Novel Oxygen Optode Sensor with Fast Response Time: In-Depth Characterization and Assessment of the HydroFlash O2 Applicable for Several Ocean Observing Platforms

Tobias Hahn^1*Arne Koertzinger^1,2

• ¹GEOMAR Helmholtz Center for Ocean Research Kiel, Helmholtz Association of German Research Centres (HZ), Kiel, Germany
• ²University of Kiel, Kiel, Schleswig-Holstein, Germany

Optics-based sensors, so-called optodes, for oxygen are used for routine operations on autonomous instrumentation and profiling platforms with great success. Observations of oxygen gradients with high spatial and temporal resolution become increasingly important, while shortcomings still exist, namely time constant problems, stability issues or accuracy levels, that limit leveraging on their full scientific and operational potential. Here we demonstrate the utility of a novel, although currently not commercially available optode: the HydroFlashO2. It was manufactured by Kongsberg Maritime Contros GmbH between 2014--2019, and peer-reviewed studies illustrate its use until today. Our work comprises its first integrated characterization with data from 13 HydroFlashO2 optodes assessing oxygen, temperature, salinity and hydrostatic pressure dependence, long-term stability and drift, response time as well as air-calibration compatibility. We multi-point calibrated this optode up to a RMSE<1umol*L-1 (mean RMSE: 1.79+-0.50umol*L-1), depending on the fit model type. Our laboratory setup yielded a temperature-dependent response time of t63%=3.31+-0.58s, showing no significant difference between a weakly-turbulent and turbulent flow, and being at least 50% faster compared to the two most-common optodes in oceanography, i.e. 4330 (Aanderaa) and SBE 63 (Sea-Bird Scientific). We assessed its pressure dependence between 0–5797dbar yielding an overall factor of 2.372+-0.409% per 1000dbar, based on three multi-point calibrated, drift-corrected optodes and five CTD profiles. Ship-underway, mooring and CTD-cast applications promise high quality observations including fast oxygen level changes. The optode revealed a strong sensitivity of the sensor spot causing erroneous oxygen measurements when exposed to direct solar irradiation during an Argo float test profile. The drift assessment covering a maximum time span of about three years is based on two optodes and yielded linear (R2=0.98) and exponential drift (t=2.35+.0.30yr, 95% CI) behaviors. The HydroFlashO2 is applicable in low to high oxygen, pressure and temperature conditions, yet we do not call for additional performance studies unless the manufacturer reactivates its production and reduces sensor spot issues. In an ocean affected by climate change, reliable oxygen optodes will contribute crucial information about the global oxygen and carbon budget, e.g., through observations in the mixed layer, thermocline or deep sea, that requires assessments of existing and promising instrumentation.