Use this URL to cite or link to this record in EThOS:
Title: Ocean data assimilation using the temperature-salinity relation and water mass diagnostics
Author: Troccoli, Alberto
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2000
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
In this thesis a novel method for assimilating upper ocean temperature profiles with salinity adjustments into numerical ocean models is presented. The approach uses a T-S relation more local in space and time than the climatological T-S used in previous studies. The assimilation method also avoids convective instability as the temperature data are introduced. In order to test the method, three sets of experiments are carried out. First, Conductivity-Temperature-Depth measurements in the western tropical Pacific, and also instantaneous fields from an ocean model, are used to test the assimilation method by combining one profile with another. These tests recover the salinity profiles and the 0-500-m dynamic height very well (differences smaller than 1 dyn cm). By contrast, analyses using a climatological T-S relation did not provide a good salinity profile or dynamic height (greater than 3 dyn cm errors). Second, a synthetic assimilation experiment using a 3-D primitive equation model is carried out. Four runs are considered: the truth (Tr), the parallel (Pa) and two assimilation runs, one in which the salinity method is applied (AST) and the other in which salinity is left unmodified during the temperature assimilation (ASZ). The only difference between Pa and Tr is that Pa is forced by a wind stress 15% larger than Tr, so as to simulate a systematic observational error. AST and ASZ use the same forcings as Pa. Vertical temperature profiles down to a depth of 525 m are taken as synthetic data from Tr and assimilated every 30 days for two years into AST and ASZ. Results show that AST yields better salinity analyses than both Pa and ASZ, which, in terms of rms errors, translate into at least 15% improvement at the end of the 2-year experiment. In addition to assessing the success of the assimilation method (e.g. using T and S rms), a more physical analysis of the model modifications, due to the assimilation, is presented.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available