Northwest Atlantic EOF-Based Temperature and Salinity Climatology

• Main Authors: Carnes, M. R., Teague, W. J.
• Other Authors: NAVAL OCEANOGRAPHIC AND ATMOSPHERIC RESEARCH LAB STENNIS SPACE CENTER MS
• Format: Text
• Language: English
• Published: 1990
• Subjects: Physical and Dynamic Oceanography; *COMPRESSION; MATHEMATICAL MODELS; DATA PROCESSING; MEASUREMENT; TEMPERATURE; OPTIMIZATION; ENVIRONMENTS; DYNAMICS; VARIABLES; VARIATIONS; REGRESSION ANALYSIS; OCEANOGRAPHIC DATA; PROFILES; VERTICAL ORIENTATION; ERRORS; INTERPOLATION; LEAST SQUARES METHOD; COVARIANCE; ALTIMETERS; SALINITY; CLIMATOLOGY; AMPLITUDE; SURFACE TEMPERATURE; SEA WATER; OCEAN MODELS; HEIGHT; DIGITAL SIMULATION; ORTHOGONALITY; FUNCTIONS(MATHEMATICS); OCEANOGRAPHY; DATA BASES; MOODS DATA SHEET; PE63207N; WUDN259001; Northwest Atlantic
• Online Access: http://www.dtic.mil/docs/citations/ADA230214; http://oai.dtic.mil/oai/oai?&verb=getRecord&metadataPrefix=html&identifier=ADA230214

Climatological relationships which allow transformations among several common oceanographic variables have been developed for the Northwest Atlantic Ocean from profiles of temperature and salinity extracted from the Navy's Master Oceanographic Observation Data Set (MOODS). All pairs of temperature and salinity profiles extending from the surface to at least 100 m were edited and then extended to 2000 m by optimal interpolation. Profiles from the Generalized Digital Environmental Model (GDEM) climatology were used as the assumed mean, and the between-depth temperature and salinity covariances were derived from the data set. Empirical Orthogonal Functions (EOFs) of the temperature and salinity vertical structure were computed from the extended data set. Each profile was then compressed to the amplitudes of the first three EOFs. The error variance for profiles reconstructed from the first three EOFs is only 4% of original profile variance (computed over the entire data set). Step-wise least-squares regression among temperature and salinity EOF amplitudes, dynamic heights, and surface temperature were used to construct several relationships: temperature or salinity amplitudes as a function of dynamic height (which may be derived from an altimeter measurement), temperature (salinity) EOF amplitudes are derived using the regression relationships, and the temperature (salinity) profile can then be constructed as the sum of the products of corresponding amplitudes and EOFs. Errors in derived profiles (or heights) were evaluated over all profiles in the original data set. Keywords: Sea water temperature, Oceanographic data processing, Ocean models.