Trailing Edge Slope Ocean Wind Speed Retrievals Using CYGNSS's Full DDMs

The launch of the CYGNSS constellation in 2016 represents NASA's first mission dedicated to the acquisition of GPS signals reflected off the earth’s surface. The major geophysical measurement objective is ocean surface winds with particular emphasis placed on the improved characterization of tropical cyclones. The conventional method for retrieving ocean surface winds using spaceborne Global Navigation Satellite System Reflectometry systems, of which CYGNSS is one example, is to measure the reflected power from the specular point on the ocean surface. Accurate measurement of surface power is complicated by calibration of the various power levels from the transmitting satellites, geometric effects, receiving system RF chain characteristics, among others. This article applies an alternative trailing edge slope (TES) method, which is more immune to end-to-end calibration uncertainties. It is nonetheless noted that because TES is a fundamentally shape-based measure of the waveform, it is susceptible to factors that may distort the measurements' slopes. This included measurements made with low SNR (on the order of 1.5 dB or less) due to low level of surface scattering or observations within the low gain portion of the receive antenna pattern. For this reason, in this work retrieval attempts are largely limited to those with a receive gain on the order of 10 dB or better. The TES method is based on the slope of the measured signal as a function of delay. Given the need for a delay space that extends beyond the
$\approx$
2.5–3.5 chips provided by CYGNSS's standard Level-1 measurements, its utility is explored using one of the constellation's special “full DDM” downlink modes. The TES method is shown to provide competitive retrievals including measurements from the surface, where the conventional method provides ocean wind speed estimates of lower accuracy. Tradeoffs using the TES method are discussed as well as simple improvements in the method are given. It is also noted that because with TES, measurements' slopes are the fundamental observable no calibration of measurements is required.