Improving Observer Data Quality Using GIS

The North Pacific Groundfish Observer Program (NPGOP) is responsible for collection and quality control of observer data used in management of the Alaskan fisheries. One of the data quality issues faced is the detection of species identification errors (outliers) in catch composition data.  Due to the quantity of data collected by observers, it is cost prohibitive and unrealistic for staff to check species composition data for potential geographic outliers in real time.  Historically, observer datasets required extensive post-field processing, and undocumented outliers remained in debriefed data. In 2009, NPGOP developed GIS models to frequently address inseason outliers and twice a year scrub post-field (debriefed) data to identify remaining outliers within the database.  The GIS model filters the data through approximately 90 species range polygons representing known distributions of various fishes, crabs and corals. Outliers remaining in debriefed data are documented, and individually verified by debriefers.  Early detection of outliers using GIS has reduced staff time required during debriefing and improved data quality by addressing outliers inseason, highlighting outliers for debriefing, and providing documentation to support the debriefed data.