Thank you for your substantial revisions to the first draft of this manuscript. If you can accommodate the two requests from the reviewer as stated below, then the manuscript may be accepted with minor revisions. These requests (quoting from the review) are:

1. My only remaining concern is Eq. 25 and the statement about variance equalling the area under the curves in Figs.5ff. The authors respond to my comment with a quote from Stull 1988, which states that "the area under any portion of the curve continues to be proportional to the variance". I agree to the statement in Stull, but it does not say that the area under the curve equals the variance. It states that any portion is proportional. The integral under $F(k)$ is the variance (see Eq.8.6.2a in Stull 1988). It becomes very clear when you look at the units of the x- and y-axis in Fig.5. The unit of the y-axis ($k F(k)$) is m^2 s^{-2}, which is the unit of the variance. Integrating over the wave number yields the unit m s^{-2} which is not the unit of a variance.

2. Additionally, I think that Eq.25 is a bit irritating, because it introduces a function of the wave number, but on the right hand side of the equation, the sampling frequency appears. I think it would be clearer to use $k$ in the right-hand side and maybe introduce $k$ as a function of $U$ and $f_s$. In p.16,l.19 I would recommend to write in the text that for cross spectra, the real part of $F_{ij}$ is used.

Thank you for your interesting and useful contribution!