Accurate Estimation of Net Community Production From O2/Ar Measurements

Under physically isolated conditions, net community production (NCP) can be accurately estimated from the rate of oxygen evasion to the atmosphere derived from local mixed layer oxygen/argon measurements. We use a simple box model to demonstrate that, when physical inputs are negligible, the sea-to-air flux of biological oxygen (bioflux) represents the average NCP exponentially weighted over the past several residence times of oxygen in the mixed layer. This new weighting scheme shows that there is no apparent lag between bioflux and exponentially weighted time-averaged NCP. Furthermore, a strict steady state assumption is unnecessary to this relationship. However, this widely used O-2/Ar method is not effective in dynamic coastal zones where low oxygen water upwells to the surface. Yet these zones are highly productive and their episodic productivity needs to be quantified. We use a quasi-2-D version of the Regional Ocean Modeling System, including oxygen and argon as prognostic variables, to explore the application of this method and the relationship between NCP and bioflux in a coastal upwelling system. We show that bioflux is an accurate measure of NCP over large regions of time and space. Bioflux is most biased near the shore following upwelling favorable winds, where bioflux is sometimes negative (flux from the atmosphere to the ocean) and even positive bioflux values can severely underestimate NCP. Assessing a range of model variables that are easily observed in the field, we show that sea surface temperature is the most effective at identifying bioflux measurements that are likely to be biased. Plain Language Summary Marine biological production affects atmospheric carbon dioxide levels, fish stocks, and oxygen levels in the ocean interior. However, productivity rates are challenging to quantify accurately, particularly in dynamic coastal zones. Measurements of dissolved oxygen/argon ratios are widely used to estimate productivity rates, but the method is compromised in coastal regions where prevailing winds cause deeper, oxygen-poor waters to rise to the surface. The method is also thought to suffer from a time lag between the production of oxygen by photosynthesis and its detection by the method. Using a simplified theoretical computer model, we show that the oxygen/argon method actually detects, with no lag, a weighted average of past productivity rates. This new conception of what the method measures will improve efforts to quantify productivity and to intercompare different field methods. Then, using a more complex model, we demonstrate that the oxygen/argon method can effectively measure productivity rates in much of the coastal region but will be highly biased near the shore during wind events that bring cold, low oxygen waters to the surface. Sea surface temperature, measured at the same time as oxygen/argon, provides an effective way to identify such biased results and exclude them from coastal observations.