학술논문

Fixed Observatories and LongTime-series of Dissolved Oxygen Measurements: Good Quality Data is a Challenge
Document Type
Conference
Source
OCEANS 2019 - Marseille OCEANS Marseille, 2019. :1-7 Jun, 2019
Subject
Communication, Networking and Broadcast Technologies
Engineering Profession
Fields, Waves and Electromagnetics
Geoscience
Power, Energy and Industry Applications
Signal Processing and Analysis
Sea measurements
Sensors
Uncertainty
Oceans
Calibration
Measurement uncertainty
Observatories
Dissolved Oxygen
Sensor calibration
Metrology
Calibration Bench
time-series
fixed observatories
data quality
Language
Abstract
Long-term marine observation is primarily aimed at acquiring Essential Ocean Variables (EOVs), for the assessment of their multi-scale temporal variability (monthly to pluriannual). EOVs also provide a unique background to depict unusual events. Feeding from these unique high quality datasets, modelling can be achieved to produce numerical ocean circulation models and atmosphere-ocean models. One of the EOVs currently measured in this context is the Dissolved Oxygen (O 2 ) content in seawater. Today, this parameter is commonly measured with a reference method, the Winkler Titration (Winkler 1888, modified Carrit & Carpenter 1966) based on discrete sampling whereas high frequency measurements are nowadays provided by autonomous O 2 sensors.On a metrological point of view, the measurement of O 2 at the required accuracy for the open ocean studies, 1 µmol.kg-1 (Gruber et al. 2010), is challenging. Moreover, the quality of the reference measurement is strongly influenced by the sampling and analyzing conditions. In situ O 2 time series are historically based on discrete sampling requesting ship based sampling and subsequent laboratory analysis. Over the last 10 years, the improvement of dedicated O 2 sensors allows the community to access to a higher sampling frequency to provide robust in situ time series. Nevertheless, the performances of the sensors are evolving slowly over time, due to embedded sensor intrinsic limitation. For these reasons, calibrations are required to know the corrections/adjustment to apply to the measurements and ensure good quality datasets. These calibrations are achieved in controlled conditions with a calibration bench and comparison with Winkler titration. Pre and post deployment Winkler titration of in situ water samples also mandatory to ensure the good quality of the O 2 dataset. In this context and in the framework of the H2020 project EMSO-Link, the EMSO-ERIC infrastructure consortium supports the collaborative construction of a O 2 sensor (whether stand alone or fitted on generic oceanographic instrumentation) calibration bench. This calibration platform aims to provide reliable metadata for the deployed sensors in order to reach and maintain high quality dataset for oceanography studies. The continued use of these standardized procedures will ensure data consistency between the different observatories in Europe.