posted on 2023-05-21, 04:58authored byBushnell, M, Waldmann, C, Seitz, S, Buckley, E, Tamburri, M, Hermes, J, Henslop, E, Ana Lara-LopezAna Lara-Lopez
Scientists who observe and distribute oceanographic data require a process to ensure high-quality data. This process includes quality assurance, quality control, quality assessment, standards, and best practices. In this paper, quality assurance is widely regarded as actions taken prior to instrument deployment to improve the probability of generating good data, while quality control is the effort made to examine the resultant data. Herein we focus on quality assurance and strive to guide the oceanographic community by identifying existing quality assurance best practices preferred by the five entities represented by the authors – specifically, the Alliance for Coastal Technology, the AtlantOS project, the Integrated Marine Observing System, the Joint Technical Commission for Oceanographic and Marine Meteorology, and the U.S. IOOS Quality Assurance/Quality Control of Real-Time Oceanographic Data project. The focus has been placed on QA in response to suggestions from the AtlantOS and QARTOD communities. We define the challenges associated with quality assurance, which include a clear understanding of various terms, the overlap in meaning of those terms, establishment of standards, and varying program requirements. Brief, “real-world” case-studies are presented to demonstrate the challenges. Following this is a description of best practices gathered by the authors from hundreds of scientists over the many years or decades the aforementioned entities have been in place. These practices address instrument selection, preparation, deployment, maintenance, and data acquisition. Varying resources and capabilities are considered, and corresponding levels of quality assurance efforts are discussed. We include a comprehensive description of measurement uncertainty with a detailed example of such a calculation. Rigorous estimates of measurement uncertainty are surprisingly complex, necessarily specific, and not provided as often as needed. But they are critical to data users who may have applications not envisioned by the data provider, to ensure appropriate use of the data. The guidance is necessarily generic because of the broad expanse of oceanographic observations. Further, it is platform-agnostic and applies to most deployment scenarios. We identify the recently created Ocean Best Practice System as one means of developing, sharing, documenting, and curating more specific QA processes. Ultimately, our goal here is to foster their development and harmonization.