Satellite altimeter calibration and validation using GPS buoy technology.
thesisposted on 2023-05-26, 07:58 authored by Watson, CS
Satellite altimeters have become an important tool for the study of global and regional mean sea level change, offering near global coverage and unprecedented accuracy. Issues of calibration and validation remain central to their ability to determine estimates of change at accuracies of better than 0.5 mm/yr.This Thesis provides an absolute calibration of the TOPEX/Poseidon (T/P) and Jason-1 satellite altimeters, undertaken in Bass Strait, Australia. The research provides a contribution to the international calibration effort, with the Bass Strait site situated as the only one of its kind in the Southern Hemisphere.\ \ A unique in situ absolute calibration methodology is presented, reliant on the episodic deployment of GPS equipped buoys at an offshore comparison point. In contrast to other calibration studies, data from the GPS buoys are used to solve for the absolute datum of an offshore oceanographic array (incorporating a pressure sensor, temperature and salinity recorders and a current meter array). Combined with data from a coastal tide gauge and a regional GPS network, the methodology enables the cycle-by-cycle computation of absolute bias, without the necessity of estimating a marine geoid. Emphasis within this Thesis is given to the design and development of the GPS equipped buoys, in addition to the standardisation requirements of the geodetic analysis. The GPS buoy design is applied to both the altimeter calibration problem, in addition to a near shore application involving the calibration of tide gauges in the Antarctic and Sub-Antarctic.The attention to standardisation ensures comparable estimates of in situ and altimeter sea surface height. Differences at the 9 mm level for the pole tide displacement and plus or minus 5 mm for the solid Earth tide displacement are revealed when using the GAMIT GPS analysis suite.The implications of non-standardisation are further illustrated with the presentation of time series analysis from various continuous GPS datasets.\ \ Absolute bias and 1-sigma uncertainties from a formal error budget are 0 plus or minus 14 mm for T/P and +152 plus or minus 13 mm for Jason-1 (for the GDR POE orbits, computed over the calibration phase, 18 Jan 2002 14 Aug 2002). Results over the duration of the T/P mission confirm a dependence on the choice of Sea State Bias (SSB), with the overall mean absolute bias not statistically different from zero. Extending the comparison period between Jason-1 cycles 1 to 101 (18 Jan 2002 06 Oct 2004) reduces the Jason-1 mean absolute bias by approximately 10 mm and reveals a significant slope of -7.6 plus or minus 5.6 mm per year.Whilst the cause for the significant absolute bias remains unexplained, the source of the drift appears attributable to the microwave radiometer, observed to be measuring drier over time (-5.9 plus or minus 2.1 mm/yr). Drift of the POE orbit relative to the JPL GPS orbit is shown to account for the remaining trend observed at the Bass Strait site.After considering geographically correlated errors, absolute bias results show excellent agreement with other international calibration studies.These results aid in understanding the performance of both the T/P and Jason-1 altimeters, further underscoring calibration and cross calibration of altimeters as essential for the study of low frequency oceanographic processes, including regional and global mean sea level change.The inference of geographically correlated orbit errors, and the significant unexplained Jason-1 absolute bias emphasises the need for maintaining globally distributed verification sites and makes it clear that further work is required to improve our understanding of the Jason-1 instrument and its algorithm behaviour.