Rusteau-1D Help
Theory and equations

Pedotransfer functions (Rosetta)

Pedotransfer functions (PTFs) predict soil hydraulic parameters from more easily measured soil properties. Rosetta Schaap et al. (2001) implements hierarchical PTFs: five models with increasing input – textural class (H1), sand, silt and clay (H2), plus bulk density (H3), plus the water content at 33 kPa (H4), plus the water content at 1500 kPa (H5) – each predicting the van Genuchten parameters θr, θs, α, n and Ks.

Method

Each model is an ensemble of artificial neural networks trained on bootstrap samples of a database of soils with measured retention curves and conductivities. The prediction is the mean of the ensemble; the spread of the ensemble gives the standard deviation, an estimate of the uncertainty of the prediction. Model H1 is a look-up table of the class averages.

VersionTraining and features
Rosetta 160 (retention) and 100 (Ks) bootstrap networks; the models of Rosetta Lite in HYDRUS-1D Schaap et al. (2001)
Rosetta 2Rosetta 1 with corrected retention parameters Schaap et al. (2004)
Rosetta 3one network for all parameters, weighted recalibration with the uncertainty of the fitted retention curves, 1000 bootstrap members Zhang and Schaap (2017)

α, n and Ks are predicted as logarithms; their ensemble means are either the mean of the back-transformed values (arithmetic) or 10 to the mean of the logarithms (geometric). Rosetta also predicts the parameters K0 and L of a conductivity function with a matching point Schaap and Leij (2000); Rusteau shows them for information.

The network weights are those of the Python package rosetta-soil; the unit tests of Rusteau reproduce its predictions within 2 × 10−5.

Limitations

The predictions are only as good as the similarity of the soil to the training data (mostly temperate agricultural soils from North America and Europe). Measured data or a parameter estimation are preferable when available; Rosetta predictions are good initial estimates for it.

See also: the Rosetta dialog · soil catalog Carsel and Parrish (1988)

Literature cited in this topic

  • Carsel, R. F., and R. S. Parrish (1988). Developing joint probability distributions of soil water retention characteristics. Water Resources Research 24(5): 755–769. doi:10.1029/WR024i005p00755
  • Schaap, M. G., A. Nemes, and M. Th. van Genuchten (2004). Comparison of models for indirect estimation of water retention and available water in surface soils. Vadose Zone Journal 3(4): 1455–1463. doi:10.2136/vzj2004.1455
  • Schaap, M. G., and F. J. Leij (2000). Improved prediction of unsaturated hydraulic conductivity with the Mualem–van Genuchten model. Soil Science Society of America Journal 64(3): 843–851. doi:10.2136/sssaj2000.643843x
  • Schaap, M. G., F. J. Leij, and M. Th. van Genuchten (2001). Rosetta: a computer program for estimating soil hydraulic parameters with hierarchical pedotransfer functions. Journal of Hydrology 251(3–4): 163–176. doi:10.1016/S0022-1694(01)00466-8
  • USDA-ARS U.S. Salinity Laboratory. rosetta-soil, version 0.3.2: Python implementation of the Rosetta pedotransfer functions (CC0 1.0). https://github.com/usda-ars-ussl/rosetta-soil
  • Zhang, Y., and M. G. Schaap (2017). Weighted recalibration of the Rosetta pedotransfer model with improved estimates of hydraulic parameter distributions and summary statistics (Rosetta3). Journal of Hydrology 547: 39–53. doi:10.1016/j.jhydrol.2017.01.004

All sources: References.