Call Prefix : PR-T

Call Suffix : 1986

Call Cutter : T - SoSc 11

Location : SEARCA Library

Format : TD

Publisher : University of the Philippines Los Banos,

Publication Date : May 1986

Publication Place : Laguna, Philippines :

Summary : Geostatistical techniques, specifically the semivariogram and kriging, were applied to characterize the spatial variability of soil properties. Data on percent sand, percent silt, percent clay, percent organic matter, ppm potassium (K) and ppm phosphorus (P) from soil samples collected in regular grid sampling scheme covering about 9,000 hectares of sugarland in Negros Occidental was used in this study. For each soil property, a semivariogram was calculated. A spherical model equation was found fitted to each semivariogram and the parametric coefficients of the equation were identified as nugget variance, sill, and range. The structure of the semivariograms revealed a strong degree of spatial dependency of the measured soil properties. However, ppm K and ppm P tended to have value of nugget variance approaching the value sill. This indicated the tendency of these properties to vary considerably at shorter distances. The sampling distance of 300 meters was optimal. The fitted spherical model equation of each semivariogram was used in the kriging process or in interpolating unknown values of kriged values. A kriged value is a weighted moving average value of known neighborhood values. The precesion of the kriged value was tested by a self-validation technique known as "jackknifing." Jackknifing showed insignificant differences between kriged and measured values of the same point using T-test at five percent level, indicating that the kriged values were optimal. Using the measured and kriged values, contour maps for percent organic matter, percent sand, percent silt, percent clay, and ppm K of the topsoil were drawn.