Comparison of modified CL and DL methods with the main Beerkan infiltration experiment algorithm (BESTslope) for determining saturated hydraulic conductivity

The Beerkan single ring infiltration experiment is a simple and inexpensive method for estimating soil hydraulic properties and is gaining scientist's attention in recent years. The purpose of present study was to compare the performance of modified forms of CL (cumulative linearization) and DL (derivative linearization) methods with the main Beerkan algorithm (i.e. BESTslope) in estimating saturated hydraulic conductivity (Ks) and the value of α (i.e. the ratio of gravity force to the capillary force) in the modified forms of CL and DL. For this purpose a number of 113 Beerkan infiltration experiments were carried out in Sistan Dam' research field. The negative values of Ks estimated by the BESTslope algorithm were eliminated and the 101 remaining experiments were selected for final analysis. According to the results, Ks values estimated by DL and CL were negative in most Beerkan experiments while the modified DL and CL methods resulted in positive values in all cases. The results were compared using a weighted coefficient of determination (ωr2). According to the results, the Ks values estimated by the modified DL were more similar to those obtained by BESTslope algorithm (ωr2 0.8135) when compared to those obtained by the CL method (ωr2=0.7506). Besides, the difference between Ks values estimated by BESTslope algorithm and the modified DL or CL methods decreased as the percentage of clay increased. The calculated value of α in modified DL (α =0.019) was more similar to that reported previous studies (in similar soil texture, i.e. loam and sandy loam) than that obtained in CL model (α =0.023).