Absorption-based algorithm for satellite estimating the particulate organic carbon concentration in the global surface ocean

Abstract

Particulate organic carbon (POC) in the surface ocean contributes to understanding the global ocean carbon cycle system. The surface POC concentration can be effectively detected using satellites. In open oceans, the blue-to-green band ratio (BG) algorithm is often used to obtain global surface ocean POC concentrations. However, POC concentrations are underestimated in waters with complex optical environments. To generate a more accurate global POC mapping in the surface ocean, we developed a new ocean color algorithm using a mixed global-scale in situ POC dataset with the concentration ranging from 11.10 to 4389.28 mg/m3. The new algorithm (a-POC) was established to retrieve the POC concentration using the strong relationship between the absorption coefficient at 490 nm (a(490)) and POC, in which a(490) was from the Ocean Color Climate Change Initiative (OC-CCI) v5.0 suite. Afterward, the a-POC algorithm was applied to OC-CCI v5.0 data for special regions and the global ocean. The performances of the a-POC algorithm and the BG algorithm were compared by combining the match-ups of satellite data and in situ dataset. The results showed that the statistical parameters of the a-POC algorithm were similar to those of the BG algorithm in the Atlantic oligotrophic gyre regions, with a median absolute percentage deviation (MAPD) value of 22.04%. In the eastern coastal waters of the United States and the Chesapeake Bay, the POC concentration retrieved by the a-POC algorithm was highly consistent with the match-ups, and MAPD values were 33.06% and 26.11%. The a-POC algorithm was also applied to the Ocean and Land Color Instrument (OLCI) data pre-processed with different atmospheric correction algorithms to evaluate the universality. The result showed that the a-POC algorithm was robust and less sensitive to atmospheric correction than the BG algorithm.