A decade of ERS satellite orbits and altimetry

Abstract

The First European Remote Sensing Satellite, ERS-1, was launched in July 1991, fol- lowed by ERS-2 in April 1995. Both satellites carry a radar altimeter to serve oper- ational applications and scientific research in the fields of geodesy, oceanography, glaciology and meteorology. Together, the satellites have now been operating for over twelve years. This thesis embarks on the a voyage along several milestones in the lifetime of these satellites, and particularly of its altimeters. Both altimeter missions were hampered by a number of misfortunes, ranging from hardware failures, via design mishaps, to a sub-optimally designed ground segment. Yet, the altimeter data, when properly processed and supplied with a precise orbit, can compete favourably with data supplied by specialised altimeter satellites like TOPEX/Poseidon and Jason-1. The upgrading of all altimeter cor- rections and the tweaking of the basic altimeter measurements of range, significant wave height and backscatter coefficient, are essential steps in this process, outlined in this thesis. Having accomplished that, the ERS altimeters allow the monitor- ing of annual, semi-annual and secular changes in sea level, which are important factors in study of climate change. A key role in the processing of altimeter data plays the position of the satellite, and its progress in time, the satellite orbit. The quality of the recovered sea surface height is limited by the precision at which the orbit can be determined. A long journey along several stepwise improvements of the orbit determination process, both in the technique as well as in the modelling, have resulted in the reduction of the orbit error by a factor of 30! A major player in this process was the estab- lishment of a gravity field model tailored to the ERS satellites that, when applied to ERS orbit determination, outperformed any of the previous and contemporary general-purpose models. Yet, it is shown that this model, DGM-E04, is not just a fabrication for the benefit of ERS only, but is equally suitable for the orbit determ- ination of TOPEX/Poseidon and as a long-wavelength geoid model. The calibration of the ERS-1 altimeter in 1991 required, at a time when the orbit errors were still counted in decimetres, if not metres, a short-arc orbit determin- ation technique centred over the calibration site off the coast of Venice. Again, through elaborate surveying, gathering of altimeter corrections and tide gauge measurements, problem solving and elimination of errors, a final absolute range bias of 41.5 cm was established, a value that is still within the error margins of more recent estimates. A timing error in the ERS altimeter data was discovered in 1995 and its cause remained a mystery ever since, especially when the timing error appeared to ex- hibit daily, annual and inter-annual variations as well. Although the exact cause may never be found, the issue is examined in fine detail and a model is constructed to eliminate the timing error for the benefit of the quality of the altimeter data.