What is the "sweeping velocity" of a turbulent flow? Background:
While reading Rubenstein and Zhou 1997 I encountered an unfamiliar term, "sweeping velocity". 
To quote the authors (pg. 3):

...the sweeping hypothesis makes the decorrelation depend on the sweeping velocity $V$, which is not an inertial range property, but an entirely independent property of the energy containing range...

I'm interested in estimating $V$ because the authors use it to predict the frequency of peak energy production by turbulent flows:
$$\omega \sim V k_{0}$$
Where $k_0$ is the inverse of the integral time scale which characterizes the intertial subrange of the turbulence:
$$\frac{1}{k_0} = \mathcal{T} = \int_{0}^{\infty} r(\tau) \ d \tau$$

(In fact, I believe they describe the entire acoustic energy spectrum generated by the flow as a piecewise function with $k_0$ as the point of discontinuity.)

Question:
What is the sweeping velocity, $V$, and how does it differ from other velocities of a flow (such as mean velocity)? Is there a way to measure $V$ for streams or rivers?
 A: Your Link to the quote is meanwhile broken, but I found this Rubenstein and Zhou 1997
On Chapter III, Page 4-7 the subject is discussed though I didn't find the exact quote. At the end of the Page 6 is said that

The peak acoustic power occurs at frequency $\omega \approx V k_0$. Some preliminary evidence supporting the 4/3 scaling is given by Lilley.

And on Page 5;

where the sweeping velocity V in Eq. (12) is a property of the most energetic scales
of motion. Thus, under the sweeping hypothesis, temporal decorrelation has a nonlocal
character since it is determined for motions of any given scale by motions of possibly much larger scale. It is therefore possible that whatever contribution the most energetic scales may make to the total acoustic power, they always determine the frequency distribution of acoustic energy through their effect on the temporal decorrelation of the inertial range scales. In free shear flows, it is very likely that the sweeping velocity is a property of large-scale coherent structures. In this case, the sweeping velocity will not be predicted by turbulence models, which compute either inertial range or mean
flow quantities.

I have bolded the text which raised my own interest. As on my own findings I have found that with Froude number $Fr=\sqrt{7/4}=1.3229$ there is a comical energetic flow character, which is definitely not predicted by any common Turbulence models, but which I have observed in the Nature and on Experiments.
As the maximum kinetic energy range of any flow is calculated $\frac{1}{2}mv^2$ and the energy losses are growing after $Fr=\sqrt{3}$ The Peak kinetic energy of the flow is reached at $Fr=3.1$, but as the losses the energy losses are already ca. 25% this flow propably can't produce that kind of "peak energy" as you are expecting.
The most energetic oscillations are present at first said froude number value, and the breaking of the wave starts with the latter. These values can be found from my paper
Navier-Stokes existency and smoothness problem, -The Answer
page 8, after the Equation 20.
