标题

A Maximum Bathymetric Depth Model to Simulate Satellite Photon-Counting Lidar Performance

作者

Wenhao Zhang, Nan Xu, Yue Ma, Bisheng Yang, Zhiyu Zhang, Xiao Hua Wang, Song Li

来源

ISPRS Journal of Photogrammetry and Remote Sensing

摘要

随着光子计数传感器的发展，星载光子计数激光雷达在测绘水下地形方面显示出诸多优势。虽然基于空间的激光雷达通常是一个剖面系统，但使用测深激光雷达实现的深度穿透和垂直精度优于图像（仅提供相对深度）。因此，许多卫星测深产品使用主动/被动星载数据来实现空间覆盖以及绝对深度测量。与船载和机载测深系统相比，卫星测深数据可以覆盖全球，特别是覆盖一些机载/船载系统难以或非常昂贵的偏远地区。本文提出了一种理论模型来定量分析卫星光子计数激光雷达的最大测深深度，该模型考虑了系统参数、环境影响和信噪比（SNR）的约束。 ICESat-2（冰、云和陆地高程卫星 2）的系统参数以及 MODIS（中分辨率成像光谱仪）和 NCEP（国家环境预测中心）数据集用于为模型，并使用 ICESat-2 实际测深数据来验证所提出的模型估计的最大深度。在六个具有不同水质、太阳角度和风速的不同地点，理论最大测深深度与实际获得的 ICESat-2 激光雷达深度非常吻合，MAE（平均绝对误差）为 0.50 m，RMSE（均方根）平方误差）为 0.60 m。所有研究区域的误差均在相应最大深度的 14% 以内。此外，对光子计数激光雷达的非线性响应、独立参数对理论模型的影响以及实验过程中的误差进行了定量分析。将来，当应用系统参数和环境参数时，该理论模型可用于评估测深调查的最大测深深度。此外，它还可用于在光子计数激光雷达的早期设计过程中优化其硬件参数。

Abstract

With the development of photon-counting sensors, spaceborne photon-counting lidars have shown many advantages in mapping underwater topography. Although a space based lidar is normally a profiling system, the depth penetration and the vertical accuracy achieved with a bathymetric lidar is superior to imagery (that only provides relative depths). Therefore, many satellite derived bathymetry products use both active/passive spaceborne data to achieve spatial coverage as well as absolute depth measurements. Compared with shipborne and airborne bathymetric systems, satellite derived bathymetry data can have a global coverage, especially covering some remote areas where airborne/shipborne systems are hard or very expensive to reach. In this paper, a theoretical model is proposed to quantitatively analyze the maximum bathymetric depth of a satellite photon-counting lidar, which considers the system parameters, environmental effects, and the constraint of the SNR (signal to noise ratio). System parameters of ICESat-2 (Ice, Cloud, and land Elevation Satellite-2) as well as the MODIS (Moderate-resolution Imaging Spectroradiometer) and NCEP (National Centers for Environmental Prediction) datasets are used to provide systematic and environmental inputs to the model, and the ICESat-2 actual bathymetric data are used to verify the maximum depth estimated by the proposed model. In six different sites with different water qualities, solar angles and wind speeds, the theoretical maximum bathymetric depths agree well with the actually achieved ICESat-2 lidar depths, in terms of MAE (mean absolute error) of 0.50 m and the RMSE (root mean square error) of 0.60 m. The errors in all study areas are within 14% of the corresponding maximum depths. Additionally, the non-linear response of a photon-counting lidar, independent parameter impacts on the theoretical model, and the errors in the experimental process were quantitatively analyzed. In the future, this theoretical model can be used to evaluate the maximum bathymetric depth for a bathymetry investigation when applying system parameters and environmental parameters. In addition, it can be used to optimize the hardware parameters of a photon-counting lidar in its early design process.

图 用于验证基于 ICESat-2 ATLAS光子计数激光雷达数据的最大测深模型的六个研究区域的位置。(a) The site near Jamaica Bay at Acklins Island; (b) the site near Yongle Atoll; (c) the site in Exmouth Gulf; (d) the site near Victor Island; (e) the site to the south of St. Thomas Island; and (f) the site near Wayangani Beach.