Improving the Estimation of Soil Cation Exchange Capacity Using Fractal Dimensions

Cation exchange capacity (CEC) is one of the most important soil chemical properties in terms of plant nutrition and pollutants adsorption in soil that its measurement is time-consuming and expensive. Therefore, this study aimed to estimate soil CEC using values of organic matter, soil textural components, and Tyler and Wheatcraft (DT) and Sepaskhah and Tafteh (DS) fractal dimensions and also to investigate the efficiency of mentioned fractal dimensions as an independent variable and its effect on the accuracy of regression relationships to estimate soil CEC. In this study, data from 100 soil samples of UNSODA soil database were used. Soil primary particles size distribution was calculated using the Skaggs approach and fractal dimension of soil primary particles was calculated using the Sepaskhah and Tafteh and Tyler and Wheatcraft approaches. Results showed that the CEC values had significant negative relationship with sand content, and significant positive relationship with logarithm (in base 10) of organic matter, clay, DS and DT values. Values of training and test data determination coefficients, normalized root mean square error (%) and Nash-Sutcliffe coefficient for multivariate regression relationship between CEC versus logarithm (in base 10) of organic matter and clay were respectively equal to 0.77, 0.84, 17.2 and 0.92; between CEC versus logarithm (in base 10) of organic matter and DS were respectively equal to 0.77, 0.85, 17.2 and 0.92 and between CEC versus logarithm (in base 10) of organic matter and DT were respectively equal to 0.77, 0.87, 14.0 and 0.93. Therefore, the most accuracy of regression relationships to estimate CEC obtained when organic matter and DT variables was used as independent variables. In other words, application of DT improved CEC estimation.