UC San Diego

As pH measurement technologies evolve and are increasingly used to study the ocean’s CO2 system, there is a need to assess the uncertainty of seawater pH measurements, particularly those based on the spectrophotometric method using an indicator dye, which is used to calibrate alternative pH measurement methods, including autonomous pH sensors. In this dissertation, I investigated methodological and instrumental contributions to uncertainty in spectrophotometric pH measurements, developed tools for their quality control, and evaluated their likely quality based on their consistency with other seawater CO2 parameters. The accuracy of seawater pH measurements has been questioned due to observations from open ocean cruises of a significant pH-dependent discrepancy between measured pH and pH calculated from dissolved inorganic carbon (CT) and total alkalinity (AT), using a thermodynamic model of seawater acid-base systems. Based on an analysis of high quality CO2 measurements on four open ocean cruises, I showed that a combination of plausible biases in the constants and assumptions used to calculate pH can explain the observed inconsistencies and that there is likely an unaccounted, possibly organic, contribution to the AT measured in the open ocean. Next, I developed methods for analyzing spectra to identify and possibly correct instrumental contributions to spectrophotometric pH measurement error, such as those that may arise from degradation of the spectrophotometer lamp. In another chapter, laboratory and chemical modeling experiments were conducted to evaluate how adjustments for effect of the indicator addition on the sample pH contribute to uncertainty in the final pH result. Assumptions about the indicator’s behavior can result in error in the pH correction, but is minor, except when using high dye concentrations in short pathlength cells. Finally, I evaluated the performance of pH measurements using impure m-cresol purple, which contains absorbing impurities that can bias pH measurements. Calibrated impure dye measurements can still be inconsistent with purified dye measurements depending on the assumptions about the impurity absorption behavior or the purity of the reference dye. This dissertation contributes to understanding of the uncertainty of spectrophotometric pH measurements and our ability to use ocean pH data to characterize the CO2 system.