Publicação
Using middle-infrared reflectance for burned area detection
| Resumo: | A strategy is presented that allows deriving a new index for burned area discrimination over the Amazon and Cerrado regions of Brazil. The index is based on information from the near-infrared (NIR) and middle-infrared (MIR) channels of the Moderate Resolution Imaging Spectroradiometer (MODIS). A thorough review is undertaken of existing methods for retrieving MIR reflectance and an assessment is performed, using simulated and real data, about the added value obtained when using the radiative transfer equation (RTE) instead of the simplified algorithm (KR94) developed by Kaufman and Remer (1994), the most used in the context of burned area studies. It is shown that use of KR94 in tropical environments to retrieve vegetation reflectance may lead to errors that are at least of the same order of magnitude of the reflectance to be retrieved and considerably higher for large values of land surface temperature (LST) and solar zenith angle (SZA). Use of the RTE approach leads to better estimates in virtually all cases, with the exception of high values of LST and SZA, where results from KR94 are also not usable. A transformation is finally defined on the MIR/NIR reflectance space aiming to enhance the spectral information such that vegetated and burned surfaces may be effectively discriminated. The transformation is based on the difference between MIR and NIR in conjunction with the distance from a convergence point in the MIR/NIR space, representative of a totally burnt surface. The transformation allows defining a system of coordinates, one coordinate having a small scatter for pixels associated to vegetation, burned surfaces and soils containing organic matter and the other coordinate covering a wide range of values, from green and dry/stressed vegetation to burned surfaces. The new set of coordinates opens interesting perspectives to applications like drought monitoring and burned area discrimination using remote-sensed information. |
|---|---|
| Autores principais: | SANTOS, Renata |
| Assunto: | Infravermelhos Queimada Sensores remotos Índice espectral MODIS (Moderate Resolution Imaging Spectroradiometer) Amazónia Teses de doutoramento - 2011 |
| Ano: | 2010 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade de Lisboa |
| Idioma: | inglês |
| Origem: | Repositório da Universidade de Lisboa |
| Resumo: | A strategy is presented that allows deriving a new index for burned area discrimination over the Amazon and Cerrado regions of Brazil. The index is based on information from the near-infrared (NIR) and middle-infrared (MIR) channels of the Moderate Resolution Imaging Spectroradiometer (MODIS). A thorough review is undertaken of existing methods for retrieving MIR reflectance and an assessment is performed, using simulated and real data, about the added value obtained when using the radiative transfer equation (RTE) instead of the simplified algorithm (KR94) developed by Kaufman and Remer (1994), the most used in the context of burned area studies. It is shown that use of KR94 in tropical environments to retrieve vegetation reflectance may lead to errors that are at least of the same order of magnitude of the reflectance to be retrieved and considerably higher for large values of land surface temperature (LST) and solar zenith angle (SZA). Use of the RTE approach leads to better estimates in virtually all cases, with the exception of high values of LST and SZA, where results from KR94 are also not usable. A transformation is finally defined on the MIR/NIR reflectance space aiming to enhance the spectral information such that vegetated and burned surfaces may be effectively discriminated. The transformation is based on the difference between MIR and NIR in conjunction with the distance from a convergence point in the MIR/NIR space, representative of a totally burnt surface. The transformation allows defining a system of coordinates, one coordinate having a small scatter for pixels associated to vegetation, burned surfaces and soils containing organic matter and the other coordinate covering a wide range of values, from green and dry/stressed vegetation to burned surfaces. The new set of coordinates opens interesting perspectives to applications like drought monitoring and burned area discrimination using remote-sensed information. |
|---|