Increased pressure on the available limited water resources for agricultural production has a significant impact on sugarcane production. Routine monitoring of evaporation with reliable accuracy is essential for irrigation scheduling, for more efficient use of the available water resources and for management purposes. An indirect method for estimating evaporation involves measuring the sensible heat flux (H) from which latent energy flux and hence total evaporation can be calculated, as a residual using the shortened energy balance from measurements of net irradiance and soil heat flux. Various methods for measuring H may include Bowen ratio energy balance, eddy covariance (EC), flux variance (FV), optical scintillation, surface renewal (SR) and temperature variance (TV). Each method has its own advantages and disadvantages, in terms of method theoretical assumptions, accuracy, complexity, cost, fetch requirements and power consumption. The TV and SR methods are inexpensive and reasonably simple with a reduced power requirement compared to other methods since they require high frequency air temperature data which is obtained by using an unshielded naturally-ventilated type-E fine-wire thermocouple at a single point above the canopy surface. The TV method is based on the Monin-Obukhov similarity theory (MOST) and uses the mean and standard deviation of the air temperature for each averaging period. Currently, there are two TV methods used for estimating sensible heat flux (HTV) at sub-hourly time intervals, one includes adjustment for stability, and a second that includes adjustment for air temperature skewness. Another method used to estimate sensible heat flux from the mean and standard deviation of air temperature is based on MOST and uses spatial second-order air temperature structure function. For the TV method adjusted for stability and the method based on MOST that uses a spatial second-order air temperature structure function, the Monin-Obukhov atmospheric stability parameter () is needed. The parameter  can be estimated from EC measurements or alternatively estimated independently using an iteration process using horizontal wind speed measurements. The TV method including adjustment for air temperature skewness requires the mean and standard deviation of the air temperature and air temperature skewness for each averaging time period as the only input. The SR method is based on the coherent structure concept. Currently, there are various SR models method for estimating sensible heat flux. These include an ideal SR analysis model method based on an air temperature structure function analysis, the SR analysis model with a finite micro-front period, combined SR with K-theory and combined SR model method based on MOST. The ideal SR analysis model based on an air temperature structure function analysis should be calibrated to determine the SR weighting factor (). The other SR approaches require additional measurements such as crop height and horizontal wind speed measurements. In all of the SR approaches, air temperature time lags are used when calculating the air temperature structure functions. In this study, the performance of TV and SR methods were evaluated for estimation of sensible heat and latent energy fluxes at different heights for air temperature time lags of 0.4 and 0.8 s for daytime unstable conditions against EC above a sugarcane canopy at the Baynesfield Estate in KwaZulu-Natal, South Africa. For all methods, latent energy flux (LE) and hence evaporation was estimated as a residual from the shortened energy balance equation using H estimates and net irradiance and soil heat flux density measurements. The ideal SR analysis model method based on an air temperature structure function analysis approach was calibrated and validated against the EC method above the sugarcane canopy using non-overlapping data sets for daytime unstable conditions during 2008. During the calibration period, the SR weighting factor was determined for each height and air temperature time lag. The magnitude of ranged from 0.66 to 0.55 for all measurement heights and an air temperature time lag of 0.8 s. The value increased with a decrease in measurement height and an increase in air temperature time lag. For the validation data set, the SR sensible heat flux (HSR) estimates corresponded well with EC sensible heat flux (HEC) for all heights and both air temperature time lags. The agreement between HSR and HEC improved with a decrease in measurement height for the air temperature time lag of 0.8 s. The best HSR vs HEC comparisons were obtained at a height of 0.20 m above the crop canopy using = 0.66 for an air temperature time lag of 0.8 s. The residual estimates of latent energy flux by SR and EC methods were in good agreement. The LESR at a height of 0.20 m above the canopy yielded the best comparisons with LEEC estimated as a residual. The performance of the TV method, including adjustment for stability, and

Thesis (Ph.D.)-University of KwaZulu-Natal, Pietermaritzburg, 2010.