The state of Mato Grosso do Sul (MS) is one of the main Brazilian producers in the agricultural sector. To maintain this status, however, it will be necessary both to know and use rationally its soil resources. This way, it will be necessary methods of soil research that are efficient and fast in obtain information, as well as of low-cost to support land use planning. Since relief is an important factor of soil formation, aerial photos and satellite images analysis can be used to detect landscape features that help to characterize and discriminate soils. Therefore, the goals of this study were: (a) to identify different soil classes and verify their relationships with landscape aspects by the interpretation of aerial photos; and (b) to use radiometric data, obtained from satellite images analysis, to discriminate soils or groups of soils on landscape. To do so, it was evaluated the relationship between soil classes and relief and drainage aspects, obtained by the interpretation of aerial photos in conjunction with spectral data obtained by satellite image analysis of 14 circular samples (CS) of the Maracaju city in the MS state. The drainage density (Dd), determined in the CS, showed positive correlation with the soil Ki index and base saturation (V%) index, thus allowing discriminate soil classes of the studied area with 85.7 % of certainty, whereas the spectral data only discriminated soils by the textural classes of their surface layer. On the other hand, it was also observed that soils with Fe2O3 content higher than 180 g kg-1 had their textural classes poorly differentiated by using the spectral data.
landform; morphometry; satellite images; reflectance
