Evapotransipration by vegetation cover is an important component of the water
budget and energy balance in any ecosystem. A key to more improved water
management therefore lies in improving our understanding of evapotranspiration, the
process that drives water use by plants.
Estimations of the turbulent fluxes are required for various applications in
micrometeorology, hydrology, environmental studies and agriculture. Numerous
methods for estimation of turbulent fluxes have been developed and tested. Direct
measurements of fluxes are usually achieved by the eddy covariance (EC) method,
which is considered as the most reliable. However, the application of the EC method
is often problematic. The necessary sensors for wind, temperature and humidity must
respond very fast (resolution of 10 Hz or better) and at the same time must not show
noticeable drift. This makes them delicate, expensive and difficult to calibrate among
other problems associated with the method.
Due to their ability to integrate atmospheric processes along a path length that
may range between a few hundred metres to a few kilometres, optical methods based
on the analysis of scintillation appear to be an alternative and possible supplement to
classical micrometeorological methods such as the EC method, which may provide
local fluxes typically at the scale of 100 m. The use of the scintillometry technique in
surface flux measurements is therefore gaining in popularity.
The accuracy of the measurements obtained by one method is judged by
comparison of the measurements obtained by those of another method considered as
the standard. For turbulent flux measurements, the EC method is taken as the standard
method for the determination of sensible heat fluxes.
This research presents the measurement of sensible heat fluxes using the
surface layer scintillometer (SLS). The SLS system used has a dual-beam and a
recommended path length of between 50 and 250 m. The method was tested against
the EC method for different Bowen ratio (f3) values, as required by the theory, under
different atmospheric stability conditions, as well as for different wind directions
relative to the SLS beam path and slanting beam path orientation. Also presented is an
analysis of the different forms of the Monin-Obukhov Similarity (MOST) functions
used in micrometeorology and suggested by various authors, done by comparing the resulting sensible heat flux measured by the SLS method with the ones calculated
through an iterative determination of the Monin-Obukhov parameters.
A comparison of the structure function parameter of temperature (Ci )
corrected for fJ and those measured (using SLS) was carried out, with the results
showing very good correspondence between the corrected and uncorrected ci values,
indicating that not correcting for fJ for SLS measured ci does not result in
significant error in the resulting ci values, and hence sensible heat flux estimates. A
comparison of the sensible heat flux Fh obtained using EC and SLS methods for fJ <
0.6 and fJ > 0.6 followed and the results also show good correspondence between the
values obtained using the EC and SLS methods, although the agreement is slightly
improved for cases when fJ > 0.6. A sensitivity analysis indicates that both the ECand
SLS-measurements of Fh are influenced by fJ values. A sensitivity analysis on
the influence of fJ on Fh measurements by both the EC and SLS methods further
indicates that the influence of fJ on Fh measurements is not large enough to warrant
correcting Fh measurements for fJ . The F" measurements by the EC method appears
to be influenced more by fJ especially for fJ values less than 0.74. A comparison of
the various methods for computing the empirical similarity functions used by MOST
was also carried out and the results show a significant difference in the Fh computed
following the various methods suggested by different researchers.
As for the agreement between the EC and SLS methods determination of Fh
for the different atmospheric stability conditions, there seems to be a better agreement
in the Fh measurements as noted by correlation coefficients closer to 1 and greater tvalues
obtained during unstable atmospheric conditions in the colder months of June
and August while reduced agreement in the values is recorded in the warmer summer
period from November to December. Also noted is a slight difference in the EC
measurements compared to the SLS measurement of F". The difference in the
measurements is noticed for unstable atmospheric conditions. Also noted is that EC
and SLS measurements of Fh differ slightly when the atmospheric condition is nearneutral.
However the agreement between the Fh values measured by the two
measurement methods is still good. was set up in an inclined position, with the receiver set at 0.68 m above the
ground level and transmitter at 1.68 m, resulting in an effective height difference of
1.00 m. There was generally good agreement in the 2-min measurements of F" by the
two methods for the SLS set up in inclined position, with the 30-min data resulting in
even better agreements. The findings confirm that the SLS set up does not impair its
performance in measuring sensible heat fluxes. This also shows that the SLS would
also work well in non-ideal (heterogeneous) conditions which the inclined optical
beam path mimics. For those days when wind direction was mainly approximately
perpendicular to the beam, the F" values obtained by SLS and EC methods are more
in agreement than when the wind direction was either irregular or parallel to the SLS
beam path. Wind speed also seems to influence the F" estimates by the two methods
since the agreement in the Fh values obtained by the two methods is greater when
wind speed is higher compared to times of the day when the wind speed is reduced.
The atmospheric stability influences the peak position of footprint with the
peak footprint position being further from the measurement point when the
atmospheric stability condition is closer to stable as denoted by the Obukhov length of
-5 and closer to the measurement point for convectively unstable atmospheric
conditions as shown by the Obukhov length of -30. Also shown is that a larger fetch is
required when the atmosphere is convectively unstable as indicated by the contours
plotted on top of the footprint plots.
In general, there seems to be very good agreement in the sensible heat flux
values obtained by the two methods, especially since SLS offers areal-averaged
sensible heat flux measurements compared to the EC method which basically
provides a point measurement. The SLS method therefore offers a better alternative
for obtaining sensible heat flux from larger and heterogeneous area - although to a
limit of250 m since beyond 250 m, the method suffers from a saturation problem. / Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2007.
| Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:ukzn/oai:http://researchspace.ukzn.ac.za:10413/3503 |
| Date | January 2007 |
| Creators | Odhiambo, George O. |
| Contributors | Savage, Michael John. |
| Source Sets | South African National ETD Portal |
| Detected Language | English |
Page generated in 0.0142 seconds