Satellite techniques for studying ocean circulation.

Satellites provide a unique semi-synoptic view of the world's oceans. In recent years, two forms of remotely sensed data have been particularly useful in providing information about ocean circulation, namely altimetric measurements of sea surface height (SSH) and infrared radiometric measurements of sea surface temperature (SST). However, in order to interpret new types of data correctly and obtain meaningful results, new techniques must be developed. In this thesis, techniques to process TOPEX/POSEIDON radar altimeter SSH data and Along-Track Scanning Radiometer (ATSR) SST data are developed. These techniques are tested in the South Atlantic Ocean. The effectiveness of an existing technique to correct for across-track variations in altimeter sampling and the associated SSH errors due to across-track mean sea surface variation is studied. The effects of orbit error removal and interpolation on altimeter data are investigated using ocean model data from the Parallel Ocean Climate Model (POCM). A technique to obtain absolute velocities from altimetry alone is implemented and its accuracy assessed through use of the POCM data. Remnant cloud contamination in the ATSR 0.5° night SST data is discovered and a new technique to remove the cloud contamination is proposed and tested. The seasonality of this cloud contamination is investigated and is found to coincide with the occurrence of marine stratiform clouds. Finally, the relationship between SST and SSH data is examined. It is found that spatial cross-correlations between SST and SSH are surprisingly high (~0.7) in regions associated with fronts and mesoscale variability such as the Agulhas, the Antarctic Circumpolar Current and the Brazil/Falkland regions. In these areas, coherency analysis reveals that the cross-correlations peak at wavelengths of 400-600 km. The strength of the cross-correlations is found to be seasonal, peaking in the winter and minimising in summer