Avaliação dos efeitos topográficos e direcionais sobre os índices de vegetação na definição de zonas de manejo em povoamentos de pinus no sul do Brasil

• Ano de defesa: 2021
• Autor(a) principal: Gaida, William
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Santa Maria; Brasil; Geografia; UFSM; Programa de Pós-Graduação em Geografia; Centro de Ciências Naturais e Exatas
• Palavras-chave em Português: Silvicultura; Resposta espectral; Manejo florestal; Geometria de visada; Modelagem topográfica; Forestry; Spectral response; Forest management; Field of view; Topographic modeling; CNPQ::CIENCIAS HUMANAS::GEOGRAFIA
• Link de acesso: http://repositorio.ufsm.br/handle/1/23197

The application of remote sensing data represents an alternative for monitoring forest stands. However, in most applications, the sources of radiometric uncertainty associated with the data are not considered. This study aims to analyze the influence of variations in illumination and viewing geometries on the delineation of management zones in pine stands. The influence of the topographic effect was analyzed using the multitemporal series of MSI/Sentinel-2A/B and OLI/Landsat-8 images, with three sample dates defined for each set covering the seasonal variations of the local illumination geometry. The analyzes consisted of graphical and correlation/linear regression analyzes between reflectance measurements, vegetation indices, and canopy illumination conditions. Analysis of the influence of directional effects on forest stand response was performed using reflectance measurements from the MODIS/Terra sensor system. Six pairs of consecutive dates with opposite viewing directions, defined to consider seasonal variations in local illumination geometry, were used for the graphical and statistical analyses. Analysis of the sensitivity of the vegetation indices was performed by analysis of variance. The delineation of management zones was carried out with the classification of vegetative vigor, into low, expected, and high. This was done using vegetative vigor maps, obtained from the average annual values of the EVI index, calculated using reflectance measurements from MSI/Sentinel-2A/B images, corrected for topographic effects. The results obtained from the seasonal variations of the local lighting geometry, associated with the relief of the management area, significantly affect the acquisitions of reflectance measurements and the calculation of canopy vegetation indices. The OLI sensor data are most affected. The topographic correction proved to be effective for the attenuation of this effect, in both types of data. The analysis of the directional effects showed that the canopy response is significantly influenced according to the viewing direction of the observations. Data obtained from the backscattering direction showed the highest values, while the forward scattering direction presented the lowest response. These oscillations of values were found even in observations with a zenith angle of view less than 10°. Among the vegetation indices analyzed, the EVI, when calculated from reflectance measurements corrected for topographic and directional effects, presented the best performance for the analysis of canopy vegetative vigor. The analysis of the influence of topographic and directional effects in the delineation of management zones showed that the absence of such corrections can significantly change the vegetative vigor estimates, producing classes that are not very representative of the real growth rates of the forest stand. However, the influence of these effects was variable according to the vegetation index used, being more pronounced in the EVI than in NDVI and WDRVI.