-
文中采用自然海洋水域中粒子散射的散射相矩阵,利用偏振蒙特卡洛仿真模型模拟垂直剖面水体中,船载海洋激光雷达垂直偏振通道和平行偏振通道的回波信号,仿真参数如表1所示。
Parameters Value Laser wavelength/nm 532 Telescope diameter/m 0.3 Field of view/mrad 10, 20, 50, 100, 200, 500, 1000 Platform height/m 5 Phase function Petzold Transmission photon counts 107 Maximum scattering times 20 Profile resolution/m 0.1 Table 1. Parameters for simulation of shipborne oceanographic lidar
文中使用高斯分布设置了三种深度分布在10~30 m的低、中、高浓度散射层,在深度20 m处出现叶绿素a浓度峰值,分别为0.1 mg/m3、1 mg/m3和10 mg/m3,背景叶绿素a浓度均为0.02 mg/m3,低、中、高浓度散射层情况下的叶绿素a浓度剖面如图4(a)所示。水体总吸收系数为纯水吸收系数与叶绿素a吸收系数之和,纯水的吸收系数采用Pope和Fry[15]的测量结果,叶绿素a的吸收系数使用Lee[16]等人建立的浮游植物吸收光谱的参数化表达式计算得到。水体总散射系数为纯水散射系数与叶绿素a散射系数之和,纯水散射系数采用Morel[17]模型,叶绿素a散射系数采用Gordon和Morel[18]提出的关于浮游植物散射系数的计算公式。水体总吸收系数剖面和总散射系数剖面分别如图4(b)和4(c)所示。
Figure 4. (a) Profile of chlorophyll-a concentration in low, medium and high scattering layer; (b) Absorption coefficient a profile; (c) Scattering coefficient b profile
深度剖面分辨率为1 m的船载海洋激光雷达仿真回波信号如图5所示,图中(a)、(b)、(c)分别对应低、中、高浓度散射层的情况,实线表示平行通道的回波能量,虚线表示垂直通道的回波能量。将0~40 m深度分为散射层上(0~10 m)、散射层中(10~30 m)和散射层下(30~40 m)三个深度区间。
Figure 5. Simulated shipborne oceanographic lidar return signals in (a) low, (b) medium and (c) high scattering layer
在散射层上(0~10 m),水体吸收系数和散射系数随深度不改变,可以看作均匀水体,回波曲线变化较为平缓。这一部分中回波能量随深度缓慢衰减,平行偏振通道在不同视场角下的衰减基本不变,三种情况的垂直通道回波曲线变化几乎一致,视场越大衰减越慢。
在散射层中(10~30 m)水体的吸收系数和散射都随颗粒物浓度增大,对光信号的衰减和散射作用都更强,二者共同作用下,随着深度的增加,散射层内的回波信号在的衰减趋势会先变缓(低浓度散射层情况)甚至增强(中浓度、高浓度散射层情况),随后以更快的速度衰减。其中高浓度散射层的衰减程度最严重,中浓度散射层的次之,低浓度散射层的衰减程度最低。另外,在散射层中,颗粒物浓度的增加会导致多次散射增多,这时在深度较大处的光子大部分是多次散射之后到达的,因此水平偏振通道的回波信号较垂直偏振通道衰减更快,在大视场角情况下更为明显,中、高浓度散射层情况下两通道的回波能量在某一深度处开始出现重合。高浓度散射层情况下,在深度25 m以下,回波信号已经完全退偏,因此两通道的信号基本重合。
船载海洋激光雷达仿真回波信号的退偏振比(根据公式(16)计算得到)剖面如图6所示,图中(a)、(b)、(c)分别对应低、中、高浓度散射层的情况。三种水体在海面处的退偏振比均约为0.118,略大于单次散射退偏振比0.1173,退偏振比随深度、叶绿素a浓度和视场角的增加逐渐变大。在散射层以上,由于水体参数相同,低、中、高三种浓度散射层情况的退偏振比随深度和视场角的变化趋势相同。
在散射层中和散射层下,不同水体不同视场角的退偏振比均随深度增大。低、中浓度散射层情况在小视场角下的退偏振比随深度变化较小,而高浓度散射层情况的退偏振比无论视场角大小,均在散射层出现一段时间后产生很大变化,深度为25 m左右时高浓度散射层情况在各个视场角下的退偏振比趋于饱和。
由于假定的水体颗粒物退偏振系数为常数,因此退偏振比相对误差随深度、散射层浓度和视场角的变化规律与退偏振比的变化趋势一致。
Simulation of polarization profiles of water measured by oceanographic lidar
doi: 10.3788/IRLA20211035
- Received Date: 2021-05-06
- Rev Recd Date: 2021-05-19
- Publish Date: 2021-06-30
-
Key words:
- oceanographic lidar /
- polarization Monte Carlo /
- water optical parameter /
- vertical profile /
- depolarization ratio
Abstract:
A Monte Carlo radiative transfer model with polarization was developed to simulate and analyze the vertical profile of received polarization signal of a ship-borne lidar. The measurement errors resulted from different seawater optical parameters and various lidar measurement modes were analyzed as well. A Gaussian distribution function was used to describe the chlorophyll-a vertical profile. The scattering layers were set at 10-30 m with the low, medium and high values of chlorophyll-a concentration ([chl-a]), respectively, and the corresponding maximum value of [chl-a] was 0.1 mg/m3, 1 mg/m3 and 10 mg/m3, respectively. The polarization return signals of the ship-borne oceanographic lidar were simulated with a laser transmission wavelength of 532 nm and field of views (FOVs) of 10-1000 mrad, and the main factors affecting the polarization measurement error were analyzed. The results suggest that the single scattering ratio of lidar return signal decreases with the enhancements of detection depth, [chl-a] and FOV due to the multiple scattering process of laser transferring in seawater. This leads to an increase in the error of the depolarization ratio directly measured by lidar. Let’s take the FOV of 100 mrad as an example. In the case of the scattering layer with a medium [chl-a], the relative errors of the depolarization ratio above (0-10 m), in (10-30 m) and under (30-40 m) the scattering layer were 16%, 125% and 281%, respectively. In the scattering layer, the relative errors of the depolarization ratio were 54%, 125% and 731% for the low, medium and high values of [chl-a], respectively. When the FOV increases from 10 mrad to 1000 mrad, the relative error of the depolarization ratio increases from 6%-28% above the scattering layer, 17%-452% in the scattering layer and 10%-734% under the scattering layer, respectively, for the case of the scattering layer with a medium [chl-a]. Therefore, when using the polarization oceanographic lidar to detect the seawater depolarization ratio, the traditional algorithm for depolarization ratio will introduce a large error due to the multiple scattering process, and a correction is required to improve the detection accuracy of lidar measurement.