Bowen ratio and surface temperature techniques for measuring evaporation from cabbages.

Abstract

Good irrigation water management requires accurate, automated, non-destructive and simple techniques to measure crop water consumption. The actual evaporation from a cabbage crop was measured using the Bowen ratio energy balance technique (BREB), the surface temperature technique and the Penman-monteith method. All models used the shortened energy balance equation to estimate latent heat in which the advected energy is assumed to be negligible. Four irrigations were applied and 17 rainfall events were recorded during the experiment. The soil at the experimental field was a clay loam. An attempt to detect and reduce measurement error that could result from using inaccurate sensors was performed by calibrating the sensors. Data from inaccurate sensors were not used to compute the latent heat. Error and sensitivity analyse were performed, and the integrity of the weather data using the estimates of weather data from an appropriate model were checked. In addition, a comparative study showed that, for daily totals, there was a very small error in the latent heat calculations when fixed "constants" (density of air, specific heat capacity of air, psychrometric constant, slope of the saturation water vapour pressure vs temperature relationship and specific heat capacity of soil) were used instead of calculated ones. The Bowen ratio (β), a fundamental input of the BREB technique, was estimated accepting the Similarity Principle and excluding nighttime data. However, an error in β was also observed during the daytime measurement of the profiles entities because the sensors were wet and the stability condition was different from neutral conditions under which the Similarity Principle could not be observed. Negative values of β were observed when there were strong winds advecting sensible heat into the field under study. Data were rejected during mornings, and during strong advection periods. Data were also rejected when the sensors were wet because of rain or irrigation. In this experiment, only 35 % of data were valid for determining latent and sensible heat estimated using the BREB technique. Comparative analysis showed that the BREB technique overestimated the latent heat by 17 % compared to the Penman-Monteith method. However, both the Penman-Monteith method and BREB technique could not be trusted because of the presence of advection, a component of the energy balance equation normally assumed to be negligible. Either the surface to air temperature differential or the aerodynamic resistance, or both, were the source of overestimation of latent heat using the surface temperature technique. The surface to air temperature differential was large in magnitude when there were high wind speeds and drier conditions in the upwind field . It was small with lighter wind speeds and wetter surface conditions. An error of less than 5 % was attributed to the use of fixed air density and specific heat capacity and acceptance of 2 % and 20 % error in measuring the net irradiance and soil heat flux density, respectively. A comparative study showed that the surface temperature latent heat was overestimated in relation to the Penman-Monteith and BREB latent heat. Generally, the technique has been reported to overestimate evaporation, although to a lesser extent than the 57 % error reported in this experiment when compared to the BREB technique. An analysis of the energy balance closure, taking the Penman-Monteith and BREB as standards, suggested that the surface temperature technique overestimated the consumption of sensible heat from the air. This observation was also confirmed when the eddy correlation technique was used to compare sensible heat estimated using the surface temperature technique. The effect of placement height of air temperature sensors suggested that the consumption of sensible heat would be overestimated if the sensor was placed far from the crop surface. This overestimation in consumption of sensible heat resulted in an overestimation of latent heat. Irrigation water management was analysed using the crop water stress index (CWSI). The CWSI was calculated using the actual to potential evaporation ratio estimated using the Penman-Monteith method and the surface temperature techniques. The estimated and measured actual surface to air temperature differential, and the estimated potential and non-transpiring surface to air temperature differential were also used to estimate the CWSI using the Penman-Monteith method, the surface temperature technique and empirical method. The estimates of the CWSI using these techniques were inaccurate because of the poor correlation between the surface to air temperature differential and the water vapour pressure deficit (or water vapour pressure deficit and net irradiance). However, use of the CWSI estimated using the actual to potential evaporation ratio (CWSI = 1 - λ(a)/ λE(p) compared well to the standard CWSI determined using the Penman-Monteith approach. The actual canopy resistance was estimated using an empirical equation based on the potential canopy resistance, solar irradiance, soil water content and the shelter factor. A value of 50 s m(-1) was estimated for potential (minimum) canopy resistance of the cabbage crop. The soil water content was poorly correlated to CWSI, while the canopy resistance was well correlated. Comparative analysis showed that the estimated soil water content using the soil water balance equation was underestimated in relation to the soil water content measured using the ThetaProbe (frequency domain reflectometry technique) when the evaporation component was overestimated, and vice versa. Soil water content was underestimated throughout the experiment when evaporation from the surface temperature technique was used. There was an underestimation of soil water content in the early stages and overestimation in later stages of the experiment when the BREB and Penman-Monteith evaporation were used. Use of the estimated soil water content using the soil water balance with the overestimated evaporation would result in an early date of irrigation application, an unnecessarily large irrigation amount and frequent irrigations. More research is needed to find the cause of overestimation of evaporation using the surface temperature technique. The robustness of the equipment allowed a long period of measurement without frequent maintenance, as was required when using the BREB technique. The technique can monitor evaporation and irrigation management aspects at a regional scale. A combination of the Penman-Monteith, surface temperature and empirical method can assist the estimation of the crop water requirement by determining the CWSI. Future research would focus on quantification of sensible and latent heat advection, and analysis of additional resistances to water vapour flow from the surface to the atmosphere. The equipment for the BREB should be refined so that it measures actual latent heat under adverse weather conditions for a protracted period. A precise use of the soil water balance equation for water management should take into consideration runoff, vertical flow of soil water through a profile, intercepted water on plant surfaces and an accurately determined evaporation.