Geoscience Reference
In-Depth Information
There are a number of caveats related to the Bowen ratio method. First, the vertical
differences in temperature and humidity can be quite small (and thus relative errors
in β quite large):
If measurements are not made suficiently close to the surface (gradients are largest close
to the surface)
When the aerodynamic resistance is small (strong turbulence)
When the luxes are small (around sunrise and sunset, or under extremely wet or dry
conditions)
Hence accurate instruments are needed with an accuracy of about 0.05 K for tem-
perature, and about 0.02 g kg -1 for humidity (note that the error in the difference is
already double the error in the observations at the individual levels). One way of elim-
inating systematic errors is to interchange the instruments periodically: for example,
measure 5 minutes with sensor 1 on top and sensor 2 below, and the next 5 minutes
with sensor 2 on top and sensor 1 below: combination of the two 5-minute averages
will give a 10-minute average vertical difference without a systematic error (see, e.g.,
McCaughey et al., 1987 ).
Another problematic condition occurs when the available energy is close to zero
(near sunrise and sunset or under cloudy conditions). Then the sensible and latent
heat luxes are of equal magnitude but opposite sign and hence the Bowen ratio will
be close to -1. As a result the luxes, as determined from Eq. ( 7.2 ), will become very
sensitive to measurement errors resulting in unrealistic values.
Question 7.1:
a) What is the value of the Bowen ratio when the available energy equals zero ( Q* -
G = 0)?
b) What are the values of sensible and latent heat luxes for this situation?
7.1.2 Trace Gases
The concept of a Bowen ratio can also be applied to the luxes of trace gases, such as
CO 2 , N 2 O, CH 4 , etc. For some gases this is still the only way to infer vertical luxes,
as sensors that can be used for eddy-covariance measurements are not yet available.
For other gases the fast and compact gas analysers have been developed only recently
(e.g., Hendriks et al., 2008 ).
For a trace gas with a surface lux F X one would deine the 'mass Bowen ratio' as
the ratio of the unknown lux to a known mass lux (e.g., the water vapour lux):
q
q
F
E
,
X X
β X
≡=
(7.4)
Search WWH ::




Custom Search