Modelling spatial variability of saturated hydraulic conductivity using fourier series analysis

Abstract

A study was conducted to develop a computer model to simulate the spatial variability of field-measured saturated hydraulic conductivity (Ks) using Fourier series analysis. Ks measurements, both in situ and in the laboratory, were made at 150 and 300 mm depths at regular intervals of 4.6 m on two perpendicular transects crossing each other at the centre of a no-till field. Three Fourier series models, with a finite number of harmonics, were developed to simulate the variability in Ks values. These models included the regular Fourier series model, the half-range Fourier series model with only cosine terms, and the half-range Fourier series model with only sine terms. Also, a test of significance was conducted with the first five and eight terms of the Fourier series models. This test indicated that with only five or eight terms model performances were not significantly different from each other at the 0.05 level. The overall results of this study indicated that spatial variability in the Ks measurements made with both techniques, field and laboratory, can be represented successfully with half-range Fourier series with only cosine terms. Statistical analysis of predicted and observed Ks values indicated that the two sets were not significantly different from each other at the 0.05 level. © 1994 Taylor and Francis Group, LLC.