Point estimation of soil moisture characteristics curve using some soil physical and mechanical properties

Soil moisture characteristic curve (SMC) is a fundamental soil property for predicting and modeling water flow and solute transport in the unsaturated soil, but its direct measurement is tedious and time consuming. Therefore, various indirect methods (e. g. , pedotransfer functions, PTFs) have been developed to predict SMC from easily available soil properties (EASP). We develop a procedure to predict SMC from ESAP and soil liquid limit (LL), and plastic limit (PL). Forty three soils were sampled from north-west of Iran. All of soil samples were divided in two groups; 28 and 15 soils samples were used to train and evaluate of the models, respectively. The SMC, (water content at the suctions 0. 1, 0. 2, 0. 3, 0. 4, 1, 2, 3, 5, 10 and 15 bar) and LL, PL and ESAP were measured through standard methods. Multiple linear regression analysis was used to make correlation between LL, PL and ESAP data as independent variables along with SMC data as the dependent variable, using the SPSS software and the stepwise algorithm. Results showed that among all measured soil properties, the clay content, bulk density, LL and PL had high correlation with the soil moisture content at different suction heads. Values of the coefficient of determination (89%) and root mean square error (0. 028), obtained by the statistical analysis, indicated the validity of the models in the all of the suction heads.