Quantitative techniques for spatial prediction in soil survey are developin
g apace. They generally derive from geostatistics and modern statistics. Th
e recent developments in geostatistics are reviewed particularly with respe
ct to non-linear methods and the use of all types of ancillary information.
Additionally analysis based on non-stationarity of a variable and the use
of ancillary information are demonstrated as encompassing modern regression
techniques, including generalised linear models (GLM), generalised additiv
e models (GAM), classification and regression trees (RT) and neural network
s (NN). Three resolutions of interest are discussed. Case studies are used
to illustrate different pedometric techniques, and a variety of ancillary d
ata. The case studies focus on predicting different soil properties and cla
ssifying soil in an area into soil classes defined a priori. Different tech
niques produced different error of interpolation. Hybrid methods such as CL
ORPT with geostatistics offer powerful spatial prediction methods, especial
ly up to the catchment and regional extent. It is shown that the use of eac
h pedometric technique depends on the purpose of the survey and the accurac
y required of the final product. (C) 2000 Elsevier Science B.V. All rights
reserved.