Atmospheric Correction of Landsat 8 and RapidEye imagery
ISBN 978-85-88783-11-9
Authors
1Rubin, I.; 2Richter, M.; 3Brasileiro, R.; 4Keidel, G.
1INCRA/UFRJ Email: igorbr1@gmail.com
2UFRRJ Email: mrichter84@hotmail.com
3UFRJ Email: roberta.brasileiro.c@gmail.com
4UFRJ Email: gabrielkeidel@hotmail.com
Abstract
The radiation from the earth's surface, which is captured by remote sensors, is subject to absorption and scattering processes. Thus, the sensor receives a flow of direct radiation from the target component and a diffuse from the atmosphere itself. In order to obtain the surface reflectance from remote sensing data with the reduction of atmospheric interference, correction techniques have been developed. Some of these techniques involves alternatives methods of self-image information to estimate the degree of correction to be introduced or physical methods which are based on the theory of radiative transfer (PONZONI et al. 2012). According CHAVEZ (1996) to generate results with minimal error, the models require atmospheric measurements in situ and the simultaneous image acquisition by the satellite, which is impractical for historical studies or made in areas that do not have those data. The author points out that although the ideal is to use methods based on field data, for without it (which happens in most cases) a simple technique called Dark Object Subtraction (DOS), can be used alternatively which is based on data of the image itself. Besides this, other authors (Gomes, 2011; Pepper, 2013) has obtained satisfactory results from radiative transfer models to rescue original or values close to the target surface reflectance: the Moderate Resolution Spectral Atmospheric Transmitance Algorithm (MODTRAN) and the Simulation of the Satellite Signal in the Solar Spectrum. These models are implemented respectively in Atcor and 6S (Vermote et al., 1997). This atmospheric correction technique considers atmospheric models relating to water vapor concentrations, O3, optical depth, type and concentration of aerosols. However, despite the literature generally stress the importance in trying to minimize the effects of the atmosphere on the response of targets, one should be aware of the fact of the necessity to infer the parameters concerning the visibility and atmospheric and aerosol models, testing different combinations. Following this questioning, the objective was to use two different methods of atmospheric correction from the radiative transfer technique, 6S and the Atcor in Landsat 8 and Rapid Eye imagery; having as area cutout portion of the Baixada Fluminense/RJ. The results were compared with surface reflectance data found in the publications reviewed, indicating that the atmospheric correction of both the Landsat8, as the RapidEye, showed satisfactory results in both programs, taking 6S the advantage as it consist of a free software. Finally, it is noteworthy that mainly visibility, which is a big weight parameter to the end result of the treatment, usually estimated, demonstrating the importance of tests to be performed. References: Chavez Jr., PSImage-based atmospheric corrections - revisited and improved. Protogrammetric Engineering and Remote Sensing, vol. 62, no. 9, p. 1025-1036, 1996. Gomes, D .; et al. Comparative assessment of atmospheric correction of Landsat images using MODTRAN and Dark Object Subtraction. In: Brazilian Symposium on Remote Sensing (RSBS), 15, 2011, Curitiba. Annals online. Pepper, M. L. F .; et al. Studies of the parameters defining the uncertainties in the atmospheric correction process. In: Brazilian Symposium on Remote Sensing (RSBS), 16, 2013, Foz do Iguaçu. Annals online. Available at: <http://www.dsr.inpe.br/sbsr2013/files/p1019.pdf>. Access: 28 October in 2014 Ponzoni, F. J .; Shimabukuro, Y. E .; Kuplich, TM. Remote Sensing of Vegetation. São José dos Campos, SP: Texts Workshop. 2012, 176 p. Vermote, F., Tanré, D., Deuze, JL, Herman, M., Morcrete, J. Second simultion of the Stellite Signal in the Solar Spectrum, 6S: an overview. IEEE transactions on Geoscience and Remote Sensing, v.35 n. 3 p. 675-232, 1997.
Keywords
Remote Sensing; image processing; Surface reflectance