Overview of the Improvements Made on the Empirical Determination of the Sea State Bias Correction
Sylvie Labroue(1) , Philippe Gaspar(1) , Joel Dorandeu(1) , Françoise Ogor(1) , and Ouan Zan Zanife(1)
8-10 rue Hermes, Parc Technologique du Canal,
31520 RAMONVILLE ST AGNE,
It has long been known that the sea level measured by radar altimeters is lower than the true sea level because the backscattered power per unit area is greater for the wave troughs than for the wave crests (Yaplee, 1971). This effect is known as the electromagnetic bias. Other sea-state-related biases additionally affect the altimeter range measurements (e.g. Chelton et al., 1989). These are generally combined with the electromagnetic bias to form the sea state bias (SSB). This bias ranges from a few centimetres up to a few decimetres and is still one of the main sources of error in satellite altimetry. Its theoretical modelling remains an important challenge (e.g. Elfouhaily et al., 2001, Gommenginger 2003). Therefore, the SSB correction still largely relies on empirical models, calibrated on the altimeter data themselves. The goal of this paper is to summarize all the improvements realised in this field at CLS, for the past ten years.
Early estimations of the SSB correction were performed by fitting an empirical parametric SSB model on altimeter-derived sea surface height (SSH) differences, either at crossover points or along collinear tracks (e.g. Born et al., 1982; Gaspar et al., 1994, Chelton, 1994, Chambers, 2003). The use of SSH differences rather than SSH measurements themselves is a simple way to directly eliminate the poorly known geoid signal from the estimation process. However, Gaspar et al (1998) demonstrated that parametric models based on SSH differences are not true least square approximations of the SSB. The origin of this problem comes from the fact that one has to fit an inevitably imperfect parametric model on SSH differences rather than on SSH measurements themselves. To overcome this artefact, Gaspar et al (1998, 2002) developed and refined a non parametric (NP) SSB estimation technique based on kernel smoothing. It was first applied to TOPEX crossover data, providing a SSB correction of 1 cm accuracy.
Recent works have focused on sensitivity studies of this NP technique. Indeed, Labroue et al (2004) analysed in details the difference between crossover and collinear data sets and resulting SSB models in order to highlight the most accurate methodology. They also showed the impact on the SSB estimates of several sources of errors affecting the SSH measurements (time-tag bias, orbit error and data filtering…).
Additional improvement was found through the use of the high frequency MOG2D correction which removes large barometer errors which were directly assimilated into the previous empirical SSB models.
Thanks to all these efforts, the NP method is now mature enough to provide a valuable tool for the SSB analysis. The latest results obtained for ENVISAT, ERS2, TOPEX, Jason 1, Poseidon 1 and GFO in Ku-band will be presented and discussed in this paper. The main difference between the models lies in the correction magnitude. The results are quite consistent for ENVISAT, Poseidon 1, Jason 1 and GFO, but appear lower for TOPEX and higher for ERS2.
The recent analysis of Jason 1 and ENVISAT SSB models in C band and S band has also been performed. These empirical SSB estimates will certainly help to understand and model more accurately the frequency dependence of the electromagnetic bias.