While at the University of Washington, I developed a mooring based air-calibration system for oxygen optodes (Bushinsky and Emerson, 2013). This system relied on pumping atmospheric air down to the sub-surface optode housing to calibrate the sensors once per day.
We subsequently modified this technique and developed Special-Oxygen-Sensor Argo floats equipped with oxygen sensors on long stalks that allowed calibration against atmospheric oxygen after every profile.
This not only provides high quality oxygen from Argo floats, fixing a major problem with sensor accuracy, but gave us the accuracy to discover and correct for an in-situ drift throughout a float’s 4-5 year deployment (Emerson and Bushinsky, 2014; Bushinsky et al., 2016).