Evaluation of infiltration models with different numbers of adjustment parameters in an oxisol with application of wastewater at the different concentrations

Abstract

The effects of the concentration of total suspended solids from municipal wastewater on the water infiltration process in a clayey soil were investigated as well as the suitability of infiltration models with different numbers of parameters. The experiment consisted of columns of soil of 20 cm long in which supply water and synthetic municipal wastewater with different concentrations of solids were applied. The soil was characterized with respect to particle size, porosity, soil and particle density, clay water-dispersed clay, and infiltration capacity. The accuracy of the Kostiakov, Kostiakov-Lewis, Horton, Phillip, Green-Ampt, and Swartzendruber infiltration models was assessed by using the statistical criteria: mean error (MAE), mean squared error (RMSE), determination coefficient (R2), index of agreement (d), and Akaike information corrected criterion (AICc). The results indicate that the application of wastewater rich in suspended solids reduces the macroporosity of the soil, due to the clogging of the pores, causing a reduction in the hydraulic conductivity of the saturated soil and the cumulative infiltration. Based upon the AICc, the Kostiakov-Lewis models proved to be the most appropriate, except for the treatment with the highest concentration of solids, in which the Swartzendruber model presented the lowest AICc. In general, the other statistical criteria corroborate this conclusion.