Error analysis of sea surface wind speed estimation based on GNSS airborne experiment

Global Navigation Satellite System (GNSS) signals can be exploited to remotely sense a number of geophysical and geochemical parameters such as ocean surface wind speed, ocean surface altitude and soil moisture in a cost-effective way. This study investigates ocean surface wind speed estimation with a focus on the analysis of the estimation error. A low-altitude airborne experiment was conducted recently and the collected data were processed to generate the delay waveforms through cross-correlation. Wind speed estimation was realised by fitting the theoretical waveforms with the measured waveform. For each of the eight satellites whose elevation angles are greater than 20deg, a sequence of 100 wind speed estimates were produced. The results show that the root mean square error of the eight error sequences ranges from 0.3m/s to 0.94m/s. Based on the limited number of samples, the error distribution is approximately bell-shaped with a Gaussian distribution.