PNNL

Abstract

Several unknowns remain surrounding marine Carbon Dioxide Removal (mCDR) monitoring, reporting, and verification (MRV) practices and capabilities. Current in-situ carbonate sensor technology is limited to proxies (primarily pH and pCO2), requiring calculations and assumptions to estimate the changes in relevant carbonate chemistry parameters, including total alkalinity (TA). While sensing technology is continuously improving, minimal testing has been done on existing sensors to assess performance and energy consumption under mCDR monitoring scenarios. Understanding sensor limitations, potential shortfalls, and how other tools can be paired with sensing will improve MRV efforts and support the development of protocols. Here we assess current sensor technology performance and model outputs to advance monitoring for mCDR via a mesocosm experiment, a field trial, and a model case study scenario. Our findings indicate that correctly constraining pH will be critical for accurate TA estimates with current sensor technology, compared to the variation caused by uncertainty in pCO2. pH sensors with manufacturer-reported 0.01 unit accuracy and 0.1 unit accuracy resulted in estimated alkalinity ranges varying by 1.4% and 29.4%, respectively. Comparatively, pCO2 sensors with manufacturer-reported 0.5% and 5% accuracy resulted in estimated alkalinity ranges varying by 0.2% to 4.4%, respectively. Our findings also indicate that ocean biogeochemistry models will need to be relied on to more accurately constrain the carbonate system given the challenges in detecting signals relative to natural variability and sensor detection limits, which we validated during a pilot field trial. Finally, we present a simplified simulation of mCDR that is used to inform the sensing needs and limitations of MRV for field trials.