Algorithms to model soil water retention are needed to study the response of vegetation and hydrologic systems to climate change. The objective of this study was to evaluate some soil water retention models to identify minimum input data requirements. Six models that function with various combinations of particle-size distribution, bulk density (ρb), and soil organic matter data were tested using data for nearly 6000 pedons. The Rawls model, which requires particle-size distribution and organic matter data, had the lowest overall absolute value of the mean error (ME) with 0.020, 0.001, and 0.007 m3 H2O m−3 soil for matric soil water pressures of −10, −33, and −1500 kPa, respectively. The Saxton model, which requires particle-size distribution data, had small MEs (0.018 and 0.007 m3 H2O m−3 soil) for −10 and −1500 kPa matric soil water pressures, and a moderately small ME (0.017 m3 H2O m−3 soil) at −33 kPa. The Vereecken model, which requires ρb, particle-size distribution, and organic matter data, had small MEs (0.016 and 0.009 m3 H2O m−3 soil) at matric soil water pressures of −10 and −33 kPa, with a larger ME (0.020 m3 H2O m−3 soil) at −1500 kPa. The remaining three models had relatively large MEs for at least two of the three matric soil water pressures. For estimating water-holding capacity only, the Saxton model is adequate. The Rawls model is recommended for characterizing the relationship of water content to matric soil water pressure.
The information in this document has been wholly funded by the U.S. Environmental Protection Agency (EPA) under Contract 68-C8-006 to ManTech Environmental Technology, Inc. It has been subjected to the agency's peer and administrative review and it has been approved as an EPA document.