This work emphasizes the importance of integrating Geostatistics and Geographic Information Syste... more This work emphasizes the importance of integrating Geostatistics and Geographic Information Systems (GIS). The Georeferenced Information Processing System (SPRING), developed at the Image Processing Division (DPI) of the National Institute for Space Research, is used to test such na integration. The conceptual basis of this study finds its foundations on geostatistics techniques. These techniques are derived from Matheron's theory on
E-MAIL: [email protected] RESUMO Propriedades ambientais são dados importantes para modelagens ... more E-MAIL: [email protected] RESUMO Propriedades ambientais são dados importantes para modelagens desenvolvidas em ambiente de Sistemas de Informação Geográfica. As técnicas geoestatísticas de inferência, conhecidas por krigeagem, consideram os atributos ambientais como variáveis aleatórias o que possibilita a modelagem das incertezas associadas às inferências. Neste contexto, o presente trabalho explora o uso da técnica de krigeagem não linear, a krigeagem por indicação, para inferências a partir de amostras observadas pontualmente. O trabalho apresenta, também, alternativas de cálculos de incertezas para propriedades ambientais de natureza numérica e temática. Por fim, o trabalho mostra e analisa os resultados da aplicação da krigeagem por indicação sobre amostras de textura de solo.
Increasingly, geostatistical procedures have been used for analysis and for attribute modeling of... more Increasingly, geostatistical procedures have been used for analysis and for attribute modeling of spatial data. These procedures are applied to an attribute sample set, represented as points, and allow the estimation of numerical and categorical attribute values in locations not sampled. Furthermore, the geostatistical estimation processes permit the inference of uncertainties related to the estimated values. In Geographical Information Systems -GIS -environment, the attribute models and their uncertainties can be used as input data for computational modeling of environmental processes. This paper addresses the main aspects involved in the incorporation of geostatistical procedures in a GIS environment. The requirements for a geostatistical module, implemented as part of the GIS, are described and analyzed. Also, the extension of a model description language to consider the uncertainties of the data models and to propagate them to the computational modeling results is presented and discussed.
Anisotropy is a common property in environment variables, implying in preferential direction for ... more Anisotropy is a common property in environment variables, implying in preferential direction for the spatial distribution of the environment variable. This directional tendency is difficult to be detected as well as modeled through deterministic procedures. To reconstruct the spatial distribution of a nature variable from a sample set, it is important that the model adequately represents its spatial autocorrelation structure. Geostatistics offers a set procedures that deals with anisotropy. It involves a carefull analysis of the variable's spatial variability, caracterized by the semivariogram. This paper presents and discusses the results of a study, using geostatistic procedures to reconstruct the spatial distribution of clay content in the soil of EMBRAPA's Canchim Farm in São Carlos -SP.
This work emphasizes the importance of integrating Geostatistics and Geographic Information Syste... more This work emphasizes the importance of integrating Geostatistics and Geographic Information Systems (GIS). The Georeferenced Information Processing System (SPRING), developed at the Image Processing Division (DPI) of the National Institute for Space Research, is used to test such na integration. The conceptual basis of this study finds its foundations on geostatistics techniques. These techniques are derived from Matheron's theory on
E-MAIL: [email protected] RESUMO Propriedades ambientais são dados importantes para modelagens ... more E-MAIL: [email protected] RESUMO Propriedades ambientais são dados importantes para modelagens desenvolvidas em ambiente de Sistemas de Informação Geográfica. As técnicas geoestatísticas de inferência, conhecidas por krigeagem, consideram os atributos ambientais como variáveis aleatórias o que possibilita a modelagem das incertezas associadas às inferências. Neste contexto, o presente trabalho explora o uso da técnica de krigeagem não linear, a krigeagem por indicação, para inferências a partir de amostras observadas pontualmente. O trabalho apresenta, também, alternativas de cálculos de incertezas para propriedades ambientais de natureza numérica e temática. Por fim, o trabalho mostra e analisa os resultados da aplicação da krigeagem por indicação sobre amostras de textura de solo.
Increasingly, geostatistical procedures have been used for analysis and for attribute modeling of... more Increasingly, geostatistical procedures have been used for analysis and for attribute modeling of spatial data. These procedures are applied to an attribute sample set, represented as points, and allow the estimation of numerical and categorical attribute values in locations not sampled. Furthermore, the geostatistical estimation processes permit the inference of uncertainties related to the estimated values. In Geographical Information Systems -GIS -environment, the attribute models and their uncertainties can be used as input data for computational modeling of environmental processes. This paper addresses the main aspects involved in the incorporation of geostatistical procedures in a GIS environment. The requirements for a geostatistical module, implemented as part of the GIS, are described and analyzed. Also, the extension of a model description language to consider the uncertainties of the data models and to propagate them to the computational modeling results is presented and discussed.
Anisotropy is a common property in environment variables, implying in preferential direction for ... more Anisotropy is a common property in environment variables, implying in preferential direction for the spatial distribution of the environment variable. This directional tendency is difficult to be detected as well as modeled through deterministic procedures. To reconstruct the spatial distribution of a nature variable from a sample set, it is important that the model adequately represents its spatial autocorrelation structure. Geostatistics offers a set procedures that deals with anisotropy. It involves a carefull analysis of the variable's spatial variability, caracterized by the semivariogram. This paper presents and discusses the results of a study, using geostatistic procedures to reconstruct the spatial distribution of clay content in the soil of EMBRAPA's Canchim Farm in São Carlos -SP.
Uploads
Papers by Eduardo Celso