基于PCA-BP特征工程的近海单点海温预报技术及应用-海洋预报

基于PCA-BP特征工程的近海单点海温预报技术及应用

单位：1. 国家海洋环境预报中心自然资源部海洋灾害预报技术重点实验室, 北京 100081;
2. 山东省海洋预报减灾中心, 山东青岛 266104

分类号：P731.31

出版年·卷·期（页码）：2023·40·第三期（35-44）

摘要：

本文将主成分分析方法（Principal Components Analysis，PCA）和误差后传（Back Propagation，BP）神经网络相结合，建立了一种PCA-BP特征工程的近海单点海温智能预报模型，并对山东荣成近岸海域气象数值预报产品和在线海温监测仪连续观测数据开展了释用技术研究和应用。2021年业务化运行结果显示：该预报模型具有占用内存小、运行速度快、预报误差低的优点，相比近岸基础单元数值预报和经验预报的24 h均方根误差降幅达1.0℃和0.8℃，均方根相对误差降幅达12%～14%，未来48 h和72 h的预报误差也降幅明显，预报计算时间小于10 s，并将预报时效进一步向前扩展了3 d，达到144 h。

An intelligent forecasting model of offshore sea surface temperature (SST) at single-point based on PCA-BP feature engineering has been established in this paper by combining principal components analysis (PCA) and back propagation (BP) neural networks. The model has been tested and implemented using meteorological numerical forecast products and continuous in-situ observation data of on-line SST monitor in Rongcheng coastal waters, Shandong Province. The operational results in 2021 show that this forecasting model has the advantages of less memory occupation, faster running speed and lower forecasting error. Compared with the numerical forecasting of offshore basic units and empirical forecasting, the 24-hour root mean square error of SST from the intelligent forecasting model decreases by 1.0 ℃、 0.8 ℃, and the root mean square prediction error decreases by 12%～14%. The forecasting errors for 48 and 72 hours also significantly decrease. In addition, the forecasting calculation time is less than 10 s, and the forecasting time is further extended by 3 day to 144 h.

参考文献：

[1] ZHANG Q, WANG H, DONG J Y, et al. Prediction of sea surface temperature using long short-term memory[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(10):1745-1749.
[2] 国家海洋局海洋环境预报中心温盐组. 海温预报传真图的内容及其应用[J]. 海洋预报服务, 1984, 1(1):85-89. Temperature and Salt Group of the Marine Environment Prediction Center of the National Oceanic Administration. Content and application of sea surface temperature forecast facsimile chart[J]. Marine Forecast Service, 1984, 1(1):85-89.
[3] 李青青, 何佩东, 蒋风芝, 等. 连云港、吕泗海域的海表温度变化及影响机制[J]. 科学技术与工程, 2013, 13(11):2930-2937. LI Q Q, HE P D, JIANG F Z, et al. The Variation of sea surface temperature with its influencing mechanism of Lianyungang and Lüsi waters[J]. Science Technology and Engineering, 2013, 13(11):2930-2937.
[4] 邓冰, 张翔. 南海海温可预报性的诊断研究[J]. 海洋预报, 1997, 14(3):32-37. DENG B, ZHANG X. Diagnosis study on the prodictability of SST over South China Sea[J]. Marine Forecasts, 1997, 14(3):32-37.
[5] LINS I D, ARAUJO M, MOURA M D C, et al. Prediction of sea surface temperature in the tropical Atlantic by support vector machines[J]. Computational Statistics & Data Analysis, 2013, 61:187-198.
[6] ZHANG Y, ZHU J S, LI Z X, et al. Sea surface temperature predictions using a multi-ocean analysis ensemble scheme[J]. Climate Dynamics, 2017, 49(3):1049-1059.
[7] 张培军, 周水华, 梁昌霞. 基于卫星遥感海温数据的南海SST预报误差订正[J]. 热带海洋学报, 2020, 39(6):57-65. ZHANG P J, ZHOU S H, LIANG C X. Study on the correction of SST prediction in South China Sea using remotely sensed SST[J]. Journal of Tropical Oceanography, 2020, 39(6):57-65.
[8] 张建华. 海温预报知识讲座第一讲海水温度预报概况[J]. 海洋预报, 2003, 20(4):81-85. ZHANG J H. Lecture on sea temperature forecast:Lecture 1:overview of sea temperature forecast[J]. Marine Forecasts, 2003, 20(4):81-85.
[9] TONANI M, BALMASEDA M, BERTINO L, et al. Status and future of global and regional ocean prediction systems[J]. Journal of Operational Oceanography, 2015, 8(S2):S201-S220.
[10] PATIL K, DEO M C, RAVICHANDRAN M. Prediction of sea surface temperature by combining numerical and neural techniques[J]. Journal of Atmospheric and Oceanic Technology, 2016, 33(8):1715-1726.
[11] 刘娜, 王辉, 凌铁军, 等. 全球业务化海洋预报进展与展望[J]. 地球科学进展, 2018, 33(2):131-140. LIU N, WANG H, LING T J, et al. Review and prospect of global operational ocean forecasting[J]. Advances in Earth Science, 2018, 33(2):131-140.
[12] 郝日栩, 赵玉新, 何忠杰, 等. 基于EOF-NAR神经网络混合模型的海温预报方法研究[C]//中国海洋学会2019海洋学术(国际)双年会论文集. 三亚:海洋出版社, 2019:31-45. HAO R X, ZHAO Y X, HE Z J, et al. Research on SST forecasting method based on EOF-NAR neural network hybrid model[C]//2019 Oceanographic (International) Biennial Meeting of China Oceanographic Society. Sanya:Ocean Publishing House, 2019:31-45.
[13] 王金萍, 张翔宇. 宁德近海海温短期预报模型的初步研究[J]. 信息技术与信息化, 2019(2):186-187. WANG J P, ZHANG X Y. Preliminary study on short-term prediction model of offshore sea temperature in Ningde[J]. Information Technology and Informatization, 2019(2):186-187.
[14] 王兆毅, 李云, 王旭. 中国近岸海域基础预报单元海温预报指导产品研制[J]. 海洋预报, 2020, 37(4):59-65. WANG Z Y, LI Y, WANG X. Development of forecast guidance product for sea temperature of basic forecast units in the Chinese coastal waters[J]. Marine Forecasts, 2020, 37(4):59-65.
[15] 张志刚, 蔡东明, 骆永军. 周期外推方法在海水水温预报中的应用[J]. 海洋技术, 2007, 26(4):69-71. ZHANG Z G, CAI D M, LUO Y J. Application of distinct periodicity extrapolation to sea water temperature forecast[J]. Journal of Ocean Technology, 2007, 26(4):69-71.
[16] 陶祖钰, 赵翠光, 陈敏. 谈谈统计预报的必要性[J]. 气象科技进展, 2016, 6(1):6-13. TAO Z Y, ZHAO C G, CHEN M. The necessity of statistical forecasts[J]. Advances in Meteorological Science and Technology, 2016, 6(1):6-13.
[17] 李燕, 张建华, 刘钦政, 等. 单站海温短期预报自动化[J]. 海洋预报, 2007, 24(4):33-41. LI Y, ZHANG J H, LIU Q Z, et al. The automation of Single Sea Station's surface sea temperature short term forecasting[J]. Marine Forecasts, 2007, 24(4):33-41.
[18] 杜碧兰, 宋学家, 张建华. 表层海水温度场的正交综合因子场分解预报方法——东海及其外缘海域月平均表层水温预报[J]. 海洋学报, 1982, 4(2):149-156. DU B L, SONG X J, ZHANG J H. An empirical orthogonal multifactorial method designed to predict sea surface temperature-on the prediction of mean monthly sea surface temperature for the East China Sea and the adjacent waters[J]. Acta Oceanologica Sinica, 1982, 4(2):149-156.
[19] 张建华. 海温预报知识讲座第二讲数理统计方法在海温预报中的应用[J]. 海洋预报, 2004, 21(1):85-90. ZHANG J H. Lecture on SST forecast:lecture 2 application of mathematical statistics in SST forecast[J]. Marine Forecasts, 2004, 21(1):85-90.
[20] 侯瑞科. 平稳时间序列分析在海温预报中的应用[J]. 海洋预报, 1996, 13(1):41-45. HOU R K. Application of stationary time series analysis in SST prediction[J]. Marine Forecasts, 1996, 13(1):41-45.
[21] LEE D E, CHAPMAN D, HENDERSON N, et al. Multilevel vector autoregressive prediction of sea surface temperature in the North Tropical Atlantic Ocean and the Caribbean Sea[J]. Climate Dynamics, 2016, 47(1-2):95-106.
[22] 周林, 杨成荫, 王汉杰, 等. 基于CCA-BP-BPNN释用模型的太平洋SST预报[J]. 解放军理工大学学报(自然科学版), 2009, 10(4):391-396. ZHOU L, YANG C Y, WANG H J, et al. Interpretation scheme of SST predicition in the tropical Pacific Ocean based on CCA-BPBPNN[J]. Journal of PLA University of Science and Technology (Natural Science Edition), 2009, 10(4):391-396.
[23] 方玥炜, 唐佑民, 李俊德, 等. 几种统计模型对热带印度洋海温异常的预报[J]. 海洋学研究, 2018, 36(1):1-15. FANG Y W, TANG Y M, LI J D, et al. Several statistical models to predict tropical Indian Ocean sea surface temperature anomaly[J]. Journal of Marine Sciences, 2018, 36(1):1-15.
[24] XUE Y, LEETMAA A. Forecasts of tropical Pacific SST and sea level using a Markov model[J]. Geophysical Research Letters, 2000, 27(17):2701-2704.
[25] NEETU, SHARMA R, BASU S, et al. Data-adaptive prediction of sea-surface temperature in the Arabian sea[J]. IEEE Geoscience and Remote Sensing Letters, 2011, 8(1):9-13.
[26] 李繁华, 尹逊福, 曾宪模, 等. 复经验正交函数在水温预报中的应用[J]. 黄渤海海洋, 1996, 14(3):2-7. LI F H, YIN X F, ZENG X M, et al. An application of complex empirical orthogonal function in the prediction of sea temperature[J]. Journal of Oceanography of Huanghai & Bohai Seas, 1996, 14(3):2-7.
[27] 侯瑞科. 方差分析及其在海洋水温预报中的应用[J]. 海洋预报, 1994, 11(4):68-73. HOU R K. Analysis of variance and its application in ocean water temperature forecast[J]. Marine Forecasts, 1994, 11(4):68-73.
[28] 张韧, 王继光, 蒋国荣, 等. 非线性模糊识别及其在海温异常检测中的应用[J]. 地球科学进展, 2002, 17(4):470-476. ZHANG R, WANG J G, JIANG G R, et al. Non-linear fuzzy recognition and its application in identifying SST abnormality[J]. Advances in Earth Sciences, 2002, 17(4):470-476.
[29] WU A M, HSIEH W W, TANG B Y. Neural network forecasts of the tropical Pacific sea surface temperatures[J]. Neural Networks, 2006, 19(2):145-154.
[30] 董兆俊, 滕军, 王骥鹏. 基于相空间重构与模糊神经网络耦合的海温预测模型[J]. 热带海洋学报, 2008, 27(4):73-76. DONG Z J, TENG J, WANG J P. Application of phase space reconstruction and ANFIS model in SST forecasting[J]. Journal of Tropical Oceanography, 2008, 27(4):73-76.
[31] 张韧, 蒋国荣. 赤道东太平洋海温及El Nino/La Nina的反演及预测[J]. 解放军理工大学学报, 2000, 1(5):7-12. ZHANG R, JIANG G R. Retrieve and prediction of equatorial east Pacific SST and El Nino/La Nina[J]. Journal of PLA University of Science and Technology, 2000, 1(5):7-12.
[32] 陈璇, 游小宝, 周广庆, 等. 基于权重调整的BP神经网络在Nino区海温预报中的应用[J]. 海洋预报, 2011, 28(5):61-68. CHEN X, YOU X B, ZHOU G Q, et al. Application of BP neural network based on weight adjustment to the temperature forcast[J]. Marine Forecasts, 2011, 28(5):61-68.
[33] LIU S Y, XU L Q, LI D L. Multi-scale prediction of water temperature using empirical mode decomposition with backpropagation neural networks[J]. Computers & Electrical Engineering, 2016, 49:1-8.
[34] 张韧. 非线性BP网络映射与赤道东太平洋海温预测[J]. 海洋通报, 2000, 19(4):1-7. ZHANG R. Non-linear BP neural network mapping and prediction of equatorial east-Pacific SST[J]. Marine Science Bulletin, 2000, 19(4):1-7.
[35] 严明良. 数值产品释用方法在预报业务系统中的应用[C]//推进气象科技创新加快气象事业发展——中国气象学2004年年会论文集. 北京:气象出版社, 2004. YAN M L, The Application of Numerical Product Interpretation Methods in Forecasting Operational Systems. Promoting Innovation in Meteorological Technology and Accelerating the Development of Meteorological Undertakings-Collection of Discussion Papers at the 2004 Annual Conference of China Meteorology. Beijing:China Meteorological Association, 2004
[36] 韩玉康, 余丹丹, 申晓莹, 等. HYCOM模式SST的预报误差订正[J]. 海洋预报, 2018, 35(3):76-80. HAN Y K, YU D D, SHEN X Y, et al. Study on the correction of SST prediction of HYCOM[J]. Marine Forecasts, 2018, 35(3):76- 80.
[37] 匡晓迪, 王兆毅, 张苗茵, 等. 基于BP神经网络方法的近岸数值海温预报释用技术[J]. 海洋与湖沼, 2016, 47(6):1107-1115. KUANG X D, WANG Z Y, ZHANG M Y, et al. An interpretation scheme of numerical near-shore sea-water temperature forecast based on BPNN[J]. Oceanologia et Limnologia Sinica, 2016, 47(6):1107-1115.
[38] 何恩业, 杨静, 李尚鲁, 等. 基于双隐层ANN模型的叶绿素a浓度智能预报方法[J]. 海洋预报, 2021, 38(1):43-53. HE E Y, YANG J, LI S L, et al. Intelligent prediction method for Chl-a based on the artificial neural network[J]. Marine Forecasts, 2021, 38(1):43-53.
[39] 张建华. 海温预报知识讲座——第三讲海表温度长期数值预报[J]. 海洋预报, 2004, 21(3):81-88. ZHANG J H. Lecture on SST Forecast:Lecture 3 Long term numerical prediction of sea surface temperature[J]. Marine Forecasts, 2004, 21(3):8
[40] 谭桂容, 孙照渤, 赵振国. 我国东部夏季降水型与北半球大气环流和北太平洋海温的关系[J]. 南京气象学院学报, 1998, 21(1):1-7. TAN G R, SUN Z B, ZHAO Z G. Relation of summer East-China rainfall patterns to northern circulations and SST[J]. Journal of Nanjing Institute of Meteorology, 1998, 21(1):1-7.

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】