An empirical model for estimating soil penetrometer resistance from relative bulk density, matric potential, and depth

Soil penetrometer resistance (PR), an indicator of soil strength, is often used to denote the force that roots need to exert to penetrate the soil. In this study, we propose an empirical model, which has a simplified form and fewer parameters than the previous ones, to estimate PR from relative bulk density, matric potential of soil water, and soil depth. The model was established by using field measurements of PR, bulk density, water content, and laboratory water retention data at various soil depths in a long-term tillage experiment during maize growing season in 2017. Relative bulk density was determined from soil texture, bulk density, and organic matter content. Model performance was evaluated by comparing the predictions from the new model against those obtained from four earlier models using independent field data collected from four soils of different textures in 2015, 2018 and 2019. The root mean square error of the new model ranged from 0.358 to 0.879 MPa, significantly lower than that of the other four models (0.404-2.689 MPa), indicating that the new model could be applied to estimate PR for a wide range of soil textures with improved accuracy.