基于QAA算法的近岸和内陆水域固有光学量反演

Inherent Optical Properties Inversion Coastal and Inland Waters Based on QAA Algorithm

摘要: 水资源在人类的生活环境中起着十分重要的作用，固有光学特性(IOPs)作为反映水生生态环境的重要因素可以使用QAAs算法进行估算。然而，大多数算法依赖412 nm波段估算水组分吸收系数。新半分析算法(QAA-LN)基于遥感反射率比值估算内陆和近岸水体的IOPs。该算法主要由两个部分组成。首先，选择常见的665 nm波段作为参考波段反演总吸收系数a和后向散射系数b_bp。其次，通过490 nm和665 nm波段遥感反射率比值计算浮游植物吸收系数(aph)和颗粒和有色溶解有机物(adg)。使用收集的近岸和内陆水域数据集对模型进行评估，a (λ)、b_bp (λ)、a_ph (λ)和a_dg (λ)的平均绝对误差(MAE)分别为0.20 m⁻¹、0.18 m⁻¹、0.19 m⁻¹、0.28 m⁻¹，结果表现良好。还将新算法用于实测数据匹配的Landsat-8 OLI卫星数据，也能展现出较好的性能，并应用到澳大利亚Hume水库。QAA-LN有潜力作为一种简单有效的算法，未来将扩展到其他卫星监测全球内陆和近岸水域IOPs的变化。

Abstract: Water resources play a crucial role in human living environments, and inherent optical properties (IOPs) serve as important indicators of aquatic ecological conditions that can be estimated using QAA algorithms. However, most algorithms rely on the absorption coefficients of water constituents estimated at the 412 nm wavelength band. The new semi-analytical algorithm (QAA-LN) estimates IOPs for inland and coastal waters based on the ratio of remote sensing reflectance. This method consists of two main components. Firstly, it selects the commonly used 665 nm band as a reference to retrieve the total absorption coefficient (a) and the backscattering coefficient (b_bp). Secondly, it calculates phytoplankton absorption coefficient (a_ph) and the absorption by particles and colored dissolved organic matter (a_dg) using the ratio of remote sensing reflectance at 490 nm and 665 nm. The model was evaluated using the collected coastal and inland water data sets. The mean absolute error (MAE) of a (λ), b_bp (λ), a_ph (λ) and a_dg (λ) were 0.20 m⁻¹, 0.18 m⁻¹, 0.19 m⁻¹, and 0.28 m⁻¹, indicating satisfactory performance. The new algorithm was also applied to the Landsat 8 OLI satellite data in conjunction with the measured data, demonstrating favorable performance, particularly in its application to the Hume Lake in Australia. QAA-LN has the potential to serve as a simple and effective algorithm, with plans for future expansion to monitor changes in IOPs of global inland and coastal waters using other satellites.