A DNS capability for obtaining underwater light field and retrieving upper ocean conditions via in-water light measurements

Abstract

Predicting the ocean surface conditions (surface elevation, temperature, wind speed, etc.) becomes more and more important for both real life and military applications. This thesis presents a direct numerical simulation (DNS) capability of solving complicated natural light field patterns in the ocean-atmosphere system. The DNS is applied by means of Monte Carlo method to solve radiative transfer for both unpolarized and polarized natural light radiation, especially strongly affected with dynamic air-sea boundary conditions and inhomogeneous ocean turbulence. In the thesis, radiative transfer theory and Monte Carlo method are introduced. The realization and rigorous code validation are given. In order to apply this software to engineering, applications of radiative transfer theory in ocean-atmosphere system is briefly introduced. To achieve most of the engineering of retrieving ocean surface properties, systematical investigations of how dynamic air-sea boundaries affect the underwater radiance and polarization are taken and discussed. To predict the upper-level ocean conditions based on radiometric underwater measurements, a scheme of inversion algorithm of reconstructing inherent optical properties based on a underwater radiance and irradiance radiometric measurements are described. The key step of the inversion is an analytical solution of Green's function of RTE under the approximation of single scattering. The preliminary trial of the inversion are being taken.