The authors have well addressed the reviewer's comments and improved the readability of the manuscript.

A minor query remains in my mind. 40 m/s folding appears rather large as compared to the 2-3 m/s statistical error of the ECMWF wind components, which is generally accessible prior knowledge on the wind. A 4-sigma value would probably be adequate to suppress ambiguity at the data assimilation stage, i.e., about 10 m/s folding time. Moreover, following wind scatterometry, ways exist to assimilate ambiguous observations successfully. What motivates the large 40 m/s folding time?