Ricardo Marques Coelho

Classification of Brazilian soils by using LIBS and variable selection in the wavelet domain

This paper proposes a novel analytical methodology for soil classification based on the use of laser-induced breakdown spectroscopy (LIBS) and chemometric techniques. In the proposed methodology, linear discriminant analysis (LDA) is employed to build a classification model on the basis of a reduced subset of spectral variables. For the purpose of variable selection, three techniques are considered, namely the successive projection algorithm (SPA), the genetic algorithm (GA), and a stepwise formulation (SW). The use of a data compression procedure in the wavelet domain is also proposed to reduce the computational workload involved in the variable selection process. The methodology is validated in a case study involving the classification of 149 Brazilian soil samples into three different orders (Argissolo, Latossolo and Nitossolo). For means of comparison, soft independent modelling of class analogy (SIMCA) models are also employed. The best discrimination of soil types was attained by SPA-LDA, which achieved an average classification rate of 90% in the validation set and 72% in cross-validation. Moreover, the proposed wavelet compression procedure was found to be of value by providing a 100-fold reduction in computational workload without significantly compromising the classification accuracy of the resulting models.

AVALIAÇÃO DA APTIDÃO AGRÍCOLA DE SOLOS DO ACRE POR DIFERENTES ESPECIALISTAS

Evaluation systems of agricultural land suitability are being used in different regions of Brazil, by distinct teams and with multiple applications. The objective of this study was to evaluate, for soils in the State of Acre, how different experts evaluate the relative importance of indicators of a land suitability system, and how they interpret properties that can define land suitability classes. Ten soil profiles analyzed for the IX Brazilian Soil Classification and Correlation Workshop were used. These soil profiles are located between Rio Branco and Cruzeiro do Sul, municipalities in the Amazon region, with sediments of the Solimoes Formation as soil parent material. The land suitability was interpreted by six experts, based on all land suitability indicators, applied in five degrees of limitation for the factors: fertility deficiency, water deficiency, oxygen deficiency, erosion susceptibility, and impediments to mechanization. From these assessments, the groups and classes of land suitability of each soil profile were established, and the results compared by discriminant analysis. The results indicated the variables nutrient stock, clay content, textural class, drainage class and local relief as the most important to discriminate the environments by the experts. The interpretation of land suitability by the experts was not entirely reproducible for all soils, and was subjected to the diverse experience and knowledge of the evaluators about the environment and different land use types.

Mapeamento pedológico digital da folha Botucatu (SF-22-Z-B-VI-3): treinamento de dados em mapa tradicional e validação de campo

O mapeamento digital de solos permite prever padroes de ocorrencia de solos com base em areas de referencia e no uso de tecnicas de mineracao de dados para modelar associacoes solo-paisagem. Os objetivos deste trabalho foram produzir um mapa pedologico digital por meio de tecnicas de mineracao de dados aplicadas a variaveis geomorfometricas e de geologia, com base em areas de referencia; e testar a confiabilidade desse mapa por meio de validacao em campo com diferentes sistemas de amostragem. O mapeamento foi realizado na folha Botucatu (SF-22-Z-B-VI-3), utilizando-se as folhas 1:50.000, Dois Corregos e Sao Pedro, como areas de referencia. Variaveis descritoras do relevo e de geologia associadas as unidades de mapeamento pedologico das areas de referencia compuseram a matriz de dados de treinamento. A matriz foi analisada pelo algoritmo PART de arvore de decisao, do aplicativo Weka (Waikato Environment for Knowledge Analysis), que cria regras de classificacao. Essas regras foram aplicadas aos dados geomorfometricos e geologicos da folha Botucatu, para predicao de unidades de mapeamento pedologico. A validacao de campo dos mapas digitais deu-se por meio de amostragem por transectos em uma unidade de mapeamento da folha Sao Pedro e de forma aleatorio-estratificada na folha Botucatu. A avaliacao da unidade de mapeamento na folha Sao Pedro verificou confiabilidade, respectivamente, de 83 e 66 %, para os mapas pedologicos digital e tradicional com legenda simplificada. Apesar de terem sido geradas regras para todas as unidades de mapeamento pedologico das areas de treinamento, nem todas as unidades de mapeamento foram preditas na folha Botucatu, o que resultou das diferencas de relevo e geologia entre as areas de treinamento e de mapeamento. A validacao de campo do mapa digital da folha Botucatu verificou exatidao global de 52 %, compativel com levantamentos em nivel de reconhecimento de baixa intensidade, e kappa de 0,41, indicando qualidade Boa. Unidades de mapeamento mais extensas geraram mais regras, resultando melhor reproducao dos padroes solo-relevo na area a ser mapeada. A validacao por transectos na folha Sao Pedro indicou compatibilidade do mapa digital com o nivel de reconhecimento de alta intensidade e compatibilidade do mapa tradicional, apos simplificacao de sua legenda, com o nivel de reconhecimento de baixa intensidade. O treinamento do algoritmo em mapas e nao em observacoes pontuais reduziu em 14 % a exatidao do mapa pedologico digital da folha Botucatu. A amostragem aleatorio-estratificada pelo hipercubo latino e apropriada a mapeamentos com extensa base de dados, o que permite avaliar o mapa como um todo, tornando os trabalhos de campo mais eficientes. A amostragem em transectos e compativel com a avaliacao da pureza de unidades de mapeamento individualmente, nao necessitando de base de dados detalhada e permitindo estudos de associacoes solo-paisagem em pedossequencias.

Availability and spatial variability of copper, iron, manganese and zinc in soils of the State of Ceará, Brazil

The aim of the present work was to: (i) determine and map the micronutrient content (copper - Cu, iron - Fe, manganese - Mn and zinc - Zn) in soils of the State of Ceara; (ii) evaluate soil attributes related to micronutrient availability; and (iii) evaluate, by ordinary kriging, the spatial dependence of the micronutrient content. The study area is located in the central region of Ceara, in the basin of the River Banabuiu. Soil samples were collected in the surface and subsurface layers to determine the chemical attributes, granulometry, and micronutrient content, extracted with Mehlich-1 solution. The sites of the principal types of soil in the region were spatialised and the results submitted to descriptive analysis and Pearson correlation analysis. The semivariograms with the best fit were tested using the ordinary kriging method, and the maps of spatial variability were prepared. The highest values for the mean, maximum and minimum levels of Cu and Fe are found in the subsurface layers, while for Mn and Zn, they are found in the surface layers. In the surface layers, the majority of soils present Fe, Mn and Zn in levels suitable for agricultural production, while the levels of Cu are classified as medium to low. The micronutrients showed a significant positive correlation, especially with the soil attributes and with each other. Cu, Fe, Mn and Zn are classified as having medium to strong spatial dependence in the surface and subsurface layers.

Adequabilidade do uso agrícola das terras do sertão central do Ceará (Adequability of agricultural land use on the central hinterland of Ceará)

O Nordeste brasileiro historicamente e uma das regioes que mais exercem pressao sobre os recursos naturais. Nesta regiao o tipo climatico predominante e o semiarido, que passa por uma faixa de transicao semiumida. Ocorrem, em grande parte dessa regiao, solos de media a alta fertilidade natural, em geral pouco profundos em decorrencia de seu baixo grau de intemperismo. O objetivo do presente manuscrito e avaliar a adequabilidade do uso agricola das terras de uma area no sertao central do Estado do Ceara, levando-se em consideracao que os solos no semiarido brasileiro apontam a agricultura de subsistencia como uma das causas de impactos ambientais negativos, assim como o uso de tecnologias inadequadas, alem da pressao antropica para diversas finalidades. Os resultados apresentados sao fruto do cruzamento do mapa de uso e cobertura das terras com o mapa de aptidao agricola. Com base no mapeamento verificou-se que os solos da area apresentam restricoes quanto a determinados tipos de uso, como para lavouras, alem de nenhuma aptidao para silvicultura. Tambem verificou-se que a maior parte das terras encontra-se subutilizada, ou seja, com outros usos, principalmente pastagem e vegetacao nativa. Porem, o que requer maior cuidado sao as terras sobreutilizadas. A B S T R A C T The Brazilian Northeast is historically one of the regions that put more pressure on natural resources. In this region the predominant climate type is semiarid, which passes to zone of semiwet transition. Occur in this region, soils with medium to high natural fertility, in general, shallow, due their low degree of weathering. The objetive of this manuscript is to evaluate the adequability of the agricultural land use of an area in the central hinterland of Ceara State. The soils in the Brazilian semiarid are used by subsistence farming, which suggests negative environmental impacts, as well as the use of inappropriate technologies, in addition of the human pressure for several purposes. The data presented are the result of crossing between land use and cover map and agricultural suitability map. Based on the mapping was found that the soils of the area present restrictions to certain types of use, such as farming, and no capability to silviculture. Also it was found that most of the land is underutilized, with another uses, particularly grassland and natural vegetation. However, more care are required for super utilized lands. Keywords: GIS, fragility, agricultural management, semiarid.