Quantitative parameterization of soil surface structure with increasing rainfall volumes

• Ano de defesa: 2016
• Autor(a) principal: Mome Filho, Edison Aparecido
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: Erosão pela água; Escala; Microrelevo; Microtopography; Porous system; Scale; Sistema poroso; Water erosiom
• Link de acesso: http://www.teses.usp.br/teses/disponiveis/11/11140/tde-30092016-095436/

The study of soil structure allows inferences on soil behavior. Quantitative parameters are oftentimes required to describe soil structure and the multifractal ones are still underused in soil science. Some studies have shown relations between the multifractal spectrum and both soil surface roughness decay by rainfall and porous system heterogeneity, however, a particular multifractal response to a specific soil behavior is not established yet. Therefore, the objectives of this research were: (i) to establish relations between multifractal parameters and soil structure changes by analyzing both soil surface roughness maps and 2D images from impregnated soil blocks; and (ii) to utilize these parameters to evaluate soil surface degradation by the processes of crusting and sealing. An experiment with simulated rainfall was assembled on a Fine Rhodic Kandiudalf with an intensity of 120 mm h-1 in quadruplicate plots at amounts of 40, 80, and 120 mm, plus a no-rainfall control. The evolution of the surface roughness was evaluated in three scales of measurement: a field microrelief meter (MRM) gathered readings on a fixed grid (10 x 10 mm, 640,000 mm²); a multistripe laser triangulation (MLT) scanner was used in the laboratory in soil blocks, creating a random mesh (0.5 mm of resolution, 5625 mm²); an X-ray tomography (XRT) scanner gathered readings of a soil block on a fixed grid (0.074 x 0.074 mm, 900 mm²). For micromorphometrical analysis, undisturbed soil samples (0.12 x 0.07 x 0.05 m) were impregnated, sliced in blocks and polished. Each block was divided into three layers (0 to 10 mm, 20 to 30 mm and 40 to 50 mm), parallel to surface, and five images (10X magnification, 156.25 μm2 pixel-1) were taken by layer. After segmentation, three representative images were chosen by layer and the pore system was evaluated. Roughness analyzes showed no differences (p > 0.10) between multifractal parameters across rainfall amounts for MRM measurements, while both MLT and XRT could be used to model roughness degradation by rainfall increase. Since the last two scales presented similar results, MLT could replace XRT in such analysis, due to its lower cost and possibility of a larger area coverage. The multifractal behavior of pores changed according to sealing development and depth of measurement, being sensitive to the changes on size distribution and fragmentation degree (number of pores) within each size class. The Hausdorff dimensions at the left side of the spectrum (Lf(α)min, LΔf(α) and D2) showed a linear behavior with increasing rainfall amount, considering both soil surface roughness and area of pores measurements. However, D2 was not different (p > 0.10) along rainfalls for the porosity closer to surface, although parameters D0-D1, D0-D2 and D1-D2 could be used to described the changes in this layer. Was concluded that the multifractal spectrum is sensitive to structure changes caused by rainfall and that it can be used to parameterize both soil surface and pores degradation.