Sigma0 Blooms in the Envisat Radar Altimeter Data

Data from satellite altimeters are often degraded by the occurrence of unrealistically high values of the ocean surface radar backscatter cross sections (sigma-0). Various studies on Topex and Jason-1 altimeters data have shown that these events called “sigma-0 blooms” affect almost 6% of the over ocean measurements. On Topex data, sigma-0 blooms have been observed and analysed by Mitchum which has shown that they occur most of the time, but perhaps not always, in regions of climatologically weak winds. Surface slicks could also be the cause of sigma-0 blooms. In any case, contamination of altimeter data by “sigma-0 blooms” is an important issue if we consider that only 60% of the sigma-0 bloom events seem to be rejected by classical altimeter data flagging. This paper gives a comprehensive description of the blooms occurrence (where, when and how long) which is important because many people of the community that make use of the altimetric data are generally unaware of this phenomenon. Statistical characteristics of sigma bloom events are presented using 1-Hz data as well as 20Hz data. Analysis is done for many cycles of Envisat data distributed along the year and also on local and limited data sets. Moreover, we show how waveforms are corrupted and we propose a criterion (such as the Mean Quadratic Error between the echo and the estimated waveform, or such as the sigma0 value or such as the estimated mispointing angle) to characterize its occurrences. The analysis is performed on Ku-band data but relation with S-band data is also considered as it has been proposed by Tournadre looking at Jason-1 data. Spatial and temporal correlations with geophysical parameters (wind, waves ...) are highlighted. Besides the precise characterisation of the blooms, one of the major interests of the study concerns the estimation of the errors induced by these events on the estimated range and sigma0 (wind) parameters.