Surfactant control of air-sea gas exchange from North Sea coastal waters and the Atlantic Meridional Transect

Suppression of gas transfer velocity (k_w) by surfactants are well established, both in laboratory wind flumes and purposeful oceanic releases. However, the effects on k_w of time and space varying concentrations of natural surfactant are inadequately studied. We have developed an automated gas exchange tank for simultaneous high precision measurement of k_w in unmodified seawater samples. Here we present data from two studies along a coastal North Sea transect during 2012-2013 and the Atlantic Meridional Transect (AMT) 24 from September to November 2014. Measurements of surfactant activity (SA), CDOM absorbance and chlorophyll-a have enabled us to characterize the effects of variable amounts of natural surfactant on k_w. North Sea coastal waters range in k₆₆₀ (k_w normalized to the value for CO₂ in freshwater at 20^oC) was 6.8-24.5 cm hr^-1 (n=20), with the ranges of SA, total CDOM absorbance (200-450 nm) and chlorophyll-a measured in the surface microlayer (SML) of our seawater samples were 0.08-0.38 mg l^-1 T-X-100, 0.13-4.7 and 0.09-1.54 µg l^-1, respectively. The AMT k₆₆₀ ranged from 7.0-23.9 cm hr^-1 (n=22), with SA measured in the SML and subsurface water (SSW) of our seawater samples ranging from 0.15-1.08 mg l^-1 T-X-100 and 0.07-0.43 mg l^-1 T-X-100, respectively.

Importantly, we found 12-45% (North Sea) and 1-43% (AMT) k₆₆₀ suppression relative to Milli-Q water that relate to seasonal and spatial differences in SA. The North Sea demonstrated notable seasonal influences on k₆₆₀ suppression that were related to CDOM absorbance and chlorophyll-a. The degree of k₆₆₀ suppression was highest in summer consistent with k₆₆₀ control by natural surfactant. The degree of k₆₆₀ suppression decreased with distance offshore in the North Sea and displayed a strong relationship with SA (r² = 0.51-0.64, p = 0.02, n = 20). The AMT demonstrated notable differences in k₆₆₀ suppression between hemispheres and across the Longhurst Provinces but the overall relationship between k₆₆₀ suppression and SA is much weaker (r² = <0.01, n = 22). While organic matter composition and sources may have variable control on air-sea gas exchange between the provinces, the poor relationship observed between SA and k₆₆₀ suggests that other environmental factors maybe more influential on air-sea gas exchange in the open ocean compared to North Sea coastal waters.