L-band GNSS-R Retrieval Algorithms for Tropical Cyclones Based on Wind-Wave Growth Similarity

In the past several decades, a combination of theory and empiricism has established robust wind-wave growth similarity relations for wave height and the spectral peak period. These relations imply the correlation of the three quantities in the wind-wave triplets: wave height, peak frequency, and wind speed. In the case of hurricanes, it has been shown that this correlation can be further constrained given a location and sector relative to the storm. Specifically, for locations more than 30 km away from the eye, wind-waves usually dominate, and a location and sector-dependent effective fetch can be found, which enables the determination of two other quantities in the triplets given the knowledge of only one.

Bistatic remote sensing using L-band GPS signals has recently been proposed for spaceborne ocean surface wind speed measurements in adverse weather conditions such as hurricanes. The physical quantity sensed by this class of sensors is the low-pass-filtered mean squared slope (LPmss). In this study, we extend the concept of the wind wave triplets to quadruplets by introducing the LPmss similarity functions. Because of a lack of existing studies on the evolution of LPmss, we exploit the growth relations embedded in a wave-age and wind speed dependent wave spectrum. Equipped with the sector-dependent effective fetch, we design retrieval relations between LPmss and wind speed, as well as LPmss and significant wave height (Hs). Comparisons are made with existing results in literature and wave model products. These relations can be leveraged to guide the retrieval of wind speed and Hs for the upcoming Cyclone Global Navigation Satellite System (CYGNSS) mission, and to help us refine our existing knowledge on the growth of the wind-wave spectrum.