Processing and characterization of fully and partial nanostructured alumina powders and ceramics / Processamento e caracterização de pós e de cerâmicas de alumina total e parcialmente nanoestruturadas

AUTOR(ES)
DATA DE PUBLICAÇÃO

2007

RESUMO

The opportune access to trustworthy agricultural information has become very important for the decision process of national and international agents. The use of remote sensing to map agricultural crops must consider their high dynamism, therefore the use of methods that consider their spectral-temporal profile is needed. The Mato Grosso State Brazil became one of the more productive agricultural regions of the country. Currently this state occupies a prominence position in the national agricultural production, since it occupies the first place in soybean production, which corresponds to 31% of this grain production in the country. In the year of 2005, the soybean cultivated in the Mato Grosso State was responsible for generating an income of 6,3 billions Reais. The present work has a hypothesis that the use of classification techniques and spectral-temporal analysis of remote sensing images allow the identification and quantification of soybean crop areas in Mato Grosso State, through an objective method. Thus, the main objective of the present work was to evaluate the capacity of the MODerate Resolution Imaging Spectroradiometer (MODIS) images to estimate the soybean crop area cultivated in the Mato Grosso State in the 2005/2006 crop season. To test the hypothesis the Spectral-Temporal Response Surface (STRS) classification method was used, that instead of using the digital numbers or the reflectance of the multi spectral-temporal images, it uses the coefficients of a polynomial generated for each STRS, from each pixel, for this spectral-temporal series. Another technique evaluated was the Crop Enhancement Index (CEI), which explores the Enhanced Vegetation Index (EVI) temporal profile in defined agricultural calendar, allowing the identification and quantification of agricultural crops. To produce a reference, 30 segments, spread throughout the State, of 30 x 30 kilometers was elaborated for the results comparison generated from the MODIS images classifications through the STRS and CEI methods, from Landsat-5/TM images. In each segment the soybean crop area was mapped. The area estimated by the STRS method was 17.56% higher than the reference areas. However, in regions with few soybean areas the classifier overestimated the areas in 56.96%, when compared with the reference areas. The global accuracy of this method was 80%, however with a Kappa value of 0.2634. The use of the CEI method to identify the soybean crop areas was very efficient. The areas identified by this method were underestimated in average by 13.75% when compared to the reference segments and in 12.64% when compared to all segments together. The global accuracy of this mapping technique was also 80%, but with a Kappa value of 0.5137.

ASSUNTO(S)

Óxido de alumínio aluminum oxides grãos nanoestruturados nanopartículas engenharia e tecnologia espacial precipitation chemistry nanostructure growth space engeneering and technology nanotechnology ceramics precipitação química nanoparticles nanotecnologia cerâmica

ACESSO AO ARTIGO

Documentos Relacionados