Bhaskaran, Shyamkumar, 1991, Application of satellite altimetry to study the Gulf of Alaska gyre: University of Colorado, Boulder, Ph.D. dissertation, 161 p., illust., maps.
Satellite altimetry has demonstrated the capability to observe a wide range of ocean phenomena. In this study, it was used to examine the Gulf of Alaska region, where the signals of the ocean circulation are fairly small. The research had two components: the first analyzed the variability while the second involved computing a mean sea surface. For the first portion of the study, crossover data from the GEOSAT Geodetic Mission (GM) and colinear data from the Exact Repeat Mission (ERM) were processed separately to form time series at grid points located at the ERM crossover locations. The time series from the GM and ERM were then combined to produce a 3.75 year data set. A portion of the analysis was dedicated to performing a sensitivity study to see the effects of different orbit error models, water vapor, and tidal error corrections. It was then shown that the altimetric time series agreed fairly well with the in situ data, lending confidence to the processing techniques. The altimetric time series during the complete, GM, and ERM time spans were analyzed using Empirical Orthogonal Functions (EOF). Examination of the EOF spatial patterns and amplitudes revealed variability that occurs primarily on annual and interannual time scales. A comparison with EOF analysis of the pressure field during the same time periods showed that the primary annual mode in each of the time spans correlated highly with the annual variation in the pressure field. A second annual mode present in the ERM data set was found to correlate well with sea level at Seward. Regarding the interannual modes, the amplitude time series of the first mode in the complete data set was very similar to the North Pacific pressure index during the 1985-1989 time frame. The maximum correlation was at a lag of 250 days. Finally, an interannual mode was present in all three data sets which was closely linked to the baroclinic variations measured by the hydrographic data. The second component of the study involved computing a mean sea surface using statistical knowledge of the orbit errors which contaminate the data. To verify the procedure and test its accuracy, simulations were performed. The simulations showed that the mean sea surface can be recovered to an accuracy of 15 to 25 cm. in a root mean square sense. These values are too large for absolute dynamic topography studies in the Gulf of Alaska, but improved geoids and orbits should prove the technique to be feasible in the future.
Theses and Dissertations