Surface water fCO2 algorithms for the high-latitude Pacific sector of the Southern Ocean

The feasibility of using remotely sensed data jointly with shipboard measurements to estimate the carbon dioxide fugacity in the surface water (fCO(2)sw) of the Pacific sector of the Southern Polar Ocean (S>60 degrees S) is evaluated using a data set obtained during austral summer 2006. A comparison between remotely sensed chlorophyll a (chl a) and sea-surface temperature (SST) with in-situ measurements, reveals the largest bias in areas with rapid and large concentration changes such as at the ice edge, the polar front and in the Ross Sea Polynya. The correlation between fCO(2)sw and SST, chl a, biological productivity estimates and mixed layer depth (MLD) are evaluated, and single and multiple regression methods are used to develop fCO(2)sw algorithms. Single regressions between the study parameters and fCO(2)sw show that most of the fCO(2)sw variability is explained by chl a. The Multi-Parameter Linear regressions were used to create fCO(2)sw algorithms derived from field measurements, and using solely remote-sensing products. Based on the best fits from the two data sets fCO(2)sw estimates have a root means square deviation of +/- 14 mu atm and coefficient of determination of 0.82. The addition of satellite derived estimates of biological productivity in the algorithm does not significantly improve the fit. We use the algorithm with remotely sensed chl a and SST data to produce an fCO2sw map for the entire high-latitude Southern Ocean south of 55 degrees S. We analyze and discuss the seasonal and spatial robustness of the algorithm based on the remotely sensed data and compare with climatologic fCO(2)sw data. (C) 2012 Elsevier Inc. All rights reserved.