Autonomous ocean carbon system observations from gliders

• Main Author: Possenti, Luca
• Format: Thesis
• Language: English
• Published: 2020
• Subjects: Norwegian Sea
• Online Access: https://ueaeprints.uea.ac.uk/id/eprint/80675/; https://ueaeprints.uea.ac.uk/id/eprint/80675/1/2020PossentiLPhD_thesis.pdf

Climate change is altering the ocean carbonate system decreasing the seawater pH. To quantify these changes novel sampling and monitoring methods are necessary. One of these methods are gliders. The sensors to fit on a glider need to have a compact size, low-cost, stability, accuracy and fast response. For the first time, a CO2 optode (Aanderaa), a potentiometric pH glass electrode (Fluidion) and a spectrophotometric labon-chip pH sensor (UK National Oceanography Centre) were tested on gliders. The CO2 optode was deployed for 8 months in the Norwegian Sea, with an O2 optode. The CO2 measurements required several corrections. The calibrated optode CO2 concentrations and a regional parameterisation of total alkalinity (AT) were used to calculate dissolved inorganic carbon concentrations (CT) with a standard deviation of 11 μmol kg-1. The O2 and CO2 data were used to calculate CT- and O2-based net community production (NCP) from inventory changes combined with estimates of air-sea exchange, diapycnal mixing and entrainment of deeper waters. Because of the summer period the NCP was largely positive. The spectrophotometric pH sensor, the glass electrode and an O2 optode were deployed on a Seaglider for 10 days in the North Sea. Before the deployment, laboratory tests showed that the main source of error for glass electrodes is drift when deployed in seawater. The spectrophotometric sensor was stable with an accuracy of 0.002 and was used as reference to calibrate the glass electrode. The potentiometric sensor failed after 2 days' deployment and was not affected by drift (<0.01), because it had been stored in seawater for 2 months. The spectrophotometric sensor had a mean bias of 0.006±0.008 (1σ) compared with pH derived from discrete AT and CT samples, higher than in the laboratory. The data were used to calculate O2 and CO2 air-sea fluxes and bottom respiration rates.