The WSRA ocean directional wave spectra, along with other calculated sea surface parameters, are produced at about 4 minute intervals. That makes the WSRA a unique instrument which routinely documents the rapid spatial variation of the sea surface inside the evolving hurricane.
Recently, ahead of the 2017 hurricane season, a few modifications were implemented to the WSRA processing algorithm to improve the quality and robustness of the WSRA data products reported in real-time:
- A covariance matrix corrects distortions in the directional wave spectra caused by ocean wavefield induced variation in the backscattered power. The backscattered power varies with the tilts of the sea surface. In turn, the variations in backscattered power distort the apparent topography of the sea surface determined by the WSRA and the resulting ocean directional wave spectra. The implemented covariance matrix was generated using a numerical simulation accounting for: (a) the wave propagation direction with respect to the aircraft track, (b) the sea surface mean square slope, (c) the variation of radar backscatter with incidence angle, and (d) spatial filtering by the antenna beam footprint near nadir.
- Modified algorithm for removal of the 180-degree ambiguity inherent in producing ocean directional wave spectra from WSRA wave topography measurements. Several refinements have been made to improve the prediction of the general direction of the ocean wave field at each observation point using only information available in real-time: hurricane maximum wind speed, radius of maximum wind, and its track and forward speed.
- The wave spectra generated from the topography have a perfect 180-degree symmetry. The 180-degree ambiguity artifact spectral lobes used to be removed before Doppler correction by looking for a spectral minimum. For complex multi-modal spectra, a more robust procedure is to identify the spectral peak, Doppler correct it (which shifts the real and artifact lobes in different directions), and extract the identified spectral peak with the surrounding spectral points above a preset threshold level to the output spectrum. After extracting that data from the original spectrum, the process is repeated until all significant spectral variance has been transferred to the output spectrum.
- Modification to the sea surface mean square slope real-time estimator to eliminate out-of-range samples.
An in-depth analysis will be presented of the verification and validation of above-listed processing algorithm modifications using WSRA data collected during the 2017 hurricane season; eight consecutive flights in TC Maria and two flights in TC Nate.