Development and sensitivity analyses of a hyperspectral optical Monte Carlo model for coastal and estuarine water quality applications

A hyperspectral optical Monte Carlo Model has been developed for shallow aquatic systems such as estuarine and coastal waters. The model simulates the propagation of optical radiant energy (light) in a homogeneous or stratified and optically deep or shallow water column. Individual photons are tracked in a 3-dimensional aquatic environment. The model can reproduce hyperspectral as well as monochromatic spectral data given a direct light (collimated light) and/or diffuse light, concentrations (depth dependent) of constituents in the water column, their inherent or apparent optical properties (absorption and scattering coefficients) and bottom reflectance. The bottom boundary can be a diffuse or specular surface, or a combination of the two, with wavelength dependent reflectance.; The model technique is described and results are compared with previous analytical solutions (Bostater et al., 1996a) and other quasi-analytical and stochastic models. The sensitivity of the results to the number of photons used in simulations of this model suggest approximately one million photons are needed to predict reflectance with an error of +/− 0.0001.; The model can be used to calculate the upwelling and downwelling diffuse irradiances, the upwelling and downwelling shape factors, the upwelling and downwelling mean cosines and conversion coefficients. The Monte Carlo model assumes the water surface is flat and thus the model has limitations with respect to the effect of air-sea interface on radiation transfer in water.; The model can be used to predict hyperspectral passive remote sensing signatures and may have future utility to active remote sensing applications. The model can also be used to generate synthetic images (as seen from airborne or spaceborne remote sensing platforms) expected to be representative of water types as determined by Bostater et al. (2002).