Predicting the Soil Moisture Characteristic Curve from Particle Size Distribution with a Simple Conceptual Model

Mohammad Hossein Mohammadi,Marnik Vanclooster
DOI: https://doi.org/10.2136/vzj2010.0080
2011-05-01
Vadose Zone Journal
Abstract:Indirect methods for predicting the soil moisture characteristic curve (SMC) from the particle size distribution (PSD) often rely on empirical coefficients, which limits their applicability to independent data sets. We have developed a robust simple PSD‐based conceptual SMC prediction model and evaluated the model performance through comparisons with the Haverkamp and Parlange (HP) and Arya and Paris (AP) models. Following the AP model, we divided the PSD into n size fractions where each fraction contained spherical particles whose packing state is described by a parameter. The moisture content is subsequently calculated from the PSD and measured saturated moisture content. The packing state is estimated from particle and bulk densities. The suction head is predicted based on the particle size, assuming a linear relationship between the suction head and packing state. Our results showed that the model can adequately predict the SMC as measured in 80 soils selected from the UNSODA database. It was also shown that the proposed model provides better predictions of SMC than the AP or HP models. The model underestimates the moisture content in the dry range of the SMC. We attribute this bias to the incomplete desorption of residual water coated on soil particles or water retained within nonspherical particles with high surface energy contents. We conclude that the main advantage of our model is the robustness and independency of model performance on soil type, allowing improving predictions of SMC from PSD at larger watershed scales.
environmental sciences,water resources,soil science
What problem does this paper attempt to address?