%0 Conference Proceedings
%@isbn 978-85-17-00056-0 (Internet)
%@isbn 978-85-17-00057-7 (DVD)
%T Discriminating burned areas in Amazon and Cerrado regions using MIR/NIR information
%8 30 abr. - 5 maio 2011
%A Santos, Renata Libonati dos,
%A Camara, Carlos do Carmo de Portugal e Castro da,
%A Pereira, José Miguel Cardoso,
%A Peres, Leonardo de Faria,
%@affiliation Instituto Nacional de Pesquisas Espaciais - INPE
%@affiliation Instituto Dom Luiz - IDL/CGUL - Portugal
%@affiliation Instituto Superior de Agronomia - ISA/DF - Portugal
%@affiliation Universidade Federal do Rio de Janeiro - UFRJ
%E Epiphanio, José Carlos Neves,
%E Galvão, Lênio Soares,
%B Simpósio Brasileiro de Sensoriamento Remoto, 15 (SBSR).
%I Instituto Nacional de Pesquisas Espaciais (INPE)
%J São José dos Campos
%1 Instituto Nacional de Pesquisas Espaciais (INPE)
%K middle-infrared, spectral index, vegetation, burns, remote sensing.
%X Although the Brazilian Amazonia together with Cerrado region presents one of the highest numbers of occurrences of fire events, hardly any studies aiming to design burned area (BA) indices specifically for these regions have been attempted. This issue is of particular interest since the accuracy of BA maps is closed related to the characteristics of the location (e.g., pre-fire land-cover type and conditions, background soil, fire severity, post-fire processes, and atmospheric conditions) and in addition, index thresholds are often subjective or vary from region to region. Accordingly, a new spectral index, specifically designed for burned land discrimination in Amazon and Cerrado regions using the middle/near-infrared spectral domain, was tested on sets of Moderate Resolution Imaging Spectroradiometer (MODIS) images. The utility of the new index for burned land discrimination was assessed against other widely used spectral indices: Normalized Difference Vegetation Index (NDVI), Burned Area Index (BAI) and Global Environmental Monitoring Index (GEMI), defined in the red/near infrared space and the modified version on them, namely VI20, BAI20 and GEMI20 defined in the middle/nearinfrared space of MODIS sensor. The ability of each index to discriminate between burned and unburned surfaces was assessed by means of a discrimination index. The new proposed (V,W) index works better than the traditional indices in all three cases analyzed, with the advantage of not requiring water masks, which is not true for the remaining indices.
%9 Queimadas, Riscos Ambientais e Poluição