Basic Kriging Methods in Geostatistics

Abstract

Measurements of environmental, hydrological, agricultural and similar studies are based on point observations over the Earth. Precipitation and temperature values are measured from meteorological stations, soil characteristics are measured from soil samples, and pollution of a lake is measured by taking samples from lake. These are some examples from spatial point measurements. These variables can be measured by taking samples from a limited number of locations or from certain locations. However, it is logically impossible to measure a variable at all parts of globe or on a field of certain size. Instead of this it is possible to make some interpolation to map spatial distributions of that variable. Observation locations which are close to each other tend to have similar values, however the ones located farther apart from each other differ more. So this knowledge is used in prediction procedure (interpolation). Kriging which will be described here, is an interpolation method. Kriging makes optimal predictions: it provides the most likely value at any location of a variable. Methodologies of most commonly used kriging methods in geostatistics; Ordinary kriging, Regression kriging and Universal kriging have been described in this review work. © 2017, Centenary University. All rights reserved.