基于LSTM构建海岸侵蚀风险预测模型的研究及应用

梁东; 高娜; 英晓明; 周曾; 舒勰俊; 徐婉明; 赵明利

doi:10.12284/hyxb2024059

基于LSTM构建海岸侵蚀风险预测模型的研究及应用

doi: 10.12284/hyxb2024059

梁东^{1, 2,},
高娜¹,
英晓明^{1, 2},
周曾³,
舒勰俊¹,
徐婉明¹,
赵明利^{1, 2, ,}

1.
自然资源部南海发展研究院，广东广州 510300
2.
自然资源部海洋环境探测技术与应用重点实验室，广东广州 510300
3.
河海大学江苏省海岸海洋资源开发与环境安全重点实验室，江苏南京 210098

基金项目: 国家重点研发计划（2022YFC3106203）课题资助；自然资源部海洋环境探测技术与应用重点实验室自主设立课题（MESTA-2022-C005）资助。

详细信息

作者简介:
梁东（1996—），男，云南省红河州人，主要从事海气浪耦合模型与神经网络融合应用的研究。E-mail：liangdong.96@qq.com

通讯作者:
赵明利（1978—），男，山东省肥城市人，正高级工程师，主要从事海洋灾害风险评估技术方法研究。E-mail：150980273@qq.com

中图分类号: P737.1
计量
- 文章访问数: 230
- HTML全文浏览量: 61
- PDF下载量: 32
- 被引次数: 0
出版历程
- 收稿日期: 2024-01-24
- 修回日期: 2024-05-13
- 网络出版日期: 2024-07-15
- 刊出日期: 2024-06-01

Research and application of constructing a coastal erosion risk prediction model based on LSTM

Liang Dong^{1, 2
,},
Gao Na¹,
Ying Xiaoming^{1, 2},
Zhou Zeng³,
Shu Xiejun¹,
Xu Wanming¹,
Zhao Mingli^{1, 2
, ,}

1.
South China Sea Development Research Institute, Ministry of Natural Resources, Guangzhou 510300, China
2.
Key Laboratory of Marine Environmental Survey Technology and Application, Ministry of Natural Resources, Guangzhou 510300, China
3.
Jiangsu Provincial Key Laboratory of Coastal Ocean Resources Development and Environment Security, Hohai University, Nanjing 210098, China

摘要

摘要: 岸线侵蚀预测是海岸动力地貌学研究的热点问题之一。本文基于长短期神经记忆网络LSTM，采用1985−2023年江苏射阳县附近海域收集到的岸线、水深和潮间带宽度数据，结合ERA5数据反演的波浪和潮流数据，构建海岸侵蚀风险预测模型。该预测模型可以准确反映海岸线的加速侵蚀、稳定侵蚀、淤积的非线性变化或线性变化趋势。预测结果表明，在沙源减少条件下，波浪和潮流增强是近20年来射阳海域海岸侵蚀的主要因素。此外，利用预测模型开展了海岸加固、消浪和弱流对海岸的防护效果试验，试验结果表明，海岸加固防护效果最佳，消浪防护效果较弱流防护效果好。预测模型设置运行过程高效，具有较高的应用价值和开发潜力。
- 海岸侵蚀预测 /
- 江苏射阳 /
- LSTM /
- 非线性变化
Abstract: Shoreline erosion prediction is one of the hot issues in coastal dynamic geomorphology research. Based on the long short term memory (LSTM), the data of shoreline, water depth, intertidal zone width , and wave and tidal current for ERA5 inversion clollected from 1985 to 2023 near Sheyang County of Jiangsu Province were used to construct a coastal erosion risk prediction model in this study. The prediction model could accurately predict the nonlinear/linear change trend of accelerated erosion, stable erosion or coastline sedimentation. The results showed that the increasing of wave and tidal currents was the main factor of coastal erosion in Sheyang area in recent 20 years under the condition of sand source reduction. Besides, an ideal experiment of coastal protection activities was conducted by using the prediction model, and the protection effects of coastal reinforcement, wave dissipation and weak current engineering were discussed. The results showed that the protection effect of coastal reinforcement is the best, and wave dissipation is better than weak current. The prediction model is reasonable, and has great application value and development potential.
- coastal erosion prediction /
- Sheyang, Jiangsu /
- LSTM /
- Nonlinear variation

HTML全文

图 1 研究区域水深变化（a）与模型网格图（b）

Fig. 1 Water depth change in this study area (a) and model mesh diagram (b)

下载: 全尺寸图片幻灯片

图 2 研究区域年平均有效波高（a），年平均波浪周期（b）和年平均风速（c）

Fig. 2 Annual average effective wave height (a), annual average wave period (b) and annual average wind speed (c) in this study area

下载: 全尺寸图片幻灯片

图 3 LSTM 神经网络模型结构示意图

Fig. 3 Schematic diagram of LSTM neural network model structure

下载: 全尺寸图片幻灯片

图 4 海岸侵蚀风险预测模型

Fig. 4 Coastal erosion risk prediction model

下载: 全尺寸图片幻灯片

图 5 研究区域岸线回归变化率（a）与实际岸线变化（b）

Fig. 5 Regression rate of shoreline in this study area (a) and actual shoreline change (b)

下载: 全尺寸图片幻灯片

图 6 研究区域波浪（a）和流速（b）变化

Fig. 6 Wave (a) and tidal velocity (b) variations in this study area

下载: 全尺寸图片幻灯片

图 7 研究区域海岸坚固性（a）和潮间带宽度（b）变化

Fig. 7 Variations of coastal firmness (a) and intertidal width (b) in this study area

下载: 全尺寸图片幻灯片

图 8 海岸侵蚀风险预测模型验证

Fig. 8 Validation diagram of coastal erosion risk prediction model

下载: 全尺寸图片幻灯片

图 9 海岸侵蚀风险预测模型预测

Fig. 9 Prediction diagram of coastal erosion risk prediction model

下载: 全尺寸图片幻灯片

图 10 海岸变化趋势

Fig. 10 Diagram of coastal change trends

下载: 全尺寸图片幻灯片

图 11 海岸防护活动侵蚀风险预测

Fig. 11 Erosion risk prediction diagram of coastal protection activities

下载: 全尺寸图片幻灯片

图 12 双导堤侵蚀风险预测

Fig. 12 Predicted erosion risk diagram of double-dyke

下载: 全尺寸图片幻灯片

表 1 海岸侵蚀风险预测模型参数表

Tab. 1 Parameter table of coastal erosion risk prediction model

模型名称	参数名称	参数值
LSTM	Time steps	8
	Layers	2
	Input dim	5
	Return sequences	{True, False}
	Neurous	{80,88}
	Dropout	{ 0,0.25}
	Batch size	{13,8}
	Epochs	{8,10}

下载: 导出CSV

表 2 岸线易损性等级评估表

Tab. 2 Assessment table of shoreline vulnerability level

岸线向陆100 m 平均高程	岸线易变性	岸线组成成分	岸线易损性
< 0	2	细砂	5
0 < X < 2	1.5	中砂/粉砂	4
> 2	1	粗砂/黏土	3
#	#	砾质	2
#	#	混凝土/岩石	1
注：#为空值。

下载: 导出CSV

表 3 预测结果统计表

Tab. 3 Statistical table of projected results

	淤积	稳定	一般侵蚀	严重侵蚀	总计
2023预测断面数量	119	636	111	292	1158
2023真实断面数量	124	728	200	311	1363
准确率/%	95.9	87.4	55.5	93.9	84.7
判断标准/m	>25	−25 < X < 25	−100 < X < −25	X < −100

下载: 导出CSV

参考文献(26)

[1]	凡姚申, 窦身堂, 裴洪杨, 等. 大河三角洲侵蚀灾害与应对策略研究进展[J]. 自然灾害学报, 2022, 31(2): 12−25. Fan Yaoshen, Dou Shentang, Pei Hongyang, et al. Research progress of erosion hazards and coping strategies in mega deltas[J]. Journal of Natural Disasters, 2022, 31(2): 12−25.
[2]	李平, 丰爱平, 孙惠凤, 等. 海岸侵蚀灾害调查和评价研究进展与展望[J]. 自然灾害学报, 2021, 30(4): 55−63. Li Ping, Feng Aiping, Sun Huifeng, et al. Research progress and prospect of coastal erosion investigation and evaluation[J]. Journal of Natural Disasters, 2021, 30(4): 55−63.
[3]	文世勇, 王紫竹, 王涛, 等. 基于遥感技术的海南省海岸侵蚀现状与趋势评估[J]. 灾害学, 2020, 35(1): 138−143. doi: 10.3969/j.issn.1000-811X.2020.01.025 Wen Shiyong, Wang Zizhu, Wang Tao, et al. Status and trend assessment of coastal erosion at Hainan Province based on remote sensing technology[J]. Journal of Catastrophology, 2020, 35(1): 138−143. doi: 10.3969/j.issn.1000-811X.2020.01.025
[4]	HsuJRC, SilvesterR, Xia Yimin. Static equilibrium bays: new relationships[J]. Journal of Waterway, Port, Coastal, and Ocean Engineering, 1989, 115(3): 285−298. doi: 10.1061/(ASCE)0733-950X(1989)115:3(285)
[5]	Hsu T W, Jan C D, Wen C C. Modified McCormicks model for equilibrium shorelines behind a detached breakwater[J]. Ocean Engineering, 2003, 30(15): 1887−1897. doi: 10.1016/S0029-8018(03)00042-8
[6]	González M, Medina R. On the applicationof static equilibrium bay formulations to natural and man-made beaches[J]. Coastal Engineering, 2001, 43(3/4): 209−225.
[7]	Kakisina T J, Anggoro S, Hartoko A, et al. NEMOS (Nearshore Modelling of Shoreline Change) model for abrasion mitigation at the northern coast of Ambon Bay[J]. Aquatic Procedia, 2016, 7: 242−246. doi: 10.1016/j.aqpro.2016.07.034
[8]	Young R S, Pilkey O H, Bush D M, et al. A discussion of the generalized model for simulating shoreline change (GENESIS)[J]. Journal of Coastal Research, 1995, 11(3): 875−886.
[9]	Castelle B, Reniers A, MacMahan J. Bathymetric control of surf zone retention on a rip-channelled beach[J]. Ocean Dynamics, 2014, 64(8): 1221−1231.
[10]	Roelvink D, Reniers A, van Dongeren A, et al. Modelling storm impacts on beaches, dunes and barrier islands[J]. Coastal Engineering, 2009, 56(11/12): 1133−1152.
[11]	朱萝云, 刘婷婷, 凡仁福, 等. 1986-2019年粤东企望湾砂质海岸线演变过程与驱动机制研究[J]. 海洋学报, 2022, 44(7): 82−94. doi: 10.12284/j.issn.0253-4193.2022.7.hyxb202207008 Zhu Luoyun, Liu Tingting, Fan Renfu, et al. Study on the evolution process and driving mechanism of the sandy shoreline of the Qiwang Bay ineastern Guangdong from 1986 to 2019[J]. HaiyangXuebao, 2022, 44(7): 82−94. doi: 10.12284/j.issn.0253-4193.2022.7.hyxb202207008
[12]	朱磊, 孙家文, 王宏, 等. 基于XBeach模型的离岸堤群防护效果评价指标[J]. 海洋环境科学, 2020, 39(5): 684−693. doi: 10.12111/j.mes.20190157 Zhu Lei, Sun Jiawen, Wang Hong, et al. Evaluation index of protection effect of breakwaters based on XBeach model[J]. Marine Environmental Science, 2020, 39(5): 684−693. doi: 10.12111/j.mes.20190157
[13]	于航, 张建鹏, 陈根发, 等. 基于改进XBeach的波浪作用下岸滩冲淤演变特性的数值研究[J]. 水动力学研究与进展, 2023, 38(3): 495−503. Yu Hang, Zhang Jianpeng, Chen Genfa, et al. Numerical study of beach profile evolution under waves based on Improved XBeach[J]. Chinese Journal of Hydrodynamics, 2023, 38(3): 495−503.
[14]	Zheng Gang, Li Xiaofeng, Zhang Ronghua, et al. Purely satellite data-driven deep learning forecast of complicated tropical instabilitywaves[J]. Science Advances, 2020, 6(29): eaba1482. doi: 10.1126/sciadv.aba1482
[15]	Nieves V, Radin C, Camps-Valls G. Predicting regional coastal sea level changes with machine learning[J]. Scientific Reports, 2021, 11(1): 7650. doi: 10.1038/s41598-021-87460-z
[16]	Gers F A, Schmidhuber J, Cummins F. Learning to forget: continual prediction with LSTM[J]. Neural Computation, 2000, 12(10): 2451−2471.
[17]	Hochreiter S, Schmidhuber J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735−1780. doi: 10.1162/neco.1997.9.8.1735
[18]	刘建伟, 宋志妍. 循环神经网络研究综述[J]. 控制与决策, 2022, 37(11): 2753−2768. Liu Jianwei, Song Zhiyan. Overview of recurrent neural networks[J]. Control and Decision, 2022, 37(11): 2753−2768.
[19]	张忍顺, 陆丽云, 王艳红. 江苏海岸侵蚀过程及其趋势[J]. 地理研究, 2002, 21(4): 469−478. doi: 10.3321/j.issn:1000-0585.2002.04.009 Zhang Renshun, Lu Liyun, Wang Yanhong. The mechanism and trend of coastal erosion of Jiangsu Province in China[J]. Geographical Research, 2002, 21(4): 469−478. doi: 10.3321/j.issn:1000-0585.2002.04.009
[20]	Zhang Changkuan, YangYaozhong, Tao Jianfeng, et al. Suspended sediment fluxes in the radial sand ridge field of SouthYellow Sea[J]. Journal of Coastal Research, 2013, 65(sp1): 624−629.
[21]	张林. 苏北废黄河三角洲海岸冲淤演变及其控制因素[D]. 上海: 华东师范大学, 2015. Zhang Lin. The coastal erosion-deposition evolution and controlling factors of the abandoned Yellow River delta in northern Jiangsu province[D]. Shanghai: East China Normal University, 2015.
[22]	刘小喜, 陈沈良, 蒋超, 等. 苏北废黄河三角洲海岸侵蚀脆弱性评估[J]. 地理学报, 2014, 69(5): 607−618. doi: 10.11821/dlxb201405004 Liu Xiaoxi, Chen Shenliang, Jiang Chao, et al. Vulnerability assessment of coastal erosion along the Abandoned Yellow River Delta of northern Jiangsu, China[J]. ActaGeographicaSinica, 2014, 69(5): 607−618. doi: 10.11821/dlxb201405004
[23]	Shi Hongyuan, Cao Xuefeng, Li Qingjie, et al. Evaluating the accuracy of ERA5 wave reanalysis in the water around China[J]. Journal of Ocean University of China, 2021, 20(1): 1−9. doi: 10.1007/s11802-021-4496-7
[24]	张达恒, 时连强, 龚照辉, 等. 冬季波浪与人工岛联合作用下日月湾海滩冲淤演变特征[J]. 热带海洋学报, 2022, 41(4): 71−81. doi: 10.11978/2021150 Zhang Daheng, Shi Lianqiang, Gong Zhaohui, et al. Evolution characteristics of beach erosion and accretion at the Riyue Bay under the combined impacts of winter waves and artificial island[J]. Journal of Tropical Oceanography, 2022, 41(4): 71−81. doi: 10.11978/2021150
[25]	Bheeroo R A, Chandrasekar N, Kaliraj S, et al. Shoreline change rate and erosion risk assessment along the Trou Aux Biches–Mont Choisy beach on the northwest coast of Mauritius using GIS-DSAS technique[J]. Environmental Earth Sciences, 2016, 75(5): 444.
[26]	张长宽, 黄婷婷, 陶建峰, 等. 江苏海岸潮滩剖面形态与动力泥沙响应关系[J]. 河海大学学报(自然科学版), 2020, 48(3): 245−251. Zhang Changkuan, Huang Tingting, Tao Jianfeng, et al. Response relationship of tidal flat profile and dynamic sediment along Jiangsu Coast[J]. Journal of Hohai University (Natural Sciences), 2020, 48(3): 245−251.