集成SSA、LSTM和ARMA方法的海面高变化预测

doi:10.3976/j.issn.1002-4026.2025084

山东科学

• •

集成SSA、LSTM和ARMA方法的海面高变化预测

张润达¹，王建全²，张宁²

1交通运输部南海航海保障中心广州海事测绘中心，广东广州510000； 2 山东省地矿局索道智能变形监测重点实验室，山东济南250109

收稿日期:2025-08-10 接受日期:2026-04-01 上线日期:2026-06-09
通信作者: 张润达 E-mail:xpeng16@163.com
作者简介:张润达（1993-），男，工程师，从事海洋测绘工作。
基金资助:
中国科学院海岸带环境过程与生态修复重点实验室开放基金（2018KFJJ04）

Integrated SSA–LSTM–ARMA framework for sea-surface-height prediction

ZHANG Rundai¹, WANG Jianquan², ZHANG Ning²

1.Guangzhou Maritime Surveying and Mapping Center, Nanhai Navigation Safety Center, Ministry of Transport of the People’s Republic of China,Guangzhou 510000, China; 2. Key Laboratory of Intelligent Deformation Monitoring, Shandong Provincial Bureau of Geology and Mineral Resources, Jinan 250109, China

Received:2025-08-10 Accepted:2026-04-01 Online:2026-06-09
Contact: ZHANG Runda E-mail:xpeng16@163.com

摘要/Abstract

摘要： 随着全球气候变化对海平面造成的影响日益显著，准确预测海面高变化对海洋环境监测、灾害预警以及沿海资源管理具有重要意义。本文基于2004—2023年间来自28个环日本沿岸验潮站的海面高观测数据，提出集成奇异谱分析（SSA）分解、长短期记忆网络（LSTM）模型和自回归移动平均（ARMA）模型的海面高预测方法，对环日本沿海的海面高进行预测，并与集成SSA和ARMA方法的预测结果对比。研究结果表明，SSA+LSTM+ARMA模型在2019—2023年的海面高预测中表现优于传统的SSA+ARMA方法，SSA+LSTM+ARMA模型的预测海面高年均平均绝对误差（MAE）相较于SSA+ARMA减少了约6%~8%。进一步的预测表明，2024—2025年海面高变化仍然遵循历史周期性波动，并未出现显著的异常波动，整体趋势平稳。该研究表明，SSA+LSTM+ARMA模型在捕捉海面高变化的非线性特征和长期趋势方面具有较强的优势，可为未来的海洋管理和灾害预警提供重要参考。

关键词: 海面高, 奇异谱分析, 长短期记忆神经网络, 自回归移动平均, 验潮站, 日本海, 预测方法

Abstract: With the increasing influence of global climate change on sea levels, accurate prediction of sea-surface-height variations is essential for marine environmental monitoring, disaster early warning, and coastal resource management. Using sea-surface-height observations from 28 tide gauge stations along the coast of Japan from 2004 to 2023, this study proposes a prediction framework that integrates Singular Spectrum Analysis (SSA), a Long Short-term Memory (LSTM) model, and an Autoregressive Moving Average (ARMA) model. The proposed method is applied to predict sea surface height along the Japanese coast and is compared with an SSA+ARMA hybrid approach. The results indicate that the SSA+LSTM+ARMA model outperforms the conventional SSA+ARMA approach in predicting sea surface height for 2019–2023. Specifically, the annual mean absolute error (MAE) is reduced by approximately 6%–8% relative to SSA+ARMA. Projections for 2024–2025 suggest that sea-surface-height variations will continue to exhibit historical periodicity without significant anomalies, with an overall stable trend. These findings demonstrate that the SSA+LSTM+ARMA model more effectively captures nonlinear dynamics and long-term trends in sea surface height, providing a useful reference for marine management and disaster early warning.

Key words: sea surface height, singular spectrum analysis, long short-term memory neural network, autoregressive moving average, tide gauge, coast of Japan, prediction method

中图分类号:

P732

张润达, 王建全, 张宁. 集成SSA、LSTM和ARMA方法的海面高变化预测[J]. 山东科学, 0, (): 1-.

ZHANG Rundai, WANG Jianquan, ZHANG Ning. Integrated SSA–LSTM–ARMA framework for sea-surface-height prediction[J]. Shandong Science, 0, (): 1-.

参考文献

[1] 周秀一, 王井利, 李如仁, 等. 基于BDS卫星信噪比反演海平面高度研究[J]. 地理空间信息, 2023, 21(7): 44-48.

[2] 徐天河, 穆大鹏, 闫昊明, 等. 近20年海平面变化成因研究进展及挑战[J]. 测绘学报, 2022, 51(7): 1294-1305.

[3] SONG T, JIANG J Y, LI W, et al. A deep learning method with merged LSTM neural networks for SSHA prediction[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2020, 13: 2853-2860. DOI:10.1109/jstars.2020.2998461.

[4] SONG T, HAN N S, ZHU Y H, et al. Application of deep learning technique to the sea surface height prediction in the South China Sea[J]. Acta Oceanologica Sinica, 2021, 40(7): 68-76. DOI:10.1007/s13131-021-1735-0.

[5] LI Y, LI Y, XU D Y, et al. Prediction of sea surface height based on recurrent neural network[J]. International Journal of Adaptive and Innovative Systems, 2021, 3(1): 43. DOI:10.1504/ijais.2021.117878.

[6].许武成,王文,马劲松等.热带太平洋海平面年际和年代际变化研究进展[J].河海大学学报(自然科学版),2017,45(3):225 - 231.

[7] DEEPA J S, GNANASEELAN C. The decadal sea level variability observed in the Indian Ocean tide gauge records and its association with global climate modes[J]. Global and Planetary Change, 2021, 198: 103427. DOI:10.1016/j.gloplacha.2021.103427.

[8] ZANCHETTIN D, BRUNI S, RAICICH F, et al. Sea-level rise in Venice: Historic and future trends (review article)[J]. Natural Hazards and Earth System Sciences, 2021, 21(8): 2643-2678. DOI:10.5194/nhess-21-2643-2021.

[9] ZANGANEH M, CHAJI A. A new aspect of the ApEn application to improve the PSO-ANFIS model to forecast Caspian Sea levels[J]. Regional Studies in Marine Science, 2024, 69: 103347. DOI:10.1016/j.rsma.2023.103347.

[10].LI X, ZHOU S J, WANG F W. A CNN-BiGRU sea level height prediction model combined with Bayesian optimization algorithm[J]. Ocean Engineering, 2025, 315: 119849. DOI:10.1016/j.oceaneng.2024.119849.

[11] AYINDE A S, YU H M, WU K J. Review of machine learning methods for sea level change modeling and prediction[J]. Science of the Total Environment, 2024, 954: 176410. DOI:10.1016/j.scitotenv.2024.176410.

[12].IMANI M, CHEN Y C, YOU R J, et al. Spatiotemporal prediction of satellite altimetry sea level anomalies in the tropical Pacific Ocean[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14(7): 1126-1130. DOI:10.1109/LGRS.2017.2699668.

[13] 牛余朋, 郭金运, 袁佳佳, 等. 集成奇异谱分析和自回归滑动平均预测日本近海海平面变化[J]. 地球物理学报, 2020, 63(9): 3263-3274.

[14] TUMSE S, ALCANSOY U. Statistical and deep learning approaches in estimating present and future global mean sea level rise[J]. Natural Hazards, 2025, 121(9): 10377-10404. DOI:10.1007/s11069-025-07203-5.

[15] GRAVES A. Long short-term memory[M]//Supervised Sequence Labelling with Recurrent Neural Networks. Berlin, Heidelberg: Springer, 2012: 37-45. DOI:10.1007/978-3-642-24797-2_4.

[16] ZARZYCKI K, ŁAWRYŃCZUK M. LSTM and GRU neural networks as models of dynamical processes used in predictive control: A comparison of models developed for two chemical reactors[J]. Sensors, 2021, 21(16): 5625. DOI:10.3390/s21165625.

[17] 李丰翔, 林琳. 基于奇异谱分析的地基GPS可降水汽含量变化规律分析[J]. 地理空间信息, 2024, 22(1): 69-73.

集成SSA、LSTM和ARMA方法的海面高变化预测

Integrated SSA–LSTM–ARMA framework for sea-surface-height prediction

PDF (PC)

赞

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 2

Metrics

本文评价

推荐阅读 0

[1]	史朕嘉, 郝增周, 厉运周, 叶枫, 黄海清, 潘德炉. 南海中尺度涡数据融合研究[J]. 山东科学, 2025, 38(3): 99-108.
[2]	周磊. 基于粗糙集和支持向量机的股指期货预测模型研究[J]. J4, 2010, 23(5): 66-70.