Concurrent estimates of carbon export reveal physical biases in ΔO2/Ar-based net community production estimates in the Southern California Bight

Net community production (NCP) represents the amount of biologically-produced organic carbon that is available to be exported out of the surface ocean and is typically estimated using measurements of the O2/Ar ratio in the surface mixed layer under the assumption of negligible vertical transport. However, physical processes can significantly bias NCP estimates based on this in-situ tracer. It is actively debated whether discrepancies between O2/Ar-based NCP and carbon export estimates are due to differences in the location of biological production and export, or the result of physical biases. In this study, we calculate export production across the euphotic depth during two months of upwelling in Southern California in 2014, based on an estimate of the consumption rate of dissolved organic carbon (DOC) and the dissolved: total organic carbon consumption ratio below the euphotic depth. This estimate equals the concurrent O2/Ar-based NCP estimates over the same period that are corrected for physical biases, but is significantly different than NCP estimated without a correction for vertical transport. This comparison demonstrates that concurrent physical transport estimates would significantly improve O2/Ar-based estimates of NCP, particularly in settings with vertical advection. Potential approaches to mitigate this bias are discussed.