Page Banner

United States Department of Agriculture

Agricultural Research Service

Title: Multimodel and ensemble simulations of water flow in variably saturated soils

Authors
item Pachepsky, Yakov
item Guber, Andrey
item Van Genuchten, Martinus
item Simunek, Jiri - U. OF RIVERSIDE, CA
item Jacques, Diederik - SCK-CEN,MOL, BELGIUM
item Nemes, Attila - U. OF RIVERSIDE, CA
item Nicholson, Thomas - US NRC, ONNR
item Cady, Ralph - US NRC, ONRR

Submitted to: Unsaturated Zone Interest Group (UZIG) Workshop
Publication Type: Abstract Only
Publication Acceptance Date: August 2, 2007
Publication Date: August 27, 2007
Citation: Pachepsky, Y.A., Guber, A.K., Van Genuchten, M.T., Simunek, J., Jacques, D., Nemes, A., Nicholson, T.J., Cady, R.E. 2007. Multimodel and ensemble simulations of water flow in variably saturated soils. Unsaturated Zone Interest Group (UZIG) Workshop, August 27-30, 2007, Los Alamos, New Mexico. p.3.

Technical Abstract: Calibration of variably saturated flow models with field monitoring data is complicated by the strong nonlinearity of the dependency of the unsaturated flow parameters on the water content. Pedotransfer functions (PTFs) are routinely utilized to relate these parameters to readily available data on soil and sediment properties. The PTF accuracy outside of the development region is generally unknown. Combining predictions using various independent models, often called multimodel prediction, has become a popular technique in climate prediction and is now increasingly being used also in subsurface hydrology. The objectives of this work were (a) to compare different methods of multimodel simulation of the field soil water regime using pedotransfer functions, and (b) to see whether the calibration of a flow model with field data can be replaced by multimodel simulations. The multimodel simulation in this work consisted of running the Richards model with outputs of individual PTFs and then combining the obtained outputs. We compared weighing simulation results from individual models by using (1) only the best model, (2) equal weights, (3) the unconstrained superensemble, i. e. regressing measured values to outputs of individual models, (4) singular value decomposition (SVD) in the regression, (5) Bayesian model averaging, and (6) weights derived from the Kullback-Leibler information. A large database of measured soil water contents and soil basic properties at 5 depths along a 6-m transect in a layered loamy soil was utilized to calibrate the Richards flow model and to develop the input for 19 published PTFs for water content time series simulations with HYDRUS1D. The unconstrained superensemble was found to be the worst weighing method characterized by low reliability but high accuracy because of multicollinearity in the regression input. The SVD multimodel was the best method that had the overall accuracy and reliability comparable with the accuracy of the calibrated model. This multimodel had average accuracy and reliability RMSE values of about 0.01 cm3cm-3 at 35 cm depth, and of about 0.005 cm3cm-3 for one month monitoring and 13 months of testing. This indicates that monitoring of the soil water regime in combination with multimodel simulations can be a viable approach to simulating field water flow in the vadose zone.

Last Modified: 10/23/2014
Footer Content Back to Top of Page