The inherent optical properties of the oceans: From closure to prediction

Precise in situ measurement of the spectral absorption and scattering coefficients in several regions has revealed patterns in the distribution of the inherent optical properties on spatial scales that were previously unobtainable. The precision of the measurements was found to be consistent and unbiased across a variety of oceanic regimes and are therefore useful for studies of radiative transfer. The spectral information obtained during this research was used to define fundamental relationships between the inherent optical properties and the apparent optical properties of the ocean.; A multiple band ratio algorithm based on the relationship between the absorption coefficient and the remote sensing reflectance was developed to provide a means to test the optical measurements for closure. A large database of synoptic measurements of the spectral absorption coefficient and the remote sensing reflectance was tested for closure using this algorithm and it was found that radiative transfer works to within instrument accuracy. Furthermore, it was demonstrated that inversion to obtain the absorption coefficient is possible using this algorithm providing that the spectral dependence of the absorption coefficient can be accurately modeled.; A model based on the horizontal variability in the vertical structure of the backscattering to absorption ratio was developed to predict the amplitude of an internal wave using the spatial information in the remote sensing reflectance. The results from a combined aircraft and in situ measurement experiment showed that the predicted amplitudes of the internal wave were comparable to the depth fluctuations of the thermocline observed in the in situ temperature profiles.; In an effort to aid primary productivity experiments, an empirical model to predict the photosynthetically available radiation light levels from the absorption coefficient profiles at 490 nm was developed based on in situ data collected in the Gulf of California. The model was able to predict the depth of the one percent light level with a standard error of 4 m. This model provides a method to estimate the daytime light levels from nighttime absorption coefficient measurements at 490 nm.