Processing of radio occultation data requires filtering and quality control for the noise reduction and sorting out corrupted data samples. We introduce a radio holographic filtering algorithm based on the synthesis of canonical transform (CT2) and radio holographic focused synthesized aperture (RHFSA) methods. The field in the CT2-transformed space is divided by a reference signal to subtract the regular phase variation and to compress the spectrum. Next, it is convolved with a Gaussian filter window and multiplied by the reference signal to restore the phase variation. This algorithm is simple to implement, and it is numerically efficient. Numerical simulation of processing radio occultations with a realistic receiver noise indicates a good performance of the method. We introduce a new technique of the error estimation of retrieved bending angle profiles based on the width of the running spectra of the transformed wavefield multiplied with the reference signal. We describe a quality control method for the discrimination of corrupted samples in the L2 channel, which is most susceptible to signal tracking errors. We apply the quality control and error estimation techniques for the processing of data acquired by Challenging Minisatellite Payload (CHAMP) and perform a statistical comparison of CHAMP data with the analyses of the German Weather Service (DWD). The statistical analysis shows a good agreement between the CHAMP and DWD error estimates and the observed CHAMP–DWD differences. This corroborates the efficiency of the proposed quality control and error estimation techniques.
展开▼
