Importance of the parallel polarization radiance for estimating inorganic particle concentrations in turbid waters based on radiative transfer simulations
Importance of the parallel polarization radiance for estimating inorganic particle concentrations in turbid waters based on radiative transfer simulations
复制标题
平行偏振辐射率对于基于辐射传输模拟估算浑浊水中无机颗粒浓度的重要性
DOI：
10.1080/01431161.2020.1727059
复制
发表时间：
2020
影响因子：
3.4
通讯作者：
Pan Delu
中科院分区：
文献类型：
--
作者：
Liu Jia;Hu Bingliang;He Xianqiang;Bai Yan;Tian Liqiao;Chen Tieqiao;Wang Yihao;Pan Delu
摘要现代海洋颜色辐射计和反演算法只依赖于离水辐射的标量处理来获得光学性质和海洋成分(L W)。然而，包含嵌入水溶胶信息的L w偏振分量在很大程度上被忽略了。在这里，我们研究了L的偏振特性，基于矢量辐射传输(RT)模拟，用于测定高度浑浊水域中的悬浮颗粒物。RT模拟结果表明，平行偏振辐射(PPR)可以改善L w的反演效果，其海洋颜色信号占总辐射的相对比例(IOc/It)高于总强度(I)。此外，与I相比，PPR()的反射率对无机粒子浓度(IPC)的变化更敏感，特别是在红色和近红外波段。此外，随着几何条件的变化，呈现出明显的方向性和光谱变化，最大值(~10%)在太阳平面的555 nm处达到极大峰值。此外，还讨论了基于反向传播神经网络的PPR用于提取IPC的优越性，并获得了很高的决定系数。PPR的反演精度显著提高，相对偏差(0.901%)低于I(3.740%)。这项研究强调，PPR可以作为一种替代方法来提取L w，从而获得生物地球化学参数，特别是在浑浊的沿海水域。
ABSTRACT Contemporary ocean colour radiometers and inversion algorithms to derive optical properties and oceanic constituents rely only on the scalar treatment of water-leaving radiance (L w). However, L w polarization components, which contain embedded hydrosol information, have been largely neglected. Here, we examine L w polarization characteristics, based on vector radiative transfer (RT) simulation, for determining suspended particulate matter in highly turbid waters. The RT simulations show that parallel polarization radiance (PPR) can improve the retrieval of L w, with a higher relative fraction of ocean colour signal to total radiance (I oc/I t) than the total intensity (I). Moreover, the reflectance for PPR ( ), compared with that for I, is more sensitive to inorganic particle concentration (IPC) variations, particularly those in the red and NIR bands. Additionally, displays significant directional and spectral variations with respect to geometrical conditions, and the maximum (~10%) is highly peaked at 555 nm in the solar plane. Furthermore, the superiority of PPR to retrieve IPCs, based on the back propagation neural network, was discussed with a very high determination coefficient. Significant improvements in inversion accuracy were observed for PPR with a lower relative deviation (0.901%) than I (3.740%). This study highlights that PPR might be used as an alternative approach to retrieve L w and thus to derive biogeochemical parameters, particularly in turbid coastal waters.