Abstract:
One of the standard techniques used to measure evapotranspiration indirectly is Bowen ratioenergy
balance (BREB) method. It is a micrometeorological technique, and it is widely used
because of its simplicity, robustness, and affordability. This method has been applied in this study
to partition available energy (Rn – G) into turbulent fluxes at the study site in Akure, Nigeria during
wet and dry periods (July to December, 2014).
BREB method estimates sensible and latent heat fluxes from measures of vertical gradients of air
temperature and vapour pressure. Both net radiation and soil heat flux are measured directly. These
are then combined in surface energy balance equation.
Results show that apart from errors of measurement of net radiation and soil heat flux introduced
into the estimates of turbulent fluxes, Bowen ratio-energy balance method often produces totally
unacceptable results. Wrong signs, and extremely inaccurate magnitude of latent and sensible heat
fluxes which occurs when Bowen ratio approaches -1 are common with this method. These
problems occur as a result of resolution limits of instruments used.
Also, under moist conditions, the BREB method can give good results for evapotranspiration
estimates, but the method may not be so accurate under very dry conditions. Under moist
conditions, error analysis showed that relative error in evapotranspiration is only small if the
relative error in Bowen ratio is also small. In dry conditions, however, the absolute error in
evapotranspiration is always fairly small because of the small value of evapotranspiration itself.
In this study, errors associated with Bowen ratio-energy balance method are evaluated. A method
to find the range of Bowen ratio around -1 which produces inaccurate flux estimates of latent and
sensible heat fluxes is presented. The excluded region of Bowen ratio is not constant but depends
on the vapour pressure differences measured in each averaging period and the resolution limits of
the device used. If temperature and vapour pressure gradients and the resolution limits of the
sensors are known, spurious data obtained from the BREB method would easily be recognized.
Objective criteria to eliminate spurious data are also presented.
