基于XGBoost和SHAP的海滩波浪爬高预测研究-海洋预报

基于XGBoost和SHAP的海滩波浪爬高预测研究

单位：1. 珠海市规划设计研究院, 广东珠海 519000;
2. 广东省滨海地区防灾减灾工程技术研究中心, 广东珠海 519000;
3. 南京水利科学研究院, 江苏南京 210029

分类号：P731.33

出版年·卷·期（页码）：2025·42·第二期（1-8）

摘要：

海滩波浪爬高预测是海岸侵蚀防护和防灾减灾的关键技术支撑。针对现有经验公式在精确度、泛化性等方面的不足，将极限梯度提升模型XGBoost引入到波浪爬高预测中，利用1 400多个来自实验室和现场观测的海滩波浪爬高数据，通过贝叶斯优化进行超参数调整，建立基于XGBoost的海滩波浪爬高预测模型。此外，还将可解释机器学习框架SHAP与XGBoost模型结合，以挖掘波浪爬高预测结果的关键特征。评估结果表明：XGBoost模型的决定系数为0.957，均方根误差为0.384 m，显著优于其他经验公式，整体预测可靠稳定；SHAP分析也表明XGBoost模型的预测趋势符合真实走向，且Iribarren数在海滩波浪爬高预测中起着关键作用。

Beach wave run-up prediction is a key technical support for coastal erosion protection, disaster prevention and mitigation. In view of the shortcomings of the existing empirical formulas in terms of accuracy and generalization, the XGBoost model is introduced into wave run-up prediction, and more than 1 400 laboratory and field observations of beach wave run-up are used as a dataset, and hyperparameter tuning is carried out by using Bayesian optimization, which in turn establishes an XGBoost-based wave run-up prediction model. The XGBoost model is used to predict beach wave height, and SHAP, an interpretable machine learning framework, is combined with the XGBoost model to explore the key features of the wave height prediction results. The evaluation results show that the R-squared of the XGBoost model is 0.957, and the root-mean-square error is 0.384 m, which is significantly better than other empirical formulas, and the overall prediction is reliable and stable, meanwhile SHAP shows that the XGBoost model predicted trend is in line with the true value direction and Iribarren number plays a key role in beach wave run-up prediction.

参考文献：

[1] 尹航. 视频图像海滩动力地貌监测与信息提取方法研究[D]. 厦门:自然资源部第三海洋研究所, 2022. YIN H. Beach dynamic geomorphology monitoring and information extraction based on video imagery[D]. Xiamen:Third Institute of Oceanography, MNR, 2022.
[2] 王广生, 童林龙, 罗梦岩, 等. 贝宁海滩上波浪传播演变特性研究[J]. 河海大学学报(自然科学版), 2023, 51(6):123-129. WANG G S, TONG L L, LUO M Y, et al. Study on wave propagation and evolution characteristics over a beach in Benin[J]. Journal of Hohai University (Natural Sciences), 2023, 51(6):123- 129
[3] 邱星, 董玉祥. 海岸沙丘对风暴潮响应研究进展与展望[J]. 地球科学进展, 2022, 37(8):811-821. QIU X, DONG Y X. Research progress and prospect of the response of coastal dunes to storm surge[J]. Advances in Earth Science, 2022, 37(8):811-821.
[4] STOCKDON H F, HOLMAN R A, HOWD P A, et al. Empirical parameterization of setup, swash, and runup[J]. Coastal Engineering, 2006, 53(7):573-588.
[5] DA SILVA P G, COCO G, GARNIER R, et al. On the prediction of runup, setup and swash on beaches[J]. Earth-Science Reviews, 2020, 204:103148.
[6] LERMA A N, PEDREROS R, ROBINET A, et al. Simulating wave setup and runup during storm conditions on a complex barred beach[J]. Coastal Engineering, 2017, 123:29-41.
[7] ABOLFATHI S, YEGANEH-BAKHTIARY A, HAMZE-ZIABARI S M, et al. Wave runup prediction using M5' model tree algorithm[J]. Ocean Engineering, 2016, 112:76-81.
[8] TARWIDI D, PUDJAPRASETYA S R, ADYTIA D, et al. An optimized XGBoost-based machine learning method for predicting wave run-up on a sloping beach[J]. MethodsX, 2023, 10:102119.
[9] BEUZEN T, GOLDSTEIN E B, SPLINTER K D. Ensemble models from machine learning:an example of wave runup and coastal dune erosion[J]. Natural Hazards and Earth System Sciences, 2019, 19(10):2295-2309.
[10] CHEN T Q, GUESTRIN C. XGBoost:A scalable tree boosting system[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco:Association for Computing Machinery, 2016:785-794.
[11] LUNDBERG S M, LEE S I. A unified approach to interpreting model predictions[C]//Proceedings of the 31st International Conference on Neural Information Processing Systems. Long Beach:Curran Associates Inc., 2017:4768-4777.
[12] HUNT JR I A. Design of seawalls and breakwaters[J]. Journal of the Waterways and Harbors Division, 1959, 85(3):123-152.
[13] HOLMAN R A. Extreme value statistics for wave run-up on a natural beach[J]. Coastal Engineering, 1986, 9(6):527-544.
[14] VOUSDOUKAS M I, WZIATEK D, ALMEIDA L P. Coastal vulnerability assessment based on video wave run-up observations at a mesotidal, steep-sloped beach[J]. Ocean Dynamics, 2012, 62(1):123-137.
[15] ATKINSON A L, POWER H E, MOURA T, et al. Assessment of runup predictions by empirical models on non-truncated beaches on the south-east Australian coast[J]. Coastal Engineering, 2017, 119:15-31.
[16] DIWEDAR A I. Investigating the effect of wave parameters on wave runup[J]. Alexandria Engineering Journal, 2016, 55(1):627- 633.
[17] WU D H, LIU H J. Effects of the bed roughness and beach slope on the non-breaking solitary wave runup height[J]. Coastal Engineering, 2022, 174:104122.
[18] TENG M H, FENG K L, LIAO T I. Experimental study on long wave run-up on plane beaches[C]//The Tenth International Offshore and Polar Engineering Conference. Seattle:OnePetro, 2000.
[19] AKIBA T, SANO S, YANASE T, et al. Optuna:A next-generation hyperparameter optimization framework[C]//Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. Anchorage:Association for Computing Machinery, 2019:2623-2631.
[20] POWER H E, GHARABAGHI B, BONAKDARI H, et al. Prediction of wave runup on beaches using Gene-Expression Programming and empirical relationships[J]. Coastal Engineering, 2019, 144:47-61

服务与反馈：

【文章下载】【发表评论】【查看评论】【加入收藏】