Multivariate analysis applied to reduce the number of predictors in digital soil mapping
The objective of this work was to assess the possibility of generating a smaller set of uncorrelated predictors, potentially applicable to digital soil mapping, by multivariate statistical analysis. The terrain attributes, elevation, slope, stream distance, planar curvature, profile curvature, relative available radiation, natural logarithm of the contributing area, topographic wetness index, and sediment transport capacity, were transformed by the Varimax method into the variables: altimetry, hydrology, and curvature. This transformation represented a variability concentration of 65.57% of the original data in the three new components. The new variables enable the use of a smaller amount of data set in prediction models, besides the fact that they are uncorrelated. Varimax rotation allows the relationship between environment and soil formation to be explicitly included in the prediction models.
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Digital revista |
| Language: | por |
| Published: |
Pesquisa Agropecuaria Brasileira
2011
|
| Online Access: | https://seer.sct.embrapa.br/index.php/pab/article/view/9731 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| id |
rev-pab-br-article-9731 |
|---|---|
| record_format |
ojs |
| spelling |
rev-pab-br-article-97312014-05-28T19:11:31Z Multivariate analysis applied to reduce the number of predictors in digital soil mapping Estatística multivariada aplicada à diminuição do número de preditores no mapeamento digital do solo Caten, Alexandre ten Dalmolin, Ricardo Simão Diniz Pedron, Fabrício Araújo Mendonça-Santos, Maria de Lourdes principal components analysis; terrain attributes; pedometry; remote sensing análise de componentes principais; atributos de terreno; pedometria; sensoriamento remoto The objective of this work was to assess the possibility of generating a smaller set of uncorrelated predictors, potentially applicable to digital soil mapping, by multivariate statistical analysis. The terrain attributes, elevation, slope, stream distance, planar curvature, profile curvature, relative available radiation, natural logarithm of the contributing area, topographic wetness index, and sediment transport capacity, were transformed by the Varimax method into the variables: altimetry, hydrology, and curvature. This transformation represented a variability concentration of 65.57% of the original data in the three new components. The new variables enable the use of a smaller amount of data set in prediction models, besides the fact that they are uncorrelated. Varimax rotation allows the relationship between environment and soil formation to be explicitly included in the prediction models. O objetivo deste trabalho foi avaliar a possibildade de se gerar um menor conjunto de preditores não correlacionados e potencialmente aplicáveis ao mapeamento digital de solos, pelo uso da estatística multivariada. Os atributos de terreno, elevação, declividade, distância à drenagem, curvatura planar, curvatura de perfil, radiação relativa disponível, logaritmo natural da área de contribuição, índice de umidade topográfica e capacidade de transporte de sedimento, foram transformados pelo método Varimax nas variáveis: altimetria, hidrologia e curvatura. Essa transformação representou uma concentração de 65,57% da variabilidade dos dados originais nas três novas componentes. As novas variáveis possibilitam o emprego de menor quantidade de dados nos modelos preditivos, além do fato de serem não correlacionados. A rotação Varimax permite que a relação com o ambiente de formação do solo seja explicitamente inserida nos modelos preditivos. Pesquisa Agropecuaria Brasileira Pesquisa Agropecuária Brasileira CNPQ e CAPES CNPq e CAPES 2011-07-28 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion application/pdf https://seer.sct.embrapa.br/index.php/pab/article/view/9731 Pesquisa Agropecuaria Brasileira; v.46, n.5, maio 2011; 554-562 Pesquisa Agropecuária Brasileira; v.46, n.5, maio 2011; 554-562 1678-3921 0100-104x por https://seer.sct.embrapa.br/index.php/pab/article/view/9731/6357 https://seer.sct.embrapa.br/index.php/pab/article/downloadSuppFile/9731/4973 |
| institution |
EMBRAPA |
| collection |
OJS |
| country |
Brasil |
| countrycode |
BR |
| component |
Revista |
| access |
En linea |
| databasecode |
rev-pab-br |
| tag |
revista |
| region |
America del Sur |
| libraryname |
Sistema de bibliotecas de EMBRAPA |
| language |
por |
| format |
Digital |
| author |
Caten, Alexandre ten Dalmolin, Ricardo Simão Diniz Pedron, Fabrício Araújo Mendonça-Santos, Maria de Lourdes |
| spellingShingle |
Caten, Alexandre ten Dalmolin, Ricardo Simão Diniz Pedron, Fabrício Araújo Mendonça-Santos, Maria de Lourdes Multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| author_facet |
Caten, Alexandre ten Dalmolin, Ricardo Simão Diniz Pedron, Fabrício Araújo Mendonça-Santos, Maria de Lourdes |
| author_sort |
Caten, Alexandre ten |
| title |
Multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| title_short |
Multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| title_full |
Multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| title_fullStr |
Multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| title_full_unstemmed |
Multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| title_sort |
multivariate analysis applied to reduce the number of predictors in digital soil mapping |
| description |
The objective of this work was to assess the possibility of generating a smaller set of uncorrelated predictors, potentially applicable to digital soil mapping, by multivariate statistical analysis. The terrain attributes, elevation, slope, stream distance, planar curvature, profile curvature, relative available radiation, natural logarithm of the contributing area, topographic wetness index, and sediment transport capacity, were transformed by the Varimax method into the variables: altimetry, hydrology, and curvature. This transformation represented a variability concentration of 65.57% of the original data in the three new components. The new variables enable the use of a smaller amount of data set in prediction models, besides the fact that they are uncorrelated. Varimax rotation allows the relationship between environment and soil formation to be explicitly included in the prediction models. |
| publisher |
Pesquisa Agropecuaria Brasileira |
| publishDate |
2011 |
| url |
https://seer.sct.embrapa.br/index.php/pab/article/view/9731 |
| work_keys_str_mv |
AT catenalexandreten multivariateanalysisappliedtoreducethenumberofpredictorsindigitalsoilmapping AT dalmolinricardosimaodiniz multivariateanalysisappliedtoreducethenumberofpredictorsindigitalsoilmapping AT pedronfabricioaraujo multivariateanalysisappliedtoreducethenumberofpredictorsindigitalsoilmapping AT mendoncasantosmariadelourdes multivariateanalysisappliedtoreducethenumberofpredictorsindigitalsoilmapping AT catenalexandreten estatisticamultivariadaaplicadaadiminuicaodonumerodepreditoresnomapeamentodigitaldosolo AT dalmolinricardosimaodiniz estatisticamultivariadaaplicadaadiminuicaodonumerodepreditoresnomapeamentodigitaldosolo AT pedronfabricioaraujo estatisticamultivariadaaplicadaadiminuicaodonumerodepreditoresnomapeamentodigitaldosolo AT mendoncasantosmariadelourdes estatisticamultivariadaaplicadaadiminuicaodonumerodepreditoresnomapeamentodigitaldosolo |
| _version_ |
1813448908347539456 |