TEST AND ANALYSIS OF A MODEL OF WATER USE BY SORGHUM1

Abstract

We carried out a field experiment with a well-developed sorghum crop to validate a water use and uptake model proposed by Van Bavel and Ahmed (1976). Actual transpiration is determined from standard daily weather data and the leaf area index and root penetration development, as inputs, and from a canopy resistance, which is a function of the effective water potential of the leaf canopy. This canopy potential is, in turn, a function of the transpiration rate, a constant hydraulic root/canopy resistance, and an effective soil water potential. The latter is a root density weighted average value of the water potential over a number of soil layers. Water uptake from each layer is determined from the relative root density and the difference in water potential between the canopy and that layer. The algorithm allows water uptake as well as water excretion by the roots. The calculated evapotranspiration over a 50-d period was not significantly different from the experimental values. Soil water content profiles calculated with the model showed some systematic deviation from the measured values, possibly because the physical nature of the profile was assumed to be homogeneous. The spread of the simulated results reflects the field variation of hydraulic soil properties. A sensitivity analysis showed minor response to rooting depth and distribution and moderate response to the excretion of water by the roots into dry soil layers. Finally, we examined the ability of the model to predict root zone water content and canopy temperatures, the latter with a view toward potential application of remote sensing in the determination of canopy water stress.