The effects of seasonal and atmospherically induced sea level variability in satellite altimeter calibration. Results from the GAVDOS Cal/Val experiment.
Thanassis Papadopoulos(1) , Stelios Mertikas(1) , and Erricos Pavlis(2)
Technical University of Crete,
Geodesy & Geomatics Engineering Lab,
GR-73 100, Chania, Crete,
(2) Joint Center for Earth Systems TechnologyUMBC/NASA, 1000 Hilltop Circle, 21250 Baltimore, United States
Shallow water processes, bathymetry, topography and sea-surface slopes give rise to differences in sea-surface heights as measured by satellite altimeters in open water and by tide gauges on the coast. These differences may bias the estimation of satellite altimeter errors if one compares directly altimetric and tide-gauge records.
In this work we attempt to reduce such differences by taking into account the seasonal cycle in shallow waters prior to satellite calibration. By subtracting estimated signals in both records at the altimeter and tide-gauge locations, data variability decreases and both measurements are closer to an equipotential surface.
Within the GAVDOS calibration experiment, the Jason-1 absolute altimeter bias has been estimated to be 144.7±15mm over the cycles 70 to 90. The seasonal differences of sea-surface heights between the GAVDOS tide-gauge site and the area extending a few tens of kilometers South of Gavdos are of the order of 1 cm. These values have been established by hydrodynamic modeling during October 2003-September 2004. In this work, we have found that the absolute bias of Jason-1 and its associated standard error is decreased by about 1cm when the seasonal cycle is subtracted from both measuring system records.
As the seasonal cycle in the Mediterranean Sea is associated with changes in the steric sea level, we also search for differences that may emerge from atmospheric forcing. This is done by employing a surge model that accounts for the atmospherically driven sea level variability.
Along-track Topex/Poseidon data collected North of Souda Bay in Greece have been compared with the Souda Bay tide-gauge readings before and after reducing the seasonal cycle and the surge. After removing the seasonal cycle from the Souda tide gauge records, the surge model accounts for about 60% of the remaining data variability. The surge model can be applied till 2001, thus the analysis has been restricted to the first ten years of Topex/Poseidon mission.