The empirical formula of Priestly and Taylor (1972) relating the evapotranspiration from a well‐watered surface to the net radiation and a function of the air temperature is modified to include a saturation deficit term to account for high local advection. The model is tested on two seasons of daily evapotranspiration measurements over irrigated potatoes (Solarium tuberosum L.), resulting in good agreement and a substantial improvement over the unmodified formula. Alfalfa (Medicago saliva L.) measurements are also represented well by the model, but no better than by the Priestley‐Taylor formula calibrated during a period of high advection.

A large number of separate experimental studies of water vapor movement in response to thermal gradients are shown to be independent of water content over a wide range of water contents. Two models, proposed by J. R. Philip and D. A. de Vries (1957) and J. W. Cary (1963), are rewritten in a compatible form and compared to the data. The phenomenological model of Cary describes the data if the phenomenological coefficient β has a value between 1.0 and 3.5, whereas the mechanistic theory of Philip and de Vries underpredicts the vapor transport observed in the experiments in all cases.A modification of the theory of Philip and de Vries is proposed which increases the influence of liquid water on vapor transport. In the new formulation the experimental information is adequately described and both models are rendered compatible. Both experimental and theoretical results suggest that water vapor movement may be described by a simple expression which is a known function of temperature, and is independent of water content over a wide range of saturation. Even without calibration it is suggested that this expression can describe water vapor flux to within a factor of 2.

Previous studies report conflicting results regarding the sensitivity of cowpeas to growth stage water deficits. Our 2‐year field study was designed to monitor the effects of water deficit imposed at three growth stages on the yield and water use of cowpeas (Vigna ungiculata (L.) Walp.). Six irrigation treatments were used to impose the water deficit including a well watered control treatment (WWW) and five deficit treatments (DWW, WDW, WWD, DWD, WDD). Cowpeas were grown on a sandy loam soil (coarse‐loamy, mixed thermic Haplic Durixeralf) on the experimental farm in Riverside, California. The pre‐dawn leaf water potential, as measured by the pressure bomb technique, was a better indicator of crop water stress than was the midday leaf water potential. The pre‐dawn leaf water potential at any growth stage progressively decreased with time following cessation of irrigation, but completely recovered after soil water storage was recharged by irrigation or rain. The most sensitive growth stages to drought were flowering and podfilling, with yield reduction from 35 to 69% depending on the timing and length of the drought treatment. A soil water deficit during the vegetative stage had the least effect on crop yield. This, coupled with decreased evaporation, resulted in a water‐use efficiency which was greater than that of the control treatment, which was adequately watered throughout the season. The water‐use efficiencies of the other stage deficit treatments were decreased below that of the control because of large decreases in crop yield. Seed yield of cowpeas was found to be linearly related to an integrated water stress indicator based on the predawn measurement of leaf water potential. We found that the relative dry matter yield was linearly related to crop transpiration as determined by the difference between a water balance measurement of ET and modeled evaporation (T = ET meas. ‐ Emodeled).

A two‐column experiment was designed to measure the primary and cross‐coupling transport coefficients for the simultaneous flow of heat and moisture through a medium sand (Plainfield sand) in the liquid‐dominated regime of moisture flow. Values for the moisture diffusivity and thermal conductivity were obtained in the range from 20% to 90% of saturation. Thermal conductivity values fell within ± 7% of a curve calculated by the method of de Vries. In addition, the coefficient for transport of heat under gradients of moisture content was found to be zero within the accuracy of the experiment. The experiment was not sufficiently accurate to yield quantitative estimates of the coefficient for transport of moisture by temperature gradients, so a different experiment was performed to measure this coefficient indirectly. In this latter experiment the dependence of soil water potential upon temperature was found to be as much as five times larger than predicted by the surface tension model used by Philip and de Vries. These large values suggest that thermal effects should be considered for long‐term movement of soil moisture, and perhaps in certain circumstances for shorter term simulations in moist soil.

A model based on papers by D. Kirkham (1949; 1958) is proposed for calculating solute travel times and effluent water quality for tile‐drained soil profiles for cases of ponded and unsaturated surface water input. Variables required for utilization of the model are drain spacing, depth of tile, depth to impermeable zone, soil porosity, and mean discharge rate over the time of study. By using dimensionless variables a characteristic travel‐time parameter is introduced which will represent a system and allow the model calculations to be summarized on a single graph. Simulations are run for cases of miscible displacement, step‐function surface solute input, and periodic surface solute input; differences between ponded and unsaturated leaching are discussed for tile systems.