Sen Lu1, Tusheng Ren2, Yuanshi Gong1, and Robert Horton3. (1) China Agricultural University, No 2, Yuan Ming Yuan Xi Lu, China Agricultural University, Beijing, 100094, China, (2) No 2 Yuan Ming Yuan Xi Lu, China Agricultural University, Department of Soil & Water, China Agricultural University, Beijing, 100094, China, (3) Iowa State University, 2543 Agronomy Hall, Iowa State University, Ames, IA 50011
Several models have been proposed to describe the soil water retention curve (SWRC) from saturation to oven dryness from measurements of limited matric suction range. The development and validation of these models are mostly based on the measurements by Campbell and Shiozawa (1992), and have not been validated by independent data from other soils. In this study, the performances of three extrapolative models were tested on eight soils using soil water retention measurements from saturation to oven dryness. When the model parameters were obtained from water retention data in 0-1500 kPa range, the Root Mean Square Error (RMSE) was generally with 0.01 for all the three models. The RMSE of the Fayer and Simmons (1995) model was relative larger than that of the Webb (2000) model and Khlosi et al. (2006) model. In addition, the Fayer and Simmons (1995) model was sensitive to the data point near the matric suction of 1500 kPa. The Khlosi et al. (2006) model produced reliable results when reduced data sets from 0-300 kPa were used for model establishment, but larger errors were observed on some soils if the measured data was reduced to the 0-100 kPa range. We conclude that when soil water retention measurements in the range of 0-1500 kPa are used, the three models are capable of giving accurate information for the entire water content range. Soil water retention measurements up to suction of 300 kPa are required if the Khlosi et al. (2006) model is applied to establish soil water retention function from saturation to oven dryness.