Quality assessment of marine geoscience data collected with the U.S. academic oceanographic research fleet
Abstract
Marine science data collected using the U.S. academic research fleet support a diverse array of investigations across the ocean sciences, targeting important questions ranging from the impacts of global change on ocean chemistry and ecosystems, to the structure and dynamics of ocean circulation, to the nature of volcanic and earthquake processes at the global plate boundaries. Each vessel within the academic fleet is equipped with a suite of sensors (geophysical, water column, and meteorological) available for continuous operation during each expedition, that provide characterization of basic environmental conditions from sea surface to subseafloor. Given the vast expanse of the oceans, many expeditions collect data in previously unexplored regions and these data are of high value for later re-use to build global syntheses, climatologies, and historical time series of ocean properties. Understanding the quality of these "underway" data streams is essential for any of these science applications. However, documentation of data quality is rare and historically, routine quality assessment (QA) of these data streams has been lacking. In most cases, QA has been the domain of the science party for each expedition but limited to the data types of interest for the scientific goals of the specific program, with the quality of other underway data types remaining unassessed. Under the Rolling Deck to Repository (R2R) program, QA of sensor data collected across the academic fleet will be routinely conducted as part of the R2R data pipeline that begins with the submission of cruise data distributions by ship operators. R2R will assemble datasets and documentation, and perform QA for submission to the NOAA National Data Centers (NDCs) for long-term archiving. The goals of the QA are two fold: 1. To provide feedback to shipboard operators to ensure that high quality data are consistently acquired, and 2. To inform future science users of the status of a field data set with detailed information regarding the nature of any problems identified. A wide array of factors impact the quality of data collected at sea, including weather conditions (e.g. winds, sea state), instrument functioning, and operator controlled settings and calibrations. Some data quality issues are best identified visually, but with the volume and diversity of data collected across the research fleet, QA is focused on tests that can be fully scripted. QA tests fall into two general categories: 1. minimal tests of the validity of a submitted data set (existence and validity of the metadata, and verification that all expected data files exist and can be read) and 2. data file level tests (each file can be read, data values are within valid limits, and calculation of statistical quality). Output of the QA is provided as a quality report included within an ISO standard data set metadata record, which is submitted with the data set to the NDCs for long-term archiving.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.U21C..06C
- Keywords:
-
- 1900 INFORMATICS;
- 1926 INFORMATICS / Geospatial;
- 1950 INFORMATICS / Metadata: Quality