Sea ice thickness is one of the most sensitive variables in the Arctic climate system. In order to quan- tify changes in sea ice thickness, CryoSat-2 was launched in 2010 carrying a Ku-band radar altimeter (SIRAL) designed to measure sea ice freeboard with a few centimeters accu- racy. The instrument uses the synthetic aperture radar tech- nique providing signals with a resolution of about 300 m along track. In this study, airborne Ku-band radar altime- ter data over different sea ice types have been analyzed. A set of parameters has been defined to characterize the dif- ferences in strength and width of the returned power wave- forms. With a Bayesian-based method, it is possible to clas- sify about 80 % of the waveforms from three parameters: maximum of the returned power waveform, the trailing edge width and pulse peakiness. Furthermore, the maximum of the power waveform can be used to reduce the number of false detections of leads, compared to the widely used pulse peakiness parameter. For the pulse peakiness the false clas- sification rate is 12.6 % while for the power maximum it is reduced to 6.5 %. The ability to distinguish between differ- ent ice types and leads allows us to improve the freeboard retrieval and the conversion from freeboard into sea ice thick- ness, where surface type dependent values for the sea ice den- sity and snow load can be used.