Skip to main content
ARS Home » Research » Publications at this Location » Publication #224590
Title: A review of the Rawls et al. (1982) soil hydraulic pedotransfer function

Author
item NEMES, ATTILA - UNIV OF MD,COLLEGE PK,MD
item Timlin, Dennis
item QUEBEDEAUX, BRUNO - UNIV OF MD,COLLEGE PK,MD

Submitted to: Soil and Water Conservation Society Proceedings
Publication Type: Abstract Only
Publication Acceptance Date: 3/13/2008
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: For many applications that involve the use of environmental simulation models, soil water retention and hydraulic conductivity data are not available and therefore need to be estimated. The current version of the APEX model – that is being used in the national scale CEAP project to evaluate on-site benefits of conservation practices on cultivated croplands uses Rawls et al. (1982) pedotransfer functions (PTFs) to estimate soil water retention. This linear regression based PTF has widely been considered to be valid and applicable for US soil conditions. In order to test the adequacy of the Rawls et al. (1982) PTF and identify inconsistencies in the estimates, we employed a specific performance measure and a non-parametric k-Nearest Neighbor (k-NN) pattern recognition technique. Data from Rawls' original publication and an independent second data set were used in the evaluation. We found that this PTF delivers sub-optimal and biased estimates for US conditions. Matching the data with the original sources revealed that this result is mainly due to incorrect conversion of organic matter/carbon data and volumetric/gravimetric soil water contents into common units, misrepresentation of organic matter data when those were in fact missing, and uneven representation of soil regions in the data collection. Correcting these significantly reduced estimation bias for the independent data set. Bias was further reduced by using the k-NN technique that does not generalize data relationships for the entire data space in the form of pre-developed equations. The alternative estimation technique and an improved data set are suggested for future use.