Empirical Model for Estimation of Soil Permeability Based on Soil Texture and Porosity

Abstract

Soil permeability is the ability of the soil to pass water or air. Soil permeability is affected by texture, structure, and soil porosity. This study aims to develop a mathematical model to predict the value of soil permeability as a function of the percentage of the constituent fraction of the soil and soil permeability as a function of porosity. The study used soil taken from 7 different locations, with 6 samples for each location, 4 samples for model building and 2 samples for model validation. Parameters observed consisted of the percentage of sand (x1), the percentage of silt (x2), the percentage of clay, (x3), soil porosity (x4) and soil permeability (y). From the analysis, the empirical model obtained is soil permeability as a function of the percentage of constituent fractions of the soil which is expressed by the equation y1=36.796-16.022x2-23.938x3 and soil permeability as a function of porosity is expressed by the equation y2=12+0.65(x4-0.06)-2.92 . The permeability equation as a function of soil constituent fraction (y1) can predict soil permeability with a value of R2 = 0.925 and an RRMSE value of 5.461%, better than the permeability equation as a function of porosity. Keywords: Empirical model, Multiple linear regression, RRMSE, Soil physical properties, Model validation