Earth, Atmospheric and Planetary Sci., MIT, Rm. 54-1524, Cambridge, MA 02139
A numerical study carried out in anticipation of acoustic thermometry of ocean climate (ATOC) measurements is described. The aim of the study is to combine long-range tomographic measurements with numerical models and other a priori information in order to improve the ocean state estimate. A 4-yr integration of the Semtner and Chervin general circulation model is assumed to represent the real ocean. The integration was performed with a nominal resolution of a 1/4 deg in the horizontal and 20 layers in the vertical. Sound-speed profiles along the proposed North Pacific ATOC paths in conjunction with canonical spectra for the scales not resolved by the model are used to generate simulated tomographic data. The estimation problem is addressed using a lower resolution primitive-equation model of the North Pacific. The second model is assumed sufficiently realistic that perturbations from the Semtner and Chervin ocean may be treated as linear. A time-independent state transition matrix is computed for large-scale density anomalies about the mean of the second model. The estimation problem is then solved using a Kalman filter and the resulting state estimate is compared with the Semtner and Chervin ocean.