中国海洋大学学报（自然科学版）2024，Vol.54Issue(2)：121-133,13.DOI:10.16441/j.cnki.hdxb.20220489
基于多层感知器和SAR参数的海浪有效波高反演方法
A retrieval Approach for Significant Wave Height Based on Multi-Layer Perceptron and Parametersacquired by Sentinel-1 SAR
摘要
Abstract
Based on the analysis of the correlation between several parameters of radar sea surface im-age and the significant wave height,the paper proposes a method which is based on a multi-layer percep-tron(MLP)model to retrieve significant wave height by applying the backscatter coefficient,normalized variance of image intensity,azimuth cut-off wavelength,peak wavelength,peak wave direction,image skewness and kurtosis parameters from Sentinel 1A satellite Level-2 SAR data.The inversion accuracy of the total significant wave height(SWH),the significant wave height of wind wave(SWHWW)and the significant wave height of swell(SWHS)by the MLP model with different combinations of parameters is discussed in detail.A correction function is used to reduce the systematic bias in the model caused by the low availability of data in extreme sea conditions,thus enhancing the applicability of the model.When compared with SWH data provided by European Centre for Medium-Range Weather Forecasts(ECM-WF)and Jason-3 satellite,the root mean square errors of the method are 0.471 and 0.535 m,with cor-relation coefficients of 0.923 and 0.922,respectively.When compared with SWHWW and SWHS data provided by ECMWF,the root mean square errors of the method are 0.534 and 0.512 m,with correla-tion coefficients of 0.898 and 0.815,respectively.关键词
哨兵1A/SAR数据/多层感知器/海浪有效波高/校正函数Key words
Sentinel-1A/SAR data/multi-layer perceptron/significant wave height/correction function分类
海洋科学引用本文复制引用
高亚飞,王运华,张彦敏,姜文正..基于多层感知器和SAR参数的海浪有效波高反演方法[J].中国海洋大学学报（自然科学版）,2024,54(2):121-133,13.基金项目
国家自然科学基金项目(41976167,52101393) （41976167,52101393）
山东省自然科学基金项目(ZR2021MD023,ZR2021QD001)资助Supported by the National Natural Science Foundation of China(41976167,52101393) （ZR2021MD023,ZR2021QD001）
the Natural Science Foundation of Shandong Province(ZR2021MD023,ZR2021QD001) （ZR2021MD023,ZR2021QD001）
