Bio-optical characteristics of the Patagonia Shelf break waters: Implications for ocean color algorithms

Bio-optical data for spring and summer phytoplankton blooms at 176 stations along the Patagonia Shelf break and adjacent areas were collected from October 2006 to January 2009 during six cruises. Data included chlorophyll-a concentration (Chla), coefficients of particulate absorption, ap(λ), phytoplankton absorption, aph(λ), beam attenuation, cp(660), diffuse attenuation, Kd(λ), and hyperspectral remote sensing reflectance, Rrs(λ). Surface Chla varied within a wide range from 0.10 to 18.87 mg m− 3 (mean of 2.82 ± 3.35), reflecting differences in both the timing of the blooms and the geographical sampling area. Considerable dispersion was verified for ap(λ), cp(660), and Kbio(440) (Kd(440) minus diffuse attenuation for pure water) as a function of Chla. A hierarchical cluster analysis (HCA) applied to the Rrs(λ) spectra (N = 116) resulted in three classes ordered by spectral features that varied according to Chla. Class 1 (N = 22), Class 2 (N = 52), and Class 3 (N = 42) were grouped according to the Rrs(λ) associated with Chla average of 0.86 (± 0.51), 2.42 (± 1.62), and 8.40 mg m− 3 (± 3.96), respectively. The estimation of Chla and aph(λ) by empirical and semi-analytical algorithms was evaluated using satellite data. Errors in the Chla estimates from the empirical algorithm OC4v6 using in situ Rrs(λ) showed significant statistical relationships with Chla, aph(440) / Chla, and bp(660) / Chla (where bp(660) = cp(660) − ap(660)). Although reasonable agreements were found between the measured and satellite-estimated values, the dependence of the OC4v6 errors on the Chla-specific phytoplankton absorption and particulate scattering coefficients reinforces the need to regionally refine both empirical and semi-analytical algorithms to improve satellite estimates of the bio-optical variables for the Patagonia region.