Dissertation Defense – Sandra Deodoro
The Postgraduate Program in Analysis and Modeling of Environmental Systems of the Institute of Geosciences of UFMG invites you to the first dissertation defense of the 2019 class:
DISSERTATION DEFENSE
Student: Sandra Cristina Deodoro
Advisor: Prof. Dr. Plínio Temba
Title: “Soil classification and predictive mapping in the Volta Grande region of the Uruguay River – SC/RS (Brazil)”
Date: October 16, 2020
Time: 9:30 a.m.
Link to watch the videoconference: https://conferenciaweb.rnp.br/webconf/sandra-cristina-deodoro
Abstract:
Soil is a natural resource that can be analyzed based on its anthropic and (geo)ecosystemic functions. Knowledge of soil texture (proportion of granulometric fractions) is important in these contexts, especially in the surface layers, which are the first to be eroded. In addition, Texture is an important characteristic due to its relationship with other soil properties such as structure, porosity, permeability, fertility, chemistry and moisture content. There is a growing need for spatially continuous and quantitative soil information for environmental modeling and management, at different cartographic scales. The lack of data sampling is usually compensated by prediction and modeling results whose procedures, known as predictive soil mapping, are specially developed to estimate spatial distribution of soil variables. Digital soil mapping constitutes a useful approach for spatial prediction of soil attributes. Such an approach involves a relationship between soil and environmental variables, based on statistical and geostatistical models, to create predictive maps or derive values of soil properties in unmeasured locations from field data. This dissertation aims to develop a map of the soil surface texture in the Uruguay River basin (border between the states of Santa Catarina and Rio Grande do Sul), in the section known as Volta Grande, using soil spectral data collected from the MSI sensor of the Sentinel-2 satellite, field data (granulometric sampling), predictive statistical modeling (Discriminant Analysis) and IDW interpolation. The results showed that the soil texture was classified with an accuracy of 71% according to the Kappa Index, with a predominance of clay. Supported by morphometric data and the MRVBF Index – Multiresolution Index of Valley Bottom Flatness –, derived from the SRTM GL1 DEM (12.5 m), most of the area was represented by clayey colluvium, consistent with field observations and with the extensive slope segments widely distributed. It was concluded that the occurrence of colluvium-alluvium on the banks of the Uruguay River, in the lowland areas, indicates the contribution of the slopes to the pedogeomorphological dynamics of the study area and not only to the fluvial dynamics. Based on the results, the methodology applied in this work demonstrated that remote sensing products and techniques, combined with statistical modeling, have potential utility as auxiliary knowledge and techniques for obtaining, analyzing and mapping soil attributes, such as granulometry. The dissertation is original and involves interdisciplinary concepts of remote sensing and pedogeomorphology, integrated by statistical modeling and geographic information system. As a practical implication, it presents soil as an important natural resource for environmental analysis. By using data from a relatively recent sensor (2015), when compared to others such as Landsat, it presents the potentialities and limitations of this instrument for applications in inferential modeling of soil attributes.9