| heal.abstract |
The phenomenon of hysteresis in hydraulic properties of porous media plays a significant role in the establishment of the soil water regime. Although a number of methodologies of various degrees of accuracy and subtlety have been developed for the determination of the hysteretic nature of these properties, nonetheless limited research work has been done in investigating the effect of hysteresis on soil water dynamics simulation. The present work has a twofold objective. One is to use existing methodologies in obtaining the wetting and drying branches of the hydraulic conductivity (K) versus water content (Θ) or water pressure head (h), and the other is to investigate the effect which the obtained hysteretic K(h) has on the soil moisture profiles in a simulated soil water infiltration process under conditions of flooding. The investigation was performed for two different porous materials. Three methods have been employed for the determination of K. All of them rely on the experimental determination of the moisture retention curve (MRC). These methods are: 1. The conceptual model of Mualem which uses the MRC to derive K (both wetting and drying branches). 2. An imbibition experiment to determine the relationship of the soil water diffusivity D(Θ), which is subsequently used together with the MRC (its main wetting branch) to derive the wetting branch of K. 3. One-step outflow experiment to determine D(Θ), which is subsequently used together with the MRC (its main drying branch) to derive the drying branch of K. After the determination of the hydraulic conductivity as a function of h, the problem of one-dimensional vertical infiltration under conditions of flooding was investigated by solving the Richard's equation numerically, using a finite difference discretization scheme. Four runs were performed for each of the porous materials with four different combinations of Θ(h) and K(h), using each time the pairs of Θ(h) and K(h) main branches. Some useful results are obtained. From our investigation it is shown that using the conventional approach according to which infiltration is simulated by using the drying K(h) and Θ(h) branches may be erroneous. Indeed, as infiltration is a wetting process it is more appropriate the use of the wetting branches K(h) and Θ(h). However, the wetting branches are not usually employed in the conventional approach since they are more difficult and time consuming to measure. The results indicated that the use of the drying branches of K(h) and Θ(h) produces soil moisture profiles advancing deeper than those obtained from using the appropriate wetting branches. Copyright © 2006 John Wiley & Sons, Ltd. |
en |