Improved air-sea CO2 flux estimates by adding sailboat measurements

Sailboats expand the observational network of sea surface partial pressure of CO₂ (pCO₂), particularly in the undersampled Southern Ocean through regularly repeating circumnavigations, however, their added value to the fCO₂-product based ocean sink estimate (S_ocean) has thus far not been quantified. Here, we show through an observing system simulation study with different sampling schemes how integrating sailboat data from different race tracks improves air-sea CO₂ flux estimates.
We find that neural network reconstruction of the air-sea CO₂ flux used within the Global Carbon Budget, when reconstructing a model that mimics present-day real-world sampling, underestimates the ocean carbon sink. This is consistent with recent studies on the interior accumulation of carbon. Increased and continuous sampling by sailboats reveals a stronger carbon sink and improves present-day estimates from 0.06 to -0.02 mol C m⁻² yr⁻¹ (0.99 μatm to -0.32 μatm for the fCO₂ estimate), particularly in the Southern Ocean between 40°S and 60°S. The improvement in reconstructions persists even when data from three circumnavigation tracks contain artificial measurement biases. However, the additional data remains insufficient to correct the overestimated air-sea CO₂ flux trend. While sailboat data has the potential to improve air-sea CO₂ flux reconstructions, expanding the observational network and maintaining long-term time series is crucial to minimize discrepancies between fCO₂-products and Global Ocean Biogeochemical Models.