SYMPOSIA PAPER Published: 01 January 1996

Geostatistical Estimation: Kriging


Geostatistics offers a variety of spatial estimation procedures which are known as kriging. These techniques are commonly used for interpolation of point values at unsampled locations and estimation of average block values. Kriging techniques provide a measure of accuracy in the form of an estimation variance. These estimates are dependent on the model of spatial variability and the relative geometry of measured and estimated locations. Ordinary kriging is a linear minimum-variance interpolator that assumes a constant, but unknown global mean. Other forms of linear kriging includes simple and universal kriging, as well as co-kriging. If measured data display non-Gaussian tendencies, more accurate interpolation may be obtained through non-linear kriging techniques, such as lognormal and indicator kriging.

Author Information

Rouhani, S
School of Civil and Environmental Engineering, Georgia Institute of Technology, Atlanta, GA
Price: $25.00
Contact Sales
Reprints and Permissions
Reprints and copyright permissions can be requested through the
Copyright Clearance Center
Developed by Committee: D18
Pages: 20–31
DOI: 10.1520/STP16111S
ISBN-EB: 978-0-8031-5337-0
ISBN-13: 978-0-8031-2414-1