During the Friday Harbor Optical Oceanography class, 1998, an in situ optical oceanographic data set was collected for several sites in the Puget Sound area. Data from the East and West Sounds have been analyzed with respect to the relationships between the downward diffuse vertical attenuation, Kd and the inherent optical properties, IOPs. The relationships available for determining the connections between these properties have been investigated for decades. The foundation of these relationships relies heavily upon radiative transfer theory, (RTE). RTE coupled with data analysis has led to several models through which one can relate the Kd to the IOPs. Historically one of the firsts was the Wilson and Honey model (1979), where; Kd » a + (b/6). In 1984, based on Monte Carlo simulations, Kirk published a relationship which depends not only on the measured IOPs but on the solar zenith angle as well; Kd » a/µsw [1+ (0.425µsw — 0.19)b/a]1/2 . In working with relationships between remotely sensed reflectance and the IOPs of the ocean (Zaneveld, 1985), the relationship was theoretically defined as: K(p ,z) » c [1-0.52 b/c —0.44 (b/c)2] . Additionally Kd was determined with the HydrolightÓ model, using WETLabs ac-9 profiles as part of the input.
As expected, the results were variable with wavelength. The West Sound results indicate that all models performed well in the blue region of the spectrum. The r2 values ranged from (Kirk) 0.76 to (Zaneveld) 0.68. Proceeding to larger wavelengths the models become progressively less accurate, with the exception of Hydrolightã . The Hydrolightã model was most accurate in the green wavelengths where r2 = .076. Results from the East Sound analyses indicate a similar trend in spectral capability. However, all models had a lesser degree of accuracy. The models failed completely in reproducing the measured Kd values in the red region of the spectrum. Errors in the Hydrolight output for this analysis could be attributed to the absorption and attenuation inputs to the model. The input was suspect as the values were from the WETLabs HISTAR (ac —100) meter. Hydrolight runs should be preformed with ac-9 data, as the HISTAR has not been well characterized.
There are many interesting trends in the data that deserve attention, however; due to time constraints, these trends were set aside. Examples are as Kd is directly proportional to absorption, the depth profiles reveal layers at both sites which on first approximation correspond to salinity, temperature, fluorescence and CDOM profiles. This relationship is seen more clearly in spectral analysis of the Kd values at those depths. Additionally having Kd and a values allows the determination of how µ varies with depth in these areas.
Figure captions
Figure 1 Comparison of Kd (412) values, (West Sound, WA), from various models and the in situ measurement by Satlantic profiler.
Figure 2 Comparison of Kd (555) values, (West Sound, WA) from various models and in situ measurement by Satlantic profiler.
Figure 3 Comparison of Kd (676) values, (West Sound, WA) from various models and in situ measurement by Satlantic profiler.
Figure 4 Regression comparison Kd (412), of model outputs to the in situ measurement.
Figure 5 Regression comparison Kd (555), of model outputs to the in situ measurement.
Figure 6 Regression comparison Kd (676), of model outputs to the in situ measurement.
Figure 7 Comparison of Kd (440) values, (East Sound, WA), from various models and the in situ measurement by Satlantic profiler.
Figure 8 Comparison of Kd (532) values, (East Sound, WA), from various models and the in situ measurement by Satlantic profiler.
Figure 9 Comparison of Kd (676) values, (East Sound, WA), from various models and the in situ measurement by Satlantic profiler.
Figure 10 Regression comparison Kd (440), of model outputs to the in situ measurement.
Figure 11 Regression comparison Kd (532), of model outputs to the in situ measurement.
Figure 12 Regression comparison Kd (676), of model outputs to the in situ measurement.
Figure 13 Spectral comparison of measured and modeled output.
Figure 14 Depth profiles of Satlantic profiler measured Kd values for East and West Sound.
Figure 15 Spectral comparison of depth = 5 m for East and West Sounds.
Figure 16 Spectral comparison of depth = 7 m for East and West Sounds.
Figure 17 Average cosine of the down welling light field, absorption, and scattering plotted as a