Algorithms to retrieve the spectral diffuse attenuation coefficient of light in the ocean from remote sensing

Area of Science:

• Oceanography
• Remote Sensing
• Optical Oceanography

Background:

• Common ocean color algorithms exhibit significant bias in spectral diffuse attenuation coefficient (K_d(λ)) retrievals compared to profiling float measurements.
• Tuning empirical coefficients of existing algorithms reduces this bias using multi-satellite data.

Purpose of the Study:

• To address the generalization limitations of tuned ocean color algorithms for K_d(λ) estimation.
• To evaluate and implement an analytical algorithm for K_d(λ) retrieval based on radiative transfer theory.

Main Methods:

• Utilized a multi-satellite match-up dataset to tune empirical coefficients of ocean color algorithms.
• Developed and applied an analytical algorithm that computes K_d(λ) from inherent optical properties (IOPs) derived from remote sensing reflectance (R_rs) inversion and transmitted radiance information.
• Validated K_d(λ) estimates using profiling float measurements, in-situ datasets, and radiative transfer simulations.

Main Results:

• Tuning algorithms reduced bias but may still lead to temporal/geographical biases in global applications.
• The new analytical algorithm demonstrated minimal bias against float measurements for K_d(λ) at 412 and 490 nm.
• Satisfactory performance was observed across various water types (clear to turbid) and evaluation methods.

Conclusions:

• An analytical algorithm offers a promising approach to overcome generalization issues in global K_d(λ) estimation.
• Further research is needed to assess the hyperspectral performance of both tuned and analytical algorithms.