Modelagem para predição de classes de solo em escala de município

Abstract

The mapping of soil classes has a fundamental role in social development. Therefore, the estimation of soil classes at the municipal scale and determination of soil classes from reference areas seek to satisfy the needs for quality data and information for adequate scientific and social progress. The methodology of digital soil mapping (MDS) offers a less expensive and more dynamic alternative, modernizing the ways of collecting and presenting data referring to soil classes, in this way, the use of digital classification methods allows the data to be interpreted in a different way. to represent the landscape and describe soil landscape relationships efficiently, attributing good coefficients of precision, accuracy, and uncertainty. It is also understood that the possibilities of updating and reviewing the results generated is an important factor for scientific development, since new tests or models can be applied to the data at any time. Finally, the studies of this thesis seek to advance issues related to the mapping of soil classes at unconventional scales in the current literature, in addition to looking for soil-landscape relationships when exploring reference areas for digital soil mapping.