Skip to main content
ARS Home » Research » Publications at this Location » Publication #200778
Title: USING ENSEMBLE PREDICTIONS TO SIMULATE FIELD-SCALE SOIL WATER TIME SERIES WITH UPSCALED AND DOWNSCALED SOIL HYDRAULIC PROPERTIES

Author
item Pachepsky, Yakov
item Guber, Andrey
item JACQUES, D - NUCLEAR AGENCY,BELGIUM
item Van Genuchten, Martinus
item SIMUNEK, J - U. OF CA, RIVERSIDE,CA
item NICHOLSON, T - USNRC
item CADY, R - USNRC

Submitted to: American Geophysical Union
Publication Type: Abstract Only
Publication Acceptance Date: 9/18/2006
Publication Date: 12/11/2006
Citation: Pachepsky, Y.A., Guber, A.K., Jacques, D., Van Genuchten, M.T., Simunek, J., Nicholson, T.J., Cady, R.E. 2006. Using ensemble predictions to simulate field-scale soil water time series with upscaled and downscaled soil hydraulic properties. American Geophysical Union, December 11-15, 2006, San Francisco, CA. 1340h:H33A-1482.

Interpretive Summary:

Technical Abstract: Simulations of soil water flow require measurements of soil hydraulic properties which are particularly difficult at field scale. Laboratory measurements provide hydraulic properties at scales finer than the field scale, whereas pedotransfer functions (PTFs) integrate information on hydraulic properties at larger scales. One way of downscaling large-scale data is to use an ensemble of PTFs to generate hydraulic properties with each of the PTFs and to obtain the ensemble prediction of soil water flow by averaging the simulations results obtained with individual PTFs. Alternatively, the prediction of field-scale soil water flow can be made with the average hydraulic properties obtained from the PTF ensemble. Similarly, the upscaling can be done using an ensemble of hydraulic properties along individual soil profiles to obtain ensemble prediction of soil water flow by averaging simulations results obtained for the individual profiles. Alternatively, the prediction of field-scale soil water flow can be made with average hydraulic properties obtained from the individual soil profiles. The objective of this work was to evaluate the two upscaling approaches and two downscaling approaches with field monitoring data on soil water contents. Data on soil water contents and pressure heads in 60 points at 5 depths over a 6-m long transect in layered loamy soil were collected during an extremely wet year in Belgium. Water retention was measured in laboratory in samples taken from 60 profiles at three depths. An ensemble of 20 PTFs developed from large datasets in different regions of the World was used to compute soil hydraulic properties at each depth. The HYDRUS-1D software was used to solve the Richards equation of water flow in soil. Ensemble predictions were evaluated against the layer-averaged monitoring data on soil water contents over 384 days. The ensemble of PTFs represented field water retention better than the laboratory data. PTF ensemble predictions were about two times more accurate compared with ensemble predictions with the laboratory data regardless of which of the two upscaling and two downscaling approaches has been used. The ensemble predictions were also made using weighted averages with weights derived from hindcasts. The accuracy generally improved, but the improvement depended on soil water regime during the hindcast periods. Overall, using PTF ensembles appeared to be viable method to downscale soil hydraulic properties for field-scale simulations purpose. Ensemble upscaling of the laboratory data resulted in less accurate simulations as compared with downscaling PTF estimates.