental circulation in cross-shelf sediment transport: implications for paleoclimate reconstructions[J]. Marine Geology, 2022, 443: 106703. doi: 10.1016/j.margeo.2021.106703 [47] 石勇. 北黄海西部细颗粒物质的跨锋面输运及其沉积环境效应[D]. 南京: 南京大学, 2020.Shi Yong. Cross-front transport of fine sediment in the western North Yellow Sea and its sedimentary effects[D]. Nanjing: Nanjing University, 2020. [48] Roundy P E, Kiladis G N. Observed relationships between oceanic Kelvin waves and atmospheric forcing[J]. Journal of Climate, 2006, 19(20): 5253−5272. doi: 10.1175/JCLI3893.1 [49] Jacobs G A, Preller R H, Riedlinger S K, et al. Coastal wave generation in the Bohai Bay and propagation along the Chinese coast[J]. Geophysical Research Letters, 1998, 25(6): 777−780. doi: 10.1029/97GL03636 [50] Wu Hui. Cross-shelf penetrating fronts: a response of buoyant coastal water to ambient pycnocline undulation[J]. Journal of Geophysical Research: Oceans, 2015, 120(7): 5101−5119. doi: 10.1002/2014JC010686 [51] Pi Zhong, Chang Fengming, Li Tiegang, et al. Sea surface temperature evolution in the Yellow Sea Warm Current pathway and its teleconnection with high and low latitude forcing during the mid-late Holocene[J]. Journal of Oceanology and Limnology, 2022, 40(1): 93−109. doi: 10.1007/s00343-021-0219-6 [52] Mantua N J, Hare S R, Zhang Yuan, et al. A Pacific interdecadal climate oscillation with impacts on salmon production[J]. Bulletin of the American Meteorological Society, 1997, 78(6): 1069−1080. doi: 10.1175/1520-0477(1997)078<1069:APICOW>2.0.CO;2 [53] Power S, Casey T, Folland C, et al. Inter-decadal modulation of the impact of ENSO on Australia[J]. Climate Dynamics, 1999, 15(5): 319−324. doi: 10.1007/s003820050284 [54] Liu Zhiqiang, Gan Jianping. Modeling study of variable upwelling circulation in the East China Sea: response to a coastal promontory[J]. Journal of Physical Oceanography, 2014, 44(4): 1078−1094. doi: 10.1175/JPO-D-13-0170.1 [55] Zheng Xiangyang, Zhang Hua, Li Yanfang, et al. The features and mechanisms of the North Shandong Coastal Current: a case study in 2014[J]. Journal of Oceanography, 2021, 77(4): 631−646. doi: 10.1007/s10872-021-00597-3 -
计量
- 文章访问数: 652
- HTML全文浏览量: 244
- PDF下载量: 145
- 被引次数: 0
