Mean and Turbulent Parts
- The Spectral Gap
- As discussed previously and shown to the right, a spectral gap separates
large-scale motions from turbulent scale motions.
- Again, the spectral gap refers to those spatial and temporal scales that
show little variation in wind speed.
- In the figure to the right, the spectral gap is centered on temporal
scales ranging from about 30 minutes to a few hours.
- This encompasses many mesoscale phenomena.
- Hence, you would not observe this power spectrum if studying growing
cumulus clouds, for example.
-
- Q: Where does the grid spacing for many of the operational NWP
models fall within Fig. 2.2? Answer
- Q: Based on your answer above, can the operational NWP models make a
deterministic forecast of turbulence? Answer
-
Separation of Variables into Mean and Turbulent
Parts
- Given an atmospheric flow situation where the large-scale flow is
distinctly separated by a spectral gap from the turbulent scales, it is
quite easy to isolate the large-scale motions from turbulence and therefore,
study the turbulent motions.
- OK, so you are given a time series of wind speed for a
period of two days.
- If you compute the mean wind speed, U,
this represents the mean wind due to the large-scale flow.
- u' represents the flow
due to turbulence, then
- U = U
+ u', where U is the actual instantaneous velocity.
- therefore, u' = U - U
- Q: So, in laymans terms, what is u'? Answer
- Given a portion of your wind speed time series, you can visualize u' and U
as shown in Fig. 2.3

- Q: How often must you take wind measurements to well sample
turbulent eddies ranging in size from a few cm to a couple km in scale?
- Assume a mean wind speed of 5 m/s and that the turbulence is not
changing with time:
- Since turbulence is a three-dimensional problem that is also characterized
by it's thermodynamic and chemical properties, we can generally partition
the following variables into their mean and turbulent parts:
- U = U + u'
- V = V + v'
- W = W + w'
- qv =
qv
+qv'
- q = q + q'
- c = c + c'