A Data Base Editor for MOODS (Master Oceanographic Observations Data Set).

Abstract : Large oceanographic data bases are valuable for many applications: climatological estimates of surface and subsurface temperatures, heat content, layer depths, currents (via geostrophy), and frontal locations. Such estimates are useful for describing and comparing ocean regions, priming numerical models, and editing or calculating anomalies from present observations. The Master Oceanographic Observations Data Set (MOODS) contains 3.5 million data observations. The data, which include temperature and salinity profiles, are edited during updates, but this editing has been very superficial and allows for erroneous values. This editor attempts to ferret out the bad data by checking for oceanographic observations that are over land, above the sea surface, below the sea bottom; that have nonmonotonic, duplicate, or negative depths; that contain impossible or all-zero temperatures or salinities; that produce temperature or density inversions; that are misplaced either by location or by season; or that are duplicates. For four MOODS test sets (two Atlantic and two Pacific), the total rejection rate ranged 17-39%. Of these rejections, 9-16% were already flagged during update editing, 1-8% were rejected because of inversions and wild values, and 7-17% were duplicate and misplaced profiles.