ABSTRACT Notwithstanding the importance of soil surveys, advances in digital soil mapping have mainly focused on mapping soil attributes or properties rather than developing digital maps of soil units or soil classes. The purpose of this research was to develop digital soil unit maps based on primary soil data collection in areas without previously collected soil information. The covariate variability, the random effect across the data subset and the map outputs were the focuses of this study. We used five datasets with four models (Random Forest - RF, Gradient Boosted Machine - GBM, C5.0, and multinomial log-linear model - MLR). The covariates were grouped into five datasets, where four were grouped by Region Of Interest per Class (ROIC) and one was not grouped by ROIC. To evaluate the random effect to split the dataset, we ran each model 50 times and observed the overall accuracy (OA) and kappa index, and uncertainty, majority and variety maps. The OA of Dataset01 to 04 was lower than to Dataset05 accuracy. However, map outputs of RF and GBM for Dataset01 and Dataset05 had the same majority prediction. It seems that RF and GBM produce consistent results in map outputs according to this methodology and pedologist expertise. To evaluate the uncertainty and the consistency of soil unit prediction, we used the majority maps process. Random Forest, similar to GBM, presented the best results. The increase in the number of covariates was not a guarantee of improvement in the OA or in the quality of the map output. Geographic position and distance raster did not improve the map output according to expert evaluation. Because the variance between the ROICs, when the training and validation datasets were split based on it, the subsets are quite different in relation to the covariates, and this is the reason for the worse results of this model, comparing with the Dataset05. On the other hand, when considering one complete dataset not based on ROICs, the variance of training and validation subsets is lower and produced more accurate parameters of quality.

ABSTRACT Some technical issues that are inherent to the mapping of permanent preservation areas (PPAs) on hilltops in Brazil have not yet been properly regulated. This results in disparities in mapping an area depending on the technique used. This paper presents a case study that exposes the influence of the topographic database, mapping scale, and method of measuring the average slope on the results obtained. The total PPAs mapped on hilltops vary significantly as a result of these factors. Subjectivity in applying legislation leads to legal uncertainty, which can be resolved by establishing precise, complementary regulations on the issue. Accordingly, it is proposed that the method of measuring the mean slope and mapping scale should be defined by regulation.

ABSTRACT The objective of this study was to map the levels of degradation of pastures and its association with chemical properties of the soil and relief. We selected a sub-basin located in the municipality of Viçosa, Minas Gerais, and we mapped the degradations through Ikonos images that were submitted for classification. Soil samples were collected and analyzed chemically; the results were transformed in spatial data using cokriging, based on variables of the digital elevation model. We associated the maps (cokriging) with data from exposure of the hillside, levels of solar radiation and levels of the pastures’ degradation. The pastures were mainly classified in moderate (44%) and very strong (30%) degradation levels. The areas with mild degradation tend to increase when there is a higher sum of bases, cationic exchange capacity and organic matter. The portions facing North and the increase of solar radiation showed more advanced stages of degradation.

Este trabalho reporta a utilização de ferramentas de um SIG (Sistema de Informação Geográfica) para o delineamento das geoformas e pedopaisagens em um estudo de caso em microbacia do município de Viçosa (MG), a partir dos dados da carta topográfica (escala 1:50.000) da região. Trabalhos de campo permitiram aferir uma exatidão global de 72 % na classificação das formas da paisagem e correlacionar as classes obtidas com as classes de solos dominantes. Após ajustes, elaborou-se um mapa síntese preliminar de distribuição de pedoformas, que auxilia no mapeamento de solos. O método de classificação das formas da paisagem baseado na geração de um MNT (Modelo Numérico do Terreno) e no cálculo dos atributos do relevo (elevação, declividade, curvatura) mostra-se uma alternativa rápida e econômica em comparação ao delineamento manual a partir da utilização de análise estereoscópica de fotografias aéreas.

This study reports on the use of GIS (Geographical Information System) tools for identifying and mapping landform units and soilscapes in a case study of a watershed in the region of Viçosa, Minas Gerais state, based on a topographical base map at a 1:50000 scale. Field work was undertaken to estimate the degree of accuracy of landform units / soilscape identification, obtaining an agreement of 72 % between the automatic and field work classification, and to correlate obtained soilscapes with dominant soil classes. After adjustments, a preliminary synthesized pedogeomorphic map was produced, which can assist soil mapping. The method of classifying landforms based on a digital elevation model (DEM) as well as on the calculation of landform attributes (elevation, slope angle, slope form) represents a quick and economic option, compared with manual drawing based on stereoscopical identification of landforms from aerial photographs.