The remote sensing reflectance of turbid estuarine environments has been difficult in the past due to difficulties in ground-truthing these environments. The development of high-resolution in situ profilers of inherent optical properties (IOPs) has made ground-truthing these areas feasible. The ac-9Ô provides a high-resolution profile of the inherent properties of absorption ("a" in units of m-1) and attenuation ("c" in units of m-1) at nine different wavelengths (l ). The HydroscatÔ allows the measurement of back scattering (bb in units of m-1) at 6 different wavelengths. The values obtained by these instruments were integrated over one optical depth at each corresponding wavelength and compared to the remote sensing reflectance (Rrs measured using a SpectrixÔ ) for the same location and time. The IOPS were related to the Rrs through the equation:
where bbp is the bb due to particulate matter, bbw is the bb due to water, aw is the absorption due to water, a g is the absorption due to dissolved substances, apf is the factor to scale the ap value, agf is the factor used to scale the ag value, and bias are scale offsets. The factor 0.0515 is the relationship between the transmission across the air water interface, ratio of upwelling radiance (Lu) to upwelling irradiance (Eu), and the subsurface radiance distribution field. The b(l )bp was modeled by the relationship:
where h is the coefficient for the shape of the particulate back scattering curve. The parameters of apf, agf, b(550)bp, h , and the bias offsets were adjusted in iterations until the modeled Rrs spectrum closely matched the measured Rrs spectrum. The resulting a(l ) and b(l )b were compared to the integrated values determined using the ac-9Ô and HydroscatÔ . The modeled values were found to be similar to the measured values. Differences were discovered at selected sites in the back scattering coefficient. The difference may be the result of factors affecting the 0.0515 coefficient, the differences due to integration of the IOPs not matching the integration by the Kd values, or improper scaling in the model. The IOPs were found to have closure to the AOPs and models for determining constituents of the water column could be improved or developed for turbid estuarine waters.
Figure 1. ;The procedure for calculation of the integrated IOPs. (a) Ed data were selected for l values that were approximately the values of the ac-9 and Hydroscat wavelength. Data that showed light values approaching near zero were used to determine Kd (b). The Kd values and diffuse attenuation depths were then put into the equation to integrate the values (c) and the process was repeated for the next l (d).
Figure 2. ;"Tweaking" of the model to achieve an Rrs spectrum. (a) Raw Spectra and model at baseline. (b) Baseline to Rrs spectrum and adjustment of b(l )bp. (c) adjustment of absorption to achieve spectrum.
Figure 3. ;Model results in Griffin Bay 8/4/98 for the Nugget.
Figure 4. ;Model results for East Sound 8/5/98 for the Nugget.
Figure 5. ;Model results for a different instrument using a different a(l )p and a(l )CDOM spectrum test model. Locations are West Sound on the Barnes on 8/3/98 and East Sound on the Barnes for 8/5/98.
