Multivariate Analysis and Visualization of Soil Data.

To evidence the multidimensionality of the soil quality concept, we propose the use of data visualization as a tool for exploratory data analyses, model building, and diagnostics. Multidimensional data visualization is used in several disciplines, including the natural sciences, geography, computer sciences, and statistics, but their use in soil science has been limited. The use of interactive data visualization aids in model development and diagnoses and can become an invaluable tool for the advancement of our understanding of soil processes and effects of management practices. Multivariate methods have been used to identify factors that are most discriminating between land or management categories. Yet a need exists to further the use of multivariate statistical methods in conjunction with multidimensional data visualization for a more comprehensive evaluation of cropping systems.