برآورد منحني مشخصه رطوبتي خاك با استفاده از توابع انتقالي

Predicting the Soil Moisture Characteristics Curve by Using Pedotransfer Functions

Soil water retention curve is needed to describe the availability of soil water to plant, and to study modeling of water and solute transport in unsaturated soils. The direct measurement of this curve is time-consuming ,difficult and costly. Recently, many attempts have been made to predict water retention curve indirectly from soil physical and chemical properties. Pedotransfer functions (PTFs) comprise the indirect methods. Predicting water retention curve and Van Genuchten equation parameters by using pedotransfer functions were the main research objectives. In current study fourty loamy soil samples (35 for prediction and 5 for validation) were randomly collected form Karaj area. Particle size distribution, bulk density, calcium carbonate equivalent and organic carbon as independent variables, were determined by the hydrometer ,clod , acid neutralization and Walkly Black methods, respectively. Water retention curve for the soils were obtained, experimentally using pressure plates. The variables were separated in two groups: (1) particle size distribution, bulk density, organic carbon and calcium carbonate percentage. (2) geometric mean and geometric standard deviation of particle size, bulk density, organic carbon and calcium carbonate percentage. The most optimum combination of independent variables for estimating soil water retention curve and Van Genuchten equation parameters were selected by the regression method. The regression equations for two independent variable groups were obtained using multiple linear regression and the PTFs were compared. The results indicate that there is a significant relationship between measured and predicted values. The relationship were significant at 0.1% probability level for pointed PTFs, at 1% for n, a and 5% for 0, of Van Genuchten predicting function. Using first group of variables for estimating soil water retention curve and Van genuchten equations parameters was better than the second group. Statistical analysis for the evaluation of PTFs indicated that GMER (Geometric mean error ratio) values were close to 1 and GSDER (Geometric standard deviation of the error ratio) values were small. The results indicate that presented functions were valid.