Submitted to: Agronomy Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: November 2, 2004
Publication Date: November 2, 2004
Citation: Nemes, A. Rawls, W.J., Pachepsky, Y.A. 2004. A nearest neighbor technique to estimate soil water retention [abstract]. Agronomy Abstracts.
Estimation of soil hydraulic properties by predictive equations (mostly termed as pedotransfer functions or PTFs) provides an alternative to the costly ' and thus often impractical ' direct measurements. Different techniques can be found in literature to estimate soil water retention and hydraulic conductivity. In this study we introduce a similarity based 'k-Nearest Neighbor' technique to estimate soil water content at -33 and -1500 kPa matric potentials from texture, bulk density and organic matter content. Model performance is tested against Neural Network models that were developed from the same data sets.