Start

CV

Research & PhD-students

Teaching & services

Software

Thesis subjects

Projects

SoilGen model

predictive soil mapping

Predictive soil mapping is the mapping of soil types, soil variables or the occurrence of soil features using observations, exhaustive auxiliary information and statistical prediction techniques. The resulting predictive map can be used to guide future sampling or mapping, or can be used as alternative to the traditional soil map.

Here an overview presentation.

The variety of variables (soil types are nominal variables; most soil properties are continuous variables; soil features either occur or not) causes that choices from a wide range of statistical prediction methods must be made. I'm interested in these methods and finding ways to choose the best one for a given application.

The updating of groundwater table dynamics is an example of digital soil mapping of continuous variables. Another example is the prediction of the probability of occurrence of archaeological finds: See this (Finke et al., 2008, JAS)

Text Box: PETER FINKE
Text Box: