Application and assessment of a membrane-based pCO₂ sensor under field and laboratory conditions

The principle, application, and assessment of the membrane-based ProOceanus ₂-Pro sensor for partial pressure of CO₂ (pCO₂) are presented. The performance of the sensor is evaluated extensively under field and laboratory conditions by comparing the sensor outputs with direct measurements from calibrated pCO₂ measuring systems and the thermodynamic carbonate calculation of pCO₂ from discrete samples. Under stable laboratory condition, the sensor agreed with a calibrated water-air equilibrator system at −3.0 ± 4.4 µatm during a 2-month intercomparison experiment. When applied in field deployments, the larger differences between measurements and the calculated pCO₂ references (6.4 ± 12.3 µatm on a ship of opportunity and 8.7 ± 14.1 µatm on a mooring) are related not only to sensor error, but also to the uncertainties of the references and the comparison process, as well as changes in the working environments of the sensor. When corrected against references, the overall uncertainties of the sensor results are largely determined by those of the pCO₂ references (± 2 and ± 8 µatm for direct measurements and calculated pCO₂, respectively). Our study suggests accuracy of the sensor can be affected by temperature fluctuations of the detector optical cell and calibration error. These problems have been addressed in more recent models of the instrument through improving detector temperature control and through using more accurate standard gases. Another interesting result in our laboratory test is the unexpected change in alkalinity which results in significant underestimation in the pCO₂ calculation as compared to the direct measurement (up to 90 µatm)