Abstract

Soil specific surface area (SSA) is an important soil property. SSA can be estimated from soil textural data or soil particle-size distribution. In this study, 20 soils with appropriate combination of texture were selected from Fars province, south of Iran. For each soil sample the values of SSA and percentages of clay, silt and sand were measured. Also, soil particle-size distribution curve of each soil was estimated with an existing modified model, and then the summation of the number of spherical particles for whole parts of the soil particle-size distribution (N) was determined. Furthermore, the geometric mean particle-size diameter (d_g) of each soil was determined. Then, two power equations based on d_gand N were calibrated to estimate SSA, and then the derived equations were validated for two independent soil data sets including 64 soils. Root mean square error (RMSE) was used to evaluate the obtained results in calibration and validation stages. The RMSE values of power equations based on d_g and N in calibration stage equaled 76.3 and 91.3, respectively, and in validation stage equaled 97.4 and 57.1, respectively. Therefore, the power equation based on d_g was better than the power equation based on N for estimating SSA in calibration stage; however, the power equation based on N was better than the power equation based on d_g for estimating SSA in validation stage.

Key words: Soil specific surface area, geometric mean particle-size diameter, summation of the number of spherical particles