Investigation of Ocean Surface Wind and Wave Coupling Using CYGNSS Observations

The Cyclone Global Navigation Satellite System (CYGNSS) mission aims to measure ocean surface wind speed and roughness in the inner core of tropical cyclones. It has two main advantages: capability to penetrate the eye wall and inner rain bands, and a short revisit time through the use of a constellation of 8 microsatellites. CYGNSS Level 2 data products include the 10m ocean surface wind speed (U10) and the mean square slope (MSS), which are derived from the Level 1 normalized bistatic radar cross section (NBRCS) and the waveform leading edge slope (LES). It has been demonstrated that CYGNSS Level 1 data has a clear dependency on waves not generated by the local winds: the CYGNSS wind speed retrieval error shows a dependence on the significant wave height (SWH) at all wind speeds, and the L1 NBRCS and LES both exhibit an increased sensitivity to SWH at lower wind speeds. These sea state conditions, including the presence of external swell and the degree of wave development, increase the uncertainty of the wind speed retrieval. An "excess MSS" correction was proposed to correct for sea state conditions, and was demonstrated to improve CYGNSS L2 wind speed retrievals. The excess MSS is calculated using ancillary data including the ECMWF U10 wind speed and IFERMER MSS produced by the WaveWatch III (WW3) numerical model, as well as the Elfouhaily et al. wave spectral model. The computed excess MSS, responsible for the sea state condition effect (e.g. the non-local swell), is highly correlated with the significant wave height at low/medium wind speed. Removing the excess MSS from the CYGNSS observed MSS reduces the correlation of the wind speed error with the SWH and improves the wind speed retrieval. The excess MSS approach also has the potential of providing knowledge of ocean surface wave development (e.g. wave age, fetch length). This presentation will review the current status of the "excess MSS" approach for compensating wave effects in CYGNSS wind retrievals, and will describe current progress in extending the approach by incorporating estimates of wave age and/or fetch length into the correction process.