Inversion of Ocean Data.

• Main Authors: Fiadeiro, Manuelo, Veronis, george
• Other Authors: YALE UNIV NEW HAVEN CT KLINE GEOLOGY LAB
• Format: Text
• Language: English
• Published: 1986
• Subjects: Physical and Dynamic Oceanography; Theoretical Mathematics; *OCEAN CURRENTS; *OCEANOGRAPHIC DATA; *INVERSION; VELOCITY; DATA BASES; MATHEMATICAL MODELS; THEORY; PROBLEM SOLVING; RELIABILITY; OCEANS; OCEAN MODELS; CIRCULATION; NORTH ATLANTIC OCEAN; REPRODUCIBILITY; TRACER STUDIES; BERMUDA; TASMAN SEA; TRIANGLES; Ocean circulation; Inverse methods; Coral Sea; North Atlantic
• Online Access: http://www.dtic.mil/docs/citations/ADA168676; http://oai.dtic.mil/oai/oai?&verb=getRecord&metadataPrefix=html&identifier=ADA168676

When we began our study of inverse procedures to determine ocean circulation, our intent was to build on past studies. It turned out, however, that much of the earlier work contained elements that were unclear so we were forced to begin again and to establish reliable and reproducible procedures. Our starting point was to accept the problem that Wunsch (1978) had posed and to sort out the methodology for solving it. We were able to derive an empirical search procedure that led to a level of no motion without requiring the mathematical apparatus of inverse theory. We obtained results for two data sets, one from the Tasman-Coral seas and one from the Bermuda Triangle. Since then several data sets that appeared at the outset to be appropriate for an inverse study have been examined. Lack of significant constraints seems to be characteristic of most oceanographic data sets which is not surprising since they were not gathered for the purpose of making the type of study we have undertaken. In most cases a traditional analysis with all its inaccuracies, assumptions and uncertainties provides as much information as one can obtain. Since our studies have indicated the need for additional considerations to constrain the system more effectively, we have formulated the problem of trying to determine the velocity field from tracer distributions.