TAMANHOS ELEMENTARES REPRESENTATIVOS DE ATRIBUTOS DO SOLO VIA ATENUAÇÃO DE RAIOS GAMA E TOMOGRAFIA COMPUTADORIZADA

Abstract

In this study, the Computed Tomography (CT) and gamma-ray attenuation (GRA) techniques were used in the investigation of representative sample sizes for attributes related to soil structure. First of all, the representative elementary length (REL) for experimental measurements of soil mass attenuation coefficient (μes), of samples from a sandy and a clayey soil, was analyzed. The study was conducted with two radioactive sources (241Am and 137Cs), three collimators (2–4 mm diameters), and 14 thickness (x) samples (2–15 cm). From these analyzes, it was possible to identify an ideal thickness range for each of the studied sources (2-4 cm and 12-15 cm for the sources of 241Am and 137Cs, respectively). The application of such results in representative elementary area evaluations, in clayey soil clods via CT, indicated that experimental soil mass attenuation coefficient average values obtained for x>4 cm and source 241Am might induce the use of samples which are not large enough for soil bulk density evaluations. Subsequently, μCT images with a total volume of 39×39×33 mm3 and spatial resolution of 60 μm were used for macroporous system morphological characterization of a Rhodic Ferralsol with clayey texture, under no-till (NT) and conventional till (CT) systems. Attributes as macroporosity (MAP), number of macropores (NMAP), tortuosity (τ) and connectivity (C) of the pores were assessed. The C degree was estimated based on the Euler-Poincaré characteristic (EPC). Once 3D images enable the study of these attributes in different sample volumes, the proposed study is ideal for the analysis of representative elementary volume (REV). Usually, the selection of subvolumes for REV analysis occurs concentrically to a small volume or in adjacent positions. Here, we introduced a new method for selecting the positions of subvolumes, which are randomly chosen within the total image volume (random selection). It was observed that higher fluctuations in amplitude of each analyzed property values occurred for smaller subvolumes and random selection. When these positions are randomly selected, all regions of the sample have the same probability of occurrence. Thus, the subvolumes can be selected from regions that exhibit similar characteristics, or which are quite different from each other. The results allowed the study of each of these attributes in function of the analyzed sample volume. Such analysis enabled the identification of a REV for MAP and τ measurements for both soils. However, NMAP and CEP showed no indication of stabilization within the total sample volume analyzed (39×39×33 mm3). From this study, it was possible to detect the influence of each management systems (NT and CT) on the macroporous system. In general, results obtained for soil under PC showed higher regularity, which is probably due to its homogenization induced by the plowing, conducted twice a year.