黄妙芬,王江颖,邢旭峰,王忠林,周运

• 1:广东海洋大学数学与计算机学院

摘要(Abstract)：

【目的】建立一种基于美国陆地卫星Landsat遥感数据和长短期记忆(Long Short-Term Memory,LSTM)网络的海洋石油污染含量预测模型。【方法】利用1984—2020年在大连新港海域过境的Landsat卫星所采集的可见光-近红外波段(0.45～0.90μm)光谱数据,基于LSTM网络,分别建立空间分辨率为30 m、时间分辨率为8 d的4波段遥感反射比Rrs预测模型,并对预测模型所涉及的神经网络层数、隐藏神经元节点数和回溯时间步长等超参数进行优化;在4波段R_(rs)预测值的基础上,结合基于水体石油污染归一化遥感反射比指数(normalized difference petroleum remote sensing reflectance index,NDPRI)的石油含量遥感反演模型,对海域石油污染含量C_o值进行预测。【结果】对于蓝光、绿光、红光和近红外4个波段,神经网络层数依次取3、3、4和3层,隐藏神经元节点取64、96、64和96个,回溯时间步长皆取6 d为最优;根据2021年1—5月现场的C_o测量值,对LSTM网络预测值进行精度分析,平均相对误差为9.17%。【结论】基于LSTM网络建立的C_o预测模型具有较好的精度,所预测的结果可弥补在有云情况C_o数据缺失的问题,也可为相关C_o未来动态演变研究提供一种新技术手段。

关键词(KeyWords)： 陆地卫星Landsat;长短期记忆网络(LSTM);遥感反射比;石油污染含量;预测模型

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金项目（41771384）;; 国家重点研发计划重点专项资助项目（2016YFC1401203）

作者(Author): 黄妙芬,王江颖,邢旭峰,王忠林,周运

参考文献(References)：

• [1] OTREMBA Z, KRóL T. Modeling of the crude oil suspension impact on inherent optical parameters of coastal seawater[J]. Polish Journal of Environmental Studies, 2002, 11(4):407-411.
• [2] LU Y C, SHI J, WEN Y S, et al. Optical interpretation of oil emulsions in the ocean-Part I:Laboratory measurements and proof-of-concept with AVIRIS observations[J]. Remote Sensing of Environment, 2019, 230:111183.
• [3]黄妙芬,王忠林,刘远,等.辽东湾海域有色可溶性有机物荧光特征分析[J].大连海洋大学学报, 2018, 33(2):137-144.
• [4]黄妙芬,庄炀,刘杨,等.大连港海域石油类物质比吸收系数特征分析[J].海洋湖沼通报, 2020(3):141-149.
• [5]黄妙芬,宋庆君,简伟军,等.水体石油类信息遥感提取模式研究[J].海洋学报, 2012, 34(5):74-80.
• [6] HUANG M F, XING X F, SONG Q J, et al. A new algorithm of retrieving a petroleum substances absorption coefficient in sea water based on a remote sensing image[J]. Acta Oceanologica Sinica, 2016, 35(11):97-104.
• [7] HUANG M F, SONG Q J, XING X F, et al. A biooptical model of retrieving petroleum concentration in seawater[J].Acta Oceanologica Sinica, 2014, 33(5):81-85.
• [8]黄妙芬,宋庆君,陈利搏,等.基于归一化遥感反射比反演水中石油含量模式研究[J].海洋技术学报, 2015,34(1):1-9.
• [9] KOVALSKYY V, ROY D P. The global availability of Landsat 5 TM and Landsat 7 ETM+land surface observations and implications for global 30 m Landsat data product generation[J]. Remote Sensing of Environment, 2013, 130:280-293.
• [10] CHASTAIN R, HOUSMAN I, GOLDSTEIN J, et al.Empirical cross sensor comparison of Sentinel-2A and2B MSI, Landsat-8 OLI, and Landsat-7 ETM+top of atmosphere spectral characteristics over the conterminous United States[J]. Remote Sensing of Environment, 2019, 221:274-285.
• [11] CAO F, TZORTZIOU M. Capturing dissolved organic carbon dynamics with Landsat-8 and Sentinel-2 in tidally influenced wetland-estuarine systems[J]. Science of the Total Environment, 2021, 777:145910.
• [12] SARI V, REIS CASTRO N M, PEDROLLO O C.Estimate of suspended sediment concentration from monitored data of turbidity and water level using artificial neural networks[J]. Water Resources Management, 2017, 31(15):4909-4923.
• [13] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M,et al. Generative Adversarial Nets[J]. Neural Information Processing System, 2014, 3:2672-2680.
• [14] CHO K, VAN MERRIENBOER B, GULCEHRE C, et al.Learning phrase representations using RNN encoder–decoder for statistical machine translation[C]//Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing(EMNLP).Stroudsburg, PA, USA:Association for Computational Linguistics, 2014:1724-1734.
• [15] ADIB A, JAHANBAKHSHAN H. Stochastic approach to determination of suspended sediment concentration in tidal rivers by artificial neural network and genetic algorithm[J]. Canadian Journal of Civil Engineering,2013, 40(4):299-312.
• [16] DIJKSTRA H A, PETERSIK P, HERNáNDEZ-GARCíA E,et al. The application of machine learning techniques to improve el ni?o prediction skill[J]. Frontiers in Physics,2019, 7:153. DOI:10.3389/fphy.2019.00153
• [17] WANG X F, XU L Y. Unsteady multi-element time series analysis and prediction based on spatial-temporal attention and error forecast fusion[J]. Future Internet,2020, 12(2):34.
• [18]周朝勉,刘明萍,王京威.基于CNN-LSTM的水质预测模型研究[J].水电能源科学, 2021, 39(3):20-23.
• [19]祁柏林,郭昆鹏,杨彬,等.基于GCN-LSTM的空气质量预测[J].计算机系统应用, 2021, 30(3):208-213.
• [20]邢旭峰,王忠林,黄妙芬,等.石油污染水体油浓度及后向散射系数垂向变化特性分析[J].广东海洋大学学报, 2020, 40(4):75-81.
• [21]黄妙芬,刘杨,邢旭峰,等.大连港海域油含量时空变化特征分析[J].大连海洋大学学报, 2020, 35(2):273-279.