Derivation of Pedotransfer functions to estimate saturated hydraulic conductivity
M. Navabian1, A.M. Liaghat1 and M. Homaee2
1.Dept. of Irrigation and Reclamation Eng., University of Tehran, Iran; 2. Dept. of Soil Science,
University of Tarbiat Modarres, Tehran, Iran. E-mail: Mhomaee@hotmail.com.
Determination of the saturated hydraulic conductivity Ks is needed for many studies and
applications related to irrigation, drainage, water movement and solute transport in the soil.
Although many advances are made for direct measurements of Ks, they are usually time consuming
and costly. Some attempts have been made to indirectly predict the saturated hydraulic conductivity
from the easily/readily obtainable parameters mainly for disturbed and un-structured soils. The
objective of this study was to investigate the possibility of derivation of some Pedotransfer
Functions PTFs and neural network models for different soil textures, using minimum input soil
parameters. Consequently, 45 soil samples with lagely different textures were taken and their in-situ
saturated hydraulic conductivity were measured, using the Geulph permeameter apparatus. The
multilinear regression method was used to develop the pedo-transfer functions. The measurements
were consisted of bulk density, particle density, field capacity, effective porosity (θe), particle size
distributions, effective particle diameters, geometric mean particle diameter (dg) and geometric
standard deviation of soil particles ( g). The Normality and multicolinearity of the parameters were
tested by MiniTab package. Three PTF functions and three neural network models were then
derived, using different input parameters. The preliminary results indicated that accuracy of the
derived PTF functions could well predict (R2=0.73) the Ks from θe, dg and g. Also compares to the
derived pedotransfer functions, the established neural network models could better predict the
saturated hydraulic conductivity with the same input parameters (θe, dg and g).
Key words: neural network, pedotransfer functions, saturated hydraulic conductivity