César da Silva Chagas

Avaliação de modelos digitais de elevação para aplicação em um mapeamento digital de solos

In Brazil, the digital elevation models (DEMs) are usually produced by users themselves and little attention has been given to their limitations as source of spatial information. The objective of this study was to evaluate different DEMs to help in choosing an appropriate model to derive topographical attributes used in a digital soil mapping based on a neural networks approach. The evaluation consisted of the following analysis: determination of root mean square error (RMSE) of elevation; analysis of the spurious depressions; comparison between mapped drainage and numeric drainage and between derived contour lines and original contour lines; and analysis of the derived contribution basins. The results demonstrated that RMSE was not enough to evaluate the quality of these models. DEMs derived from contour lines (CARTA, obtained using the TOPOGRID module) presented better quality than those derived from remote sensors (ASTER and SRTM). These presented great amount of errors that can negatively affect the establishment of relationships between topographical attributes and local conditions of soils.

Atributos topográficos e dados do Landsat7 no mapeamento digital de solos com uso de redes neurais

The objective of this study was to evaluate discriminant variables in digital soil mapping using artificial neural networks. The topographic attributes elevation, slope, aspect, plan curvature and topographic index, derived from a digital elevation model, and the indexes of clay minerals, iron oxide and normalized difference vegetation, derived from a Landsat7 image, were combined and evaluated for their ability to discriminate soils of an area at the northwest of Rio de Janeiro State. The Java neural simulator and the backpropagation learning algorithm were used. The maps generated by each of the six tested sets of variables were compared with reference points for determining the rating accuracy. This comparison showed that the map produced with the use of all the variables reached a performance (73.81% of agreement) superior to maps produced by other sets of variables. Possible sources of error in the use of this approach are mainly related to the great lithological heterogeneity of the area, which hindered the establishment of a more realistic model of environmental correlation. The approach can help make the soil survey in Brazil faster and less subjective.

Modelos de elevação para obtenção de atributos topográficos utilizados em mapeamento digital de solos

The objective of this work was to evaluate digital elevation models (DEM) obtained by different data sources and to select one of them for deriving morphometric variables used in digital soil mapping. The work was performed in the Guapi‑Macacu river basin, RJ, Brazil. The primary data used in the models generated by interpolation (DEM map and DEM hybrid) were: contour lines, drainage, elevation points, and remote sensor data transformed into points. The obtained models by remote sensing and aero‑restitution (DEM SRTM and DEM IBGE) were used in the comparison. All models showed spatial resolution of 30 m. The elevation model evaluations were based on: the terrain derived attribute analysis (slope, aspect, and curvature); spurious depressions (sink); comparison between features derived from the models and the original ones originated from planialtimetric maps; and the analysis of derived watersheds. The DEM hybrid showed a superior quality than the other models.

Digital soilscape mapping of tropical hillslope areas by neural networks

Geomorphometric variables are applied in digital soil mapping because of their strong correlation with the disposition and distribution of pedological components of the landscapes. In this research, the relationship between environmental components of tropical hillslope areas in the Rio de Janeiro State, Brazil, artificial neural networks (ANN), and maximum likelihood algorithm (MaxLike) were evaluated with the aid of geoprocessing techniques. ANN and MaxLike were applied to soilscape mapping and the results were compared to the original map. The ANN architectures with seven and five neurons in the hidden layer produced the best classifications when using samples obtained systematically. When random samples were applied, the best neural net architectures were within 22 and 16 neurons in the hidden layer. In conclusion, the ANN can contribute to soilscape surveys, making map delineation faster and less expensive. The digital elevation model (DEM) and its derived attributes can contribute to the understanding of the soil-landscape relationship of tropical hillslope areas; the use of artificial neural networks and MaxLike is feasible for digital soilscape mapping. The systematic sampling method provided a global accuracy of 70 % and 65.9 % for the ANN and the MaxLike, respectively. When the random sampling method was applied, the ANN had a global accuracy of 69.6 %, and the MaxLike had an accuracy of 62.1 %, considering the total study area in relation to the reference map.

Elaboração de zoneamentos agropedoclimáticos por geoprocessamento: soja em municípios do Rio Grande do Sul

Environmental databases and several variables based on geoprocessing operations are used for a classification of Agropedoclimatic Zoning. This paper presents a case study of soybean carried out in the State of Rio Grande do Sul. PC-ArcInfo, Arcview, and SGI VGA softwares were used for data input, geoprocessing, and result presentation. Geoprocessing included the reclassification and overlay operations of information layers and their associated tables. Soil characteristics, such as fertility, texture, relief, effective depth, erosion susceptibility, drainage, rocks, and sodium saturation were assessed for soybean culture for all components of the soil map units of Rio Grande do Sul. An information layer for the pedologic aptitude of this culture was created and then overlaid with the sowing time layer for municipal districts, generating the final information for Agropedoclimatic Zoning. This method allowed the evaluation of: (a) every single component of the map units, and (b) results obtained in agreement with the percentage of occurrence of each component within the map unit.

Comparison between artificial neural networks and maximum likelihood classification in digital soil mapping

O levantamento de solos e a principal fonte de informacao espacial sobre solos para diferentes usos, principalmente o uso agricola. No entanto, a continuidade dessa atividade tem sido grandemente comprometida, principalmente pela escassez de recursos financeiros. O objetivo deste estudo foi avaliar a eficiencia da utilizacao de dois classificadores distintos (redes neurais artificiais - RNAs e o algoritmo da maxima verossimilhanca - Maxver) na predicao de classes de solos em uma area na regiao noroeste do Estado do Rio de Janeiro. As variaveis discriminantes usadas incluem atributos do terreno, como elevacao, declividade, aspecto, plano de curvatura e indice topografico combinado (CTI) e indices clay minerals, iron oxide e de vegetacao NDVI, derivados de uma imagem do sensor ETM+ do LANDSAT 7. Para o treinamento e a validacao dos classificadores, foram utilizadas, respectivamente, 300 e 150 amostras por classe de solo, representativas das caracteristicas dessas classes, com relacao as variaveis discriminantes utilizadas. De acordo com os testes estatisticos realizados, o classificador com base na RNA produziu maior exatidao do que o classificador classico da maxima verossimilhanca (Maxver). A comparacao com 126 pontos de referencia coletados no campo evidenciou que o mapa produzido pela RNA teve desempenho superior (73,81 %) ao mapa produzido pelo algoritmo Maxver (57,94 %). As principais causas de erros detectadas na utilizacao desses classificadores foram: a heterogeneidade geologica da area aliada a problemas no mapa geologico utilizado; profundidade do contato litico e, ou, exposicao da rocha; e problemas com o modelo de correlacao ambiental utilizado em razao da natureza poligenetica dos solos. Os resultados obtidos permitem inferir que a utilizacao de atributos do terreno juntamente com dados de sensoriamento remoto em uma abordagem por RNAs pode contribuir para facilitar o mapeamento de solos no Brasil, principalmente por causa da disponibilidade de dados de sensores remotos a custos mais baixos e da facilidade de obtencao dos atributos do terreno.

Mapeamento digital de solos por redes neurais artificiais com base na relação solo-paisagem

Digital mapping techniques can help reduce the lack of soil information in areas where no 1st and 2nd order soil surveys were performed. The aim of this study was to obtain a digital soil map (DSM) by artificial neural networks (ANN) using the correlation between soil mapping units and environmental covariates. The study area of approximately 11,000 ha is located in Barra Bonita, SP, Brazil. Based on a cluster analysis of environmental covariates, five reference areas were chosen for conventional mapping. The selected soil mapping units supported the application of ANN. We used the neural network simulator JavaNNS and the backpropagation learning algorithm. Reference points were collected to evaluate the efficiency of the resulting digital map. The position in the landscape and the underlying parent material were fundamental to the recognition of the designs of the mapping units. There was good agreement between the mapping units delineated by DSM and the conventional method. The comparison between the reference points and the digital soil map showed an accuracy of 72 %. The use of the DSM approach can help reduce the lack of soil information in unmapped places, based on soil information obtained from adjacent reference areas.

Artificial neural networks applied for soil class prediction in mountainous landscape of the Serra do Mar

Soil information is needed for managing the agricultural environment. The aim of this study was to apply artificial neural networks (ANNs) for the prediction of soil classes using orbital remote sensing products, terrain attributes derived from a digital elevation model and local geology information as data sources. This approach to digital soil mapping was evaluated in an area with a high degree of lithologic diversity in the Serra do Mar. The neural network simulator used in this study was JavaNNS and the backpropagation learning algorithm. For soil class prediction, different combinations of the selected discriminant variables were tested: elevation, declivity, aspect, curvature, curvature plan, curvature profile, topographic index, solar radiation, LS topographic factor, local geology information, and clay mineral indices, iron oxides and the normalized difference vegetation index (NDVI) derived from an image of a Landsat-7 Enhanced Thematic Mapper Plus (ETM+) sensor. With the tested sets, best results were obtained when all discriminant variables were associated with geological information (overall accuracy 93.2 - 95.6 %, Kappa index 0.924 - 0.951, for set 13). Excluding the variable profile curvature (set 12), overall accuracy ranged from 93.9 to 95.4 % and the Kappa index from 0.932 to 0.948. The maps based on the neural network classifier were consistent and similar to conventional soil maps drawn for the study area, although with more spatial details. The results show the potential of ANNs for soil class prediction in mountainous areas with lithological diversity.

Using Soil Depth Functions to Distinguish Dystric from Xanthic Ferralsols in the Landscape

The soil texture is a key parameter and is widely used as input in predictive models to estimate other soil properties. The general goal was creating numerical parameters to describe the variability of soil particle size (sand, silt, and clay) components using continuous depth functions to characterize Ferralsols from Guapi-Macacu watershed in Rio de Janeiro State (Brazil). The profile collection comprises fifteen profiles, seven classified as Haplic Ferralsols (Dystric) and eight as Haplic Ferralsols (Xanthic). The analysis was performed in the R software through “aqp” package (Algorithms for Quantitative Pedology) and using equal-area quadratic spline function. A numerical aggregation of soil texture components was used to build a mean, a median, and spline depth functions, fitting the dataset to six predefined depths (GlobalSoilMap project) and to most-likely horizon depths. The analysis revealed sand and silt content with decreasing values with soil depth and the opposite trend for clay. The topsoil layer (0–30 cm) had dominantly a clay loam texture (32–40 % clay; 49–53 % sand; and 15–20 % silt). The most-likely diagnostic B-horizon (45–150 cm depth) presented clayey texture (43–47 % of clay and 40–55 % of sand). Ferralsols usually have low silt contents; and the silt range was from 10 to 20 % in the soil profile collection. The organized data can be useful to many purposes, including profile database harmonization and soil classification.

Quantitative pedology to evaluate a soil profile collection from the Brazilian semi-arid region§

This work applies pedometric tools to analyse soil property information relevant to morphological characterisation and soil classification. The objective of this paper was further to identify similarities in soil properties among a soil profile collection. The harmonisation of soil data enables the comparison between soil profiles, transference of information and modelling of soil horizons distribution. The statistical procedures were implemented in R software, through the Algorithms for Quantitative Pedology (AQP package), which contains a collection of algorithms to model soil resources and aid soil classification, soil profile aggregation and visualisation. The procedures allowed definition of values for soil properties in every one-centimetre layer of the soil profile, by regrouping the data in a different layer thickness, and it was possible to analyse similarity between profiles using a dissimilarity matrix for each depth slice. The AQP allowed analysis of a large number of soil profiles, in terms of vertical variability of soil continuous properties (e.g. sand and clay content, and pH) and for categorical variables, such as diagnostic horizons. Soil depth functions were developed to represent soil properties and probability of occurrence to diagnostic horizons to a large data set, and the dissimilarity analysis allowed separation of a small group of similar soil profiles and further qualitative comparison among the select profiles.