基于机载激光测深信息的近岸海域悬沙浓度高精度反演

Retrieval of suspended sediment concentration (SSC) in estuarine and coastal waters by using airborne LiDAR bathymetry (ALB) has various advantages, such as having high spatial and temporal resolution and a large area. However, this method has low accuracy and poor adaptability due to the complex environment and different performances of single-/dual-wavelength ALB systems. This project aims to achieve the highly accurate and simple SSC retrieval by using the waveform and 3D point-cloud data of single- or dual-wavelength ALB. First, water-land classification based on the features in the time and frequency domains of multichannel green laser waveforms and the soft classification algorithm is studied and proposed. A high-accuracy water-land classification method that is suitable for single-/dual-wavelength ALB is then realized. Second, SSC retrieval based on the improved waveform decomposition is studied and proposed by using the waveform data to achieve high-accuracy retrieval. SSC retrieval based on the height biases of green bottom points is studied and proposed by using the 3D point-cloud data, and simple SSC retrieval that is suitable for single-/dual-wavelength ALB is then realized. The accuracy and adaptability of SSC retrieval based on ALB are improved. This project provides theoretical and methodological support in solving problems on SSC acquisition in estuarine and coastal waters.

机载激光测深(Airborne LiDAR Bathymetry, ALB)反演近岸海域悬沙浓度(Suspended Sediment Concentration, SSC)具有高时空分辨率和大面积优点，但受复杂环境、单频和双频ALB性能差异影响，SSC反演精度较低、普适性差。为此，本项目开展普适于单频和双频ALB的近岸海域SSC高精度和简便反演方法研究。首先，研究并提出基于绿激光多通道波形时域、频域特征综合选取和软分类算法的ALB水陆回波分类方法，实现普适于单频和双频ALB的水陆回波高精度分类；然后借助水体回波波形数据，研究并提出基于改进的波形分解法的SSC反演方法，实现SSC高精度反演；借助水体回波三维点云数据，研究并提出基于水底点高度偏差的SSC反演方法，实现普适于单频和双频ALB的SSC简便反演。研究将提高基于ALB的SSC反演精度及普适性，为解决SSC获取难题提供理论和方法支持。

海水悬沙浓度（Suspended Sediment Concentration, SSC）是河口海岸带侵蚀和沉积过程分析预测、生态系统保护以及海岸工程建设等所需的重要海洋环境参数。传统现场测量和卫星遥感反演方法难以满足SSC高精度、高空间分辨率获取需求。为此，本项目从机载激光测深（Airborne LiDAR bathymetry, ALB）全波形和测量偏差两方面入手，开展了基于机载激光测深信息的近岸海域SSC高精度和简便反演研究。.主要研究内容和取得重要结果如下：.（1）开展了ALB全波形去噪和海陆波形分类研究。提出了基于离散傅里叶变换和低通滤波的绿激光全波形背景噪声去噪法和随机噪声自适应小波阈值去噪法，解决了绿激光测深全波形去噪“过平滑”和“欠平滑”问题，提高了全波形信号保真度。提出双频ALB红外激光全波形双层聚类法和单频ALB绿激光多通道投票的海陆波形分类方法，实验区域内获得了优于99%的总体精度，实现了单/双频ALB高精度海陆波形分类。.（2）开展了基于ALB绿激光波形分解的SSC反演研究。提出了附加波形参数边界约束的改进的波形分解法，将海水后向散射回波振幅提取精度提高了约1.4倍。提出了基于改进的波形分解法的SSC反演方法，实验区域内反演SSC均方误差为2.28 mg/L，实现了适用于单/双频ALB的SSC高精度、高空间分辨率反演。.（3）研究并提出了基于单/双频ALB海底点高度测量偏差的SSC反演方法，为单/双频ALB反演SSC提供了一种简便方法。.项目研究取得了如下成果：.（1）形成了一套完整的、普适于单频和双频ALB的近岸海域SSC高精度和简便反演理论与方法体系，包括ALB激光全波形去噪、单/双频ALB海陆波形分类、绿激光全波形分解、基于ALB绿激光波形分解的SSC反演以及基于ALB绿激光海底点高度偏差的SSC反演。.（2）在本项目资助下，截止2022年底，项目负责人以第一作者或通讯作者在IEEE GRSL、IEEE JSTARS等国内外核心刊物上共发表第一标注本项目的期刊论文8篇，其中SCI论文7篇，核心1篇。.（3）研制软件5套，登记软件著作权3项。.（4）培养硕士研究生8名，其中2名获得研究生国家奖学金。.（5）形成了完整的课题研究报告。