A Comparison of Waveform Model Re-Tracking Methods Using Data from CYGNSS

The CYGNSS mission is a new GNSS-R ocean remote sensing constellation of eight small spacecraft. The measurements from CYGNSS allow for the first time a large, densely sampled, and calibrated GNSS-R dataset to be used for ocean surface altimetry. Re-tracking of the reflected signal correlation waveform is one of the most significant performance limiting components of ocean altimetry retrievals. This work demonstrates the use of empirical methods and physical scattering delay/delay-Doppler waveform models, fit to the measured correlation functions in a least-squares sense, to retrieve the specular reflection delay. We examine the performance of waveform modelling methods as compared to single-point re-tracking methods for the first time with a spaceborne dataset. Particular attention is given to the presence of narrow waveforms in the data set, indicative of coherent, rather than purely diffuse reflection. Re-tracking performance is quantified with statistics of the accuracy and precision and examples are presented for returns demonstrating both diffuse and coherent characteristics.