Bio-optical properties during the summer season in the Sea of Okhotsk

• We evaluated the bio-optical properties in terms of Chl a concentration, Rrs, and the light absorption coefficients.
• At 443 nm, CDOM absorption dominated the total non-water absorption.
• The summer Chl a values derived from the ocean-color algorithms were significantly higher than the in situ data.
• The largest overestimation occurred in the region where the contribution of aCDOM to at-w at 443 nm reached more than 80%.

The Sea of Okhotsk is one of the most productive ocean regions in the world. However, the in situ bio-optical properties, which are crucial for satellite ocean-color of the productivity, remain uncertain in this region because little data have been available. We conducted an in situ observation and evaluated the bio-optical properties in terms of chlorophyll a (Chl a) concentration, spectral remote sensing reflectance (Rrs), and the light absorption coefficients of phytoplankton (aφ), non-algal particles (aNAP), and chromophoric dissolved organic matter (aCDOM) in the summer of 2006. The data covered a wide range of Chl a levels in surface waters from 0.3 to 8.5 mg m−3. At 443 nm, aCDOM dominated (64% on average) the total non-water absorption (at-w) in this study area. Based on the in situ Rrs data, surface Chl a concentrations that were estimated using the sea-viewing wide field-of-view sensor (SeaWiFS) OC4v6 and the moderate-resolution imaging spectroradiometer (MODIS) OC3M algorithms were significantly higher than the in situ data by more than 160% and 260%, respectively. In particular, the largest overestimation occurred in the region where aCDOM at 443 nm accounted for more than 80% of at-w near the mouth of the Amur River. However, except the CDOM-rich stations, the performance of the OC4v6 and OC3M algorithms became better (i.e., their mean normalized biases were reduced to 50% and 66%, respectively). We conclude that the operational global algorithms were applicable to the summer season in the Sea of Okhotsk except the CDOM-rich region, in which new approaches for ocean-color algorithms (i.e., local algorithms) would be required.