Modelagem de dados do satélite Ikonos II para estimativa de micronutrientes na floresta ombrófila mista montana
Resumo
Resumo: O presente trabalho teve como objetivo desenvolver uma metodologia utilizando imagens do satélite IKONOS II para realizar estimativas de micronutrientes na Floresta Ombrófila Mista Montana. Para o desenvolvimento desta metodologia utilizou-se de material coletado no município de General Carneiro (PR) na propriedade das Indústrias Pedro N. Pizzatto Ltda em 20 unidades amostrais primárias de 12 x 12 metros. Os materiais coletados foram separados, secos e pesados, obtendo-se, então, os teores de micronutrientes; para a seleção das bandas e índices de vegetação foi necessário o método de eliminação de variáveis "Stepwise" sendo a reflectância das bandas MS-1, MS-2, MS-3, MS-4, razão de bandas (MS-4/MS-3), NDVI e SAVI consideradas variáveis independentes e os teores de micronutrientes considerados como variáveis dependentes. O modelo selecionado para cada micronutriente foi baseado no grau de associação das variáveis e que melhor representassem a distribuição dos micronutrientes, possibilitando sua quantificação e mapeamento. O estudo mostrou ser possível modelar as equações que estimaram a quantidade de micronutrientes, por meio de técnicas de sensoriamento remoto Abstract: The objective of the present research is to develop A methodology using images from the IKONOS II satellite in order to accomplish estimates at Ombrófila Mista Montana Forest. In order for this methodology to be develop it was used data collected on the property of Industrial Pedro N. Pizatto LTDA in General Carneiro (PR) in 20 units primaries of 12 x12 meters. The litter that had been collected was separated, dried and weighed. From that it was obtained concentrations of micronutrients. For the selection of the bands and indexes of vegetation it was necessary are elimination method of variables called Stepwise using reflectances of MS-1, MS-2, MS-3 reason of the bands (MS-4/ MS-3), NDVI and SAVI considering the independent variable and the concentrations of micronutrients as dependent variable. The selected model for each micronutrients was based on the variable association degree and what better could represent the distribution of the micronutrients. The research showed that is possible to model equations making possible the use of remote sensing techniques to estimate micronutrients from forest litter.
Collections
- Teses & Dissertações [10806]
