Abstract's details

Altika Cloud liquid water/rain flag validation

Jean Tournadre (IFREMER, France)


Jean-Christophe Poisson (CLS, France); Nathalie Steunou (CNES, France)

Event: 2014 SARAL/AltiKa workshop

Session: Instrument processing

Presentation type: Type Oral

Contribution: PDF file


AltiKa on board SARAL, the indo-french satellite launched in February 2013, is an altimeter developed by the French Centre National d'Etudes Spatiales (CNES). It is a wide-band Ka-band altimeter (35.75 GHz, 500 MHz bandwidth).
It is the first oceanography altimeter to operate at such a high frequency.
This unique technical characteristic of the instrument offers improvements both in terms of spatial and vertical resolution. This improved accuracy should lead to improved observation of ice, coastal areas, inland waters and wave height.

The one major drawback of Ka band is its sensitivity to atmospheric liquid water (both rain and clouds) is high, an order of magnitude than that at Ku band used for Jason2 or Envisat. Studies of Ka band altimeter have shown that even light rain or heavy clouds can strongly attenuate the radar signal, distort the altimeter echo waveform and hampers the retrieval of geophysical parameters.
It is thus necessary to detect and flag the samples potentially affected by atmospheric liquid water. The same problem was also encountered for Ku-band altimeters, such as Topex, Jason-1 or Envisat, for medium and heavy rain.

For these dual frequency altimeters rain flags, based on the differential attenuation of the main (Ku-band) and secondary channels (C-band for Topex and Jason-1 and S-band for Envisat) by rain droplets, were defined and are currently used operationally.
Unfortunately, AltiKa is a single frequency altimeter and this kind of simple and efficient flag could obviously not be used. A rain flag based on the analysis of the altimeter measurement alone has thus been developed for AltiKa.
Based on past experience with Ku-band altimeter data and on the modeling of AltiKa echo waveform in presence of cloud and rain that showed that rain cells or clouds are characterized by sharp coherent along-track variations of off-nadir angle estimates, a method of identification, detection and localization of these along-track transient features series has been developed.
The flagging algorithm is based on the analysis of the variations of the off-nadir angle estimate by Matching Pursuit (MP) algorithm. MP allows the decomposition of a signal into a few salient features or atoms chosen from a dictionary of elementary functions defined by the wavelet decomposition of the signal.

The flag has been tested pre-launch on simulated Ka-band waveforms, and on real Jason-1 Ku-band data affected rain. Different parameters of the algorithm such as the level of noise of the rain-free estimate of off-nadir angles have to be defined post-launch and the results have to be validated.

Firstly, the results of the definition of the different parameters of the algorithm are presented as well as the improvements and modification of the operational flagging algorithm. The statistics of the number of flagged data both at 40 Hz and 1 Hz for the first year of operations are presented and are compared to the pre-launch planned availability of data. The flag is then validated by comparison with rain data from passive radiometers, showing the good performances of the flag. Finally, the degradation of the geophysical parameters estimates in presence of rain is analyzed to show the necessity of a good rain flagging.

Oral presentation show times:

Room Start Date End Date
Ballroom Mon, Oct 27 2014,11:15 Mon, Oct 27 2014,11:30
Jean Tournadre