The U.S. National glider network: Application of QARTOD recommended quality control methods to glider CTD data sets
作者
John Kerfoot,Rebecca Baltes,Mark Bushnell,Luke Campbell,Kelly Knee
标识
DOI:10.1109/oceans.2016.7761356
摘要
A current focus of the Integrated Ocean Observing System's National Glider Data Assembly Center (NGDAC) is on the adoption and implementation of quality control (QC) methods in use by the broader oceanographic community. In 2015, U.S. IOOS, soliciting input from this community, began drafting a glider specific QC manual focusing on temperature and salinity measurements made by autonomous underwater gliders. The manual implements guidance provided by the Quality Assurance/Quality Control of Real-Time Oceanographic Data (QARTOD) Manual for the Real-Time Quality Control of In-situ Temperature and Salinity Data manual [1] and focuses on the observations of temperature, conductivity, and on the calculation of practical salinity, collectively referred to herein as TS observations. The manual has recently been officially accepted and the NGDAC has started to design and implement a subset of the tests outline in the document. The NGDAC recognizes the unique knowledge and understanding, by the glider operators, of the global and local environments they work in. As such, we propose a two-fold approach to glider data quality control: 1) Glider operators may (or may not) apply global and/or locally unique QC checks of both real-time and delayed-mode glider data sets prior to submitting the data sets to the NGDAC, 2) In the event that these QC tests are not applied at the local level, the NGDAC will apply QARTOD recommended methods to the received data sets, storing the results of this process in an additional set of variable written to the NetCDF files distributed to the public and transmitted on the Global Telecommunication System. This approach preserves the original data sets while providing the first step in a comprehensive approach to ensuring that the data delivered by the National Glider DAC are suitable for science-quality analysis and model assimilation. We expect to provide this level of QC by the end of 2016.