Artificial Neural Networks for Estimating Soil Water Retention Curve Using Fitted and Measured Data

Figure 3

Models diagrams of the ANNs developed in this study. Bulk density (Bd), particle density (Pd), total porosity (), and sand, silt, and clay percentages are the common input predictors of the three models. In ANNs 1 and 2, (for ) are the water content at the matric potentials 0, −10, −20, −30, −50, −100, −200, −500, and −1000 cm, respectively. In ANN 3, the matric potential was considered as an additional predictor for estimating the corresponding water content .