Performance Evaluation of Scaling Methods of Richard's Equation in Infiltration Modeling in a Watershed (Case Study: Marghmalek Watershed)

Spatial variability of soils makes difficult analysis of soil water flow phenomena especially in a large area such as a watershed. Using scaling methods is a solution in variability problems. The objective of this study was to investigate the effect of the non-linear variability on performance of the scaling methods of Richards’ equation for modeling infiltration in a watershed. The method of Warrick et al. by adopting van Genuchten hydraulic functions was used and variability of n values (power of van Genuchten hydraulic functions) was considered as the nonlinear variability. Marghmalek watershed, a sub watershed of Zayanderoud, with 97 Sq. kilometers was studied. In addition, ten virtual watersheds with various degrees of variability of n were evaluated which were generated by stochastic method of Monte Carlo. Using HYDRUS-1D model, original and scaled Richards’ equations were solved for infiltration condition with constant hydraulic head and uniform initial soil water content. The results indicated that coefficient of variations of n values in the Marghmalek watershed (equal to 2.57%) is small enough that the scaling method can be used efficiently in modeling infiltration. Therefore, in this watershed, generalized solutions of Richards’ equation can be adequately used instead of individual solutions for every points of the watershed. Evaluations in the virtual watersheds indicated that variability of n values considerably affect the error between the generalized and individual solutions. Based on the result of this study, it can be concluded that scaling methods of Richards’ equation can be adequately applied in the watersheds in which coefficient of variations of n values does not exceed 3%.