PERFORMANCE EVALUATION OF THE POPULAR SWRC MODELS AT DIFFERENT SOIL BULK DENSITY RANGES

Abstract

Soil bulk density is an important soil property which influences the hydraulic properties of soil. Soil bulk density has never been a fixed soil property, but it varies spatially due to soil changes and temporally due to human activities, soil management and effects of climate change. This changes in soil bulk density affects the characterization of hydraulic properties of soil especially when mathematical soil retention models are used. Since the measurement of the soil water retention characteristics in the field is expensive, labor intensive and time consuming, mathematical models have been developed to predict these properties. Therefore, it is important to establish the impact of the soil bulk density on the performance of these mathematical models especially their accuracy to fit measured data using measured data. The aim of this study was performance evaluation of the water retention models at different soil bulk density ranges. The study used the dataset downloaded from the global database (ISRIC). The study used Residual Standard Error (RSE) and R2 to measure the accuracy of the model in predicting or quantify the goodness of fit to the experimental data. The low values of RSE shows that experimental data are close to the model predictions. This paper discussed the popular Soil Water Retention Characteristic (SWRC) models then it assessed their performance on varying soil bulk density ranges. The results from the study showed that the performance of the models improved with increase in values of field bulk density. It was established that Gardner and Bi-exponential models were better suited to describe SWRC irrespective of the variation of soil bulk density because they displayed consistent best performance (R2 greater than 0.98) at low and high soil bulk density. From the study, the soil scientist can choose these models to apply especially when the soil bulk density varies greatily.