A variational data assimilation method for a reduced gravity model is developed. The method is applied to the equatorial Pacific Ocean. In the variational formalism a cost function measuring the "distance" between the model solution and the observations is minimized. The phase speed of the model is used as a control parameter and the optimal spatial structure giving the best fit of the model to the observations is determined. In the minimization algorithm a conjugate gradient descent direction is used. The method is computationally effective, and for the experiments considered convergence is achieved in ten iterations or less. / Several experiments are performed using the model solutions as observations. It is shown that the assimilation algorithm is able to determine the spatial structure of the phase speed, even if observations are available at only three stations. The estimated phase speed is not sensitive to errors in the observations, and the algorithm gives a unique solution to the problem. / Real sea level observations from three stations are assimilated for two different periods. The year 1979 was chosen to represent a year without an El Nino, while 1982/83 was chosen to represent an El Nino year. For 1979 the assimilation gave a phase speed with higher values in the west and lower values in the east compared to the initial guess of a constant phase speed. Assimilation of observations in 1982/83 gave the opposite picture, with lower values in the west and higher values in the east. This result is consistent with observations. The phase speed is proportional to the depth of the thermocline, and during normal conditions the basic stratification consists of a deep thermocline in the west and a shallow thermocline in the east. During an El Nino the picture is reversed. Calculating the correlation coefficient between the model results and the observations shows that the correlation increased for all the stations during the assimilation, even at stations which were not a part of the assimilation. / Source: Dissertation Abstracts International, Volume: 50-08, Section: B, page: 3374. / Major Professor: James J. O'Brien. / Thesis (Ph.D.)--The Florida State University, 1989.
Identifer | oai:union.ndltd.org:fsu.edu/oai:fsu.digital.flvc.org:fsu_78069 |
Contributors | Smedstad, Ole Martin., Florida State University |
Source Sets | Florida State University |
Language | English |
Detected Language | English |
Type | Text |
Format | 137 p. |
Rights | On campus use only. |
Relation | Dissertation Abstracts International |
Page generated in 0.0015 seconds