Multivariate Analysis in the Characterization and Classification of Soils

Abstract

Soil is a fundamental natural resource in the balance for the ecosystems as well as for agriculture, food, and housing. The soil is very susceptible to changes in its structure due to contamination or degradation of anthropogenic origin. Therefore, its evaluation, whether for environmental purposes or as an agricultural or housing resource, must be carried out in depth. This evaluation comprises the analysis of multiple physical, physicochemical and chemical-biological parameters. However, due to these multiple parameters, the use of multivariate statistical methods becomes necessary. In this chapter, the soil data analysis was performed by the method of Principal Components Analysis for a reduction of dimensions and, to carry out a better interpretation of results. This method was applied to carry out a characterization and classification of soil samples. The analysis was performed with data obtained from soil samples from the Bolivian Altiplano. The results show the potential of the principal component of the method in processing data.