Understanding the water composition in marine environments is essential to preserve and manage their resources. Since the component distribution influences the propagation of light, the water content may be estimated from light-propagation measurements if their inherent optical properties (IOPs) are known (absorption and scattering). But actual instrumentation cannot provide an accurate characterization of IOPs (all angular scattering values are difficult to obtain), and therefore, particle modeling techniques become fundamental to complement the measured data. Mie and T-matrix theories have already been adopted to characterize living cells and suspended mineral particles. Additionally, Mie theory, in combination with optimization tools, have also been used to estimate the intrinsic particle properties, such as their internal refractive index. However this method only assumes spherical particles, and significant deviations can be obtained when dealing with more complex structures. In this contribution, a heuristic-search methodology to estimate the refractive index of marine particles will be presented. It is based on a genetic algorithm that exploits the principles of evolution to find the optimal solutions. It does not present limitations regarding to boundary values nor not-feasible measurements and can be easily programmed to select between Mie and T-matrix approaches, thereby allowing more complex structures than a sphere. The algorithm is experimentally tested considering as inputs the attenuation and scattering coefficients, both measurable quantities, in different simulated scenarios of mono and polydisperse particle size distributions with different shapes

Aquatic Sciences Meeting, Aquatic Sciences: Global And Regional Perspectives - North Meets South, 22-27 February 2015, Granada, Spain

Peer Reviewed