Spectral analysis of vegetation with Hyperion data and his relationship with the concentration and the flux of CO2 in different environments in the Brazilian Amazon

Abstract

The spectral modeling is one of the important tools for determining biophysical parameters estimation of a location. Correlate the concentration and carbon dioxide flux through indirect measurements is a large advance in the field of geosciences. The methodology developed for forest carbon sink CO2flux, CO2fluxEVI and CO2fluxEVI2 from Hyperion involving natural and anthropic environments, making it effective when associated with variety of ecotypes existing in the Brazilian Amazon. Results of this study showed that NDVI indices, EVI and EVI2 could compose the rate structure of the forest sink of carbon, CO2flux. However, it was necessary to evaluate the potential and comparison of ratios in relation to detection of forest carbon sink associated to the direct measurements of CO2 in micrometeorological towers. After processing the hyperspectral imaging and statistical analysis of the suggested ratios, it could be inferred that CO2flux has a better adjustment when using NDVI, when associated to CO2 (ppm) concentration data measured from the top of the towers, as well as the data flux (mol.m-2.s-1), since the correlations have similarities and are statistically significant at the Hyperion data.