Identification of the Best Infiltration Model in Order to Investigation of Spatial Variability of Infiltration Parameters (Case Study: Darabkola River Basin)

Abstract

Infiltration is the process of water penetration from the ground surface into the soil and is an important process  in  the  hydrological  cycle  by  which  surface  runoff and groundwater recharge can be linked. Over the years, the importance of the infiltration process resulted in the development of several simplified analytical models for predicting infiltration. These infiltration  models  range  from  entirely  empirical  to  physically based  models.  The most serious problem associated with infiltration modelling of a catchment is how to express the spatial soil variability. In the present study, various infiltration models were fitted to the observed infiltration data of 27 double ring infiltrometer tests and the best-fit infiltration model  for  Darabkola  watershed was  identified  and  evaluated.  In  addition,  the  spatial variability of the selected infiltration model parameters was analyzed using the geostatistical techniques.  Results  showed  that  among  of  four models,  the  Green  -  Ampt  model  could determine the infiltration rate with smallest values of RMSE. Hence, saturated hydraulic conductivity parameter (KS) and suction head at the wetting front (Sw) were estimated for all  the  test  points.  Evaluation  of  spatial  variability  of  these  parameters  indicated   that parameters KS and Sw had the spatial dependencies of 0.49 and 0.25 respectively, showing medium spatial dependencies  of  both  parameters.  Also,  investigation  of  interpolation parameter maps showed that in the upland with forest land use, relative to other areas in the watershed, the saturated hydraulic conductivity (KS) and suction head at the wetting front (Sw) have larger (1.57-2.69 cm/hr) and smaller values (12.12-1737 cm), respectively.