POSTER Session 2

Systematic quality control for underway bio-optical data

Most larger research vessels are equipped with flow-through systems measuring next to temperature or salinity also bio-optical data such as chlorophyll-a fluorescence and turbidity in surface waters. Often underway data are only archived as raw data which limits their usability by the scientific community. The underway research data project of the German Marine Research Alliance (DAM) aims at increasing the research vessel’s efficiency by making underway data accessible under FAIR principles. We present workflows from data acquisition on the ship via data transfer and quality control to archiving and visualization. The data undergo a systematic and well-documented quality control procedure based on international standards. The quality control also includes calibration of the chlorophyll-a fluorescence using reference values from discrete water samples or satellite data depending on data availability. Here, we focus on a novel approach based on AI to complement classic quality control steps. A random forest approach is used to effectively detect sections of the bio-optical time series influenced by air bubbles in the flow-through systems. The quality-controlled data is published open access and provides the scientific community with consistent data sets. These can be easily combined with other underway data, e.g., ADCP, temperature and salinity, which are likewise systematically quality-controlled.

Julia Oelker, Carl von Ossietzky Universität Oldenburg, Germany, https://orcid.org/0000-0002-0783-3617

Daniela Voss, Carl von Ossietzky Universität Oldenburg, Germany

Daniela Pieck, Carl von Ossietzky Universität Oldenburg, Germany

Jochen Wollschläger, Carl von Ossietzky Universität Oldenburg, Germany

Code of Conduct
Privacy Policy