Effects of PRN-Dependent ACF Deviations on GNSS-R Wind Speed Retrieval

Abstract

In global navigation satellite systems reflectometry (GNSS-R), the autocorrelation property of the transmitted GNSS signal is an important factor in the delay-Doppler map of the scattered power. The autocorrelation functions (ACFs) of the GPS Level 1 (L1) coarse/acquisition signals exhibit sidelobes. For some of the pseudorandom noise (PRN) codes, the high-amplitude sidelobes are located within one chip from the main lobe and, thus, broaden or narrow its width. These PRN-dependent ACF deviations can introduce significant biases into the GNSS-R scatterometric observables and the wind speed measurements. Simulation results for Cyclone GNSS (CYGNSS) mission show that the ACF deviation induced wind speed error can be larger than 10% (e.g., 2–3 m/s for 20 m/s and 7–8 m/s for 40 m/s wind speeds). These effects can be corrected efficiently by applying empirical factors on the CYGNSS L1 observables.