Sir,

In a recent paper in the Journal of Glaciology the performance of three radar-altimeter “retracking” algorithms was investigated using simulated waveforms (Reference Femenias, Remy, Raizonville and MinsterFemenias and others, 1993). One of the techniques used was described as the Offset Center of Gravity (OCOG) method. There appears to be a misunderstanding about the function of this algorithm which, in itself, is not a retracking procedure but a means of determining the amplitude of the waveform. This amplitude is then used to find the position on the leading edge of the waveform which equals some percentage of the amplitude (e.g. 0.3, used by Reference Partington, W and RapleyPartington and others (1991)). Furthermore, to reduce the effects of the leading edge on the amplitude estimate, each waveform sample is squared. The centre of gravity and waveform width that can be obtained from using the OCOG procedure were never intended to be used to calculate a retrack position in the way that they have been by Reference Femenias, Remy, Raizonville and MinsterFemenias and others (1993). Instead, they were designed to be used as part of a satellite onboard-tracking loop (such as is used for the ice mode of the ERS-1 altimeter). A complete and correct description of a retracking procedure using the OCOG algorithm is given in Reference BamberBamber (1994). It should be noted that using the waveform width and centre of gravity to find the retrack point on the waveform and using the amplitude to “threshold” retrack it give very different results.