Seasonal variation in the apparent height of the East Antarctic ice sheet

Abstract

A satellite radar altimeter can be used to monitor surface elevation change over polar ice sheets. Thirty-five months of Geosat Exact Repeat Mission (ERM) data from November 1986 to September 1989 over a section of East Antarctica (69–72.1 ∘S, 80–140∘ E) have been used in this study. A model that considers both surface and volume scattering was used to retrack the altimeter waveforms. Surface elevations for each month after the first three were compared to the average elevations for the first 3 months through a crossover method. The averaged crossover elevation difference changed with time in a way that suggests a yearly cycle in surface elevation. The average amplitude of the cycle is about 0.6 m. We have been unable to find any satisfactory explanation for the observed changes, in terms of either sources of error or contributors to real surface-height changes. We strongly suspect that orbit error plays a major role in producing the variations, although we know of no quantitatively satisfactory source of a quasi-seasonal variation in orbit error. Other possible contributors include a real seasonal variation in accumulation rate, seasonal changes in the delay of the radar signal as it propagates through the atmosphere, unmodeled variations in the depth of penetration of the radar pulse into the firn, changes in the thickness of the ice and the firn zone in response to seasonal variations in pressure and temperature, and the inverted barometer effect. Even though we do not know the cause of the variations, the results show the importance of comparing elevations at the same time of year for observations that are not continuous, while at the same time showing that even annually spaced measurements may not be free of substantial errors associated with interannual variability. The quasi-periodic variations obscure any evidence of a moderate secular change in surface height, if there is one, but a dramatic lowering at rates approaching 1 ma–1, such as are known elsewhere in Antarctica, can definitely be ruled out.