Evaluation and Validation of Models for Estimating Oxygen Diffusion Coefficient in Different Soil Texture Classes

The diffusion of gas in the soil, which is usually expressed in terms of its diffusion in the atmosphere (D_p⁄D_0 ), varies based on the characteristics of shape, size, pores distortion, ventilation porosity and soil moisture content (Ɵ), and its direct measurement is generally difficult and time consuming. In this study, while preparing soil samples from different texture classes, oxygen gas diffusion (D_p) was measured in various ventilation porosity (ɛ) and moisture (Ɵ) by non-sustainable methods. Then its changes with ɛ and Ɵ were investigated as a regression equation. The results showed that with increasing ɛ and decreasing the amount of soils Ɵ, D_p⁄D_0 increased and its maximum and minimum values were obtained in low and high humidity (Equal to saturation), respectively. The results of estimating models that presented in resource and obtained regression model in this study (Equation 11) were evaluated and compared using RMSE, GMER and GSDER criteria. The results showed that the estimated D_p⁄D_0 with equation 11 had maximum agreement with the measured data (RMSE = 0.022) and had a minimum overestimation or underestimation (GMER = 1.019), compared with other models that used in this research. Due to the high correlation of the obtained data from this regression equation with the measured data (GSDER = 1.251), the degree of accuracy of this equation is higher than the previous five models and it can be a suitable alternative for them, if it is validated with a wider range of soils.