Abstract
The signal of Clark‐type oxygen sensors is controlled not only by the oxygen concentration but also by sensor dimensions, temperature, and salinity. We have developed a mathematical model that predicts the O 2 microsensor signal as a function of these parameters. The only model variable, membrane permeability, can be determined by a one‐point calibration; hence, the model can be used for calibrating the signal measured under any temperature and salinity. The model describes the oxygen concentration profile through the sensor and thus allows optimization of sensor construction with respect to stirring sensitivity, signal‐to‐noise ratio, and response time. The model is demonstrated using oxygen water column profiles off the coast of central Chile.