Neural network approach for correction of multiple scattering errors in the LISST-VSF instrument.

The LISST-VSF is a commercially developed instrument used to measure the volume scattering function (VSF) and attenuation coefficient in natural waters, which are important for remote sensing, environmental monitoring and underwater optical wireless communication. While the instrument has been shown to work well at relatively low particle concentration, previous studies have shown that the VSF obtained from the LISST-VSF instrument is heavily influenced by multiple scattering in turbid waters. High particle concentrations result in errors in the measured VSF, as well as the derived properties, such as the scattering coefficient and phase function, limiting the range at which the instrument can be used reliably. Here, we present a feedforward neural network approach for correcting this error, using only the measured VSF as input. The neural network is trained with a large dataset generated using Monte Carlo simulations of the LISST-VSF with scattering coefficients b=0.05-50m-1, and tested on VSFs from measurements with natural water samples. The results show that the neural network estimated VSF is very similar to the expected VSF without multiple scattering errors, both in angular shape and magnitude. One example showed that the error in the scattering coefficient was reduced from 103% to 5% for a benchtop measurement of natural water sample with expected b=10.6m-1. Hence, the neural network drastically reduces uncertainties in the VSF and derived properties resulting from measurements with the LISST-VSF in turbid waters.

Experimental and theoretical investigation of waveguided plasmonic surface lattice resonances.

Plasmonic nanostructures are good candidates for refractive index sensing applications through the surface plasmon resonance due to their strong dependence on the surrounding dielectric media. However, typically low quality-factor limits their application in sensing devices. To improve the quality-factor, we have experimentally and theoretically investigated two-dimensional gold nanoparticle gratings situated on top of a waveguide. The coupling between the localized surface plasmon and waveguide modes results in Fano-type resonances, with high quality-factors, very similar to plasmonic surface lattice resonances. By combining plasmonic surface lattice resonance and waveguide theory, we present a theoretical framework describing the structures. By immersing the fabricated samples in three different media we find a sensitivity of ∼50 nm/RIU and figure of merit of 8.9, and demonstrate good agreement with the theory presented. Further analysis show that the sensitivity is very dependent on the waveguide parameters, grating constant and the dielectric environment, and by tuning these parameters we obtain a theoretical sensitivity of 887 nm/RIU.

Efficient Monte Carlo simulation reveals significant multiple scattering errors in underwater angular scattering measurements.

Multiple scattering can severely affect the accuracy of optical instrumentation. Variance reduction methods have been implemented to improve a Monte Carlo model developed to simulate volume scattering functions measured by LISST-VSF instruments. The implemented methods can result in more than a tenfold increase in efficiency. The simulation is used to analyze multiple scattering errors for a range of Fournier-Forand (FF) phase functions. Our results demonstrate significant errors in the scattering coefficient, backscattering coefficient and phase function, where multiple scattering errors may only be considered negligible (<10%) for scattering coefficients <1 m-1. The errors depend strongly on the scattering coefficient but also increase when phase functions become more forward-peaked.

Analysis of multiple scattering errors in LISST-VSF volume scattering function measurements using Monte Carlo simulations and experimental data.

A Monte Carlo algorithm has been developed to investigate the effects of multiple scattering on the volume scattering function measured by the LISST-VSF instrument. The developed algorithm is compared to experimental results obtained from bench-top measurements using 508nm spherical polystyrene beads and Arizona test dust as scattering agents. The Monte Carlo simulation predicts measured volume scattering functions at all concentrations. We demonstrate that multiple scattered light can be a major contributor to the detected signal, resulting in errors in the measured volume scattering function and its derived inherent optical properties. We find a relative error of 10% in the scattering coefficient for optical depths ∼0.4, and it can reach 100% at optical depths ∼2.