Propriedades espectrais das plantas no infravermelho termal (2,5 - 14 um): da química ao dossel. / Spectral properties of plants in the thermal infrared (2.5 - 14 um): from the chemistry to the canopy.

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

This work explores the thermal infrared spectral properties of plants to evaluate how such data might be used in laboratory, field, and remote sensing studies of ecosystems, and to analyze diverse chemical aspects of plant species. Attenuated total reflectance (ATR) spectra of plant leaves display absorption bands caused by the fundamental molecular vibrations of various compounds. By comparing ATR spectra of fresh leaves to reference spectra of selected pure compounds, it was possible to assign a number of leaf absorption bands related to cellulose, cutin, silica ( quartz micro-crystalline), water and acid triterpene. By using spectral search/match procedures it was possible to locate species within a database of leaf spectra that had similar chemical characteristics. Potential sources of spectral variation were explored, including temporal, spatial, and positional variations. For example, sun leaves showed spectral differences compared to shaded leaves. Spectra of adaxial leaf surfaces were commonly different from those of abaxial surfaces. Individuals of the same species consistently showed very similar spectra. In a simulated ecological study using field ATR measurements as a tool for species identification 82% of the individuals were correctly identified. Scanning electron microscope images were utilized in conjunction with directional hemispherical reflectance (DHR) measurements of leaves to study 3-dimensional structural effects on spectral behavior. For example, small-scale structures formed by waxes on a leaf surface can cause the attenuation of spectral features due to the Holblaum (cavity) effect. DHR measurements can be linked to emissivity using Kirchhoff’s law (ε=1-R), and therefore are relevant to understanding the kinds of information concerning plants that may be available via remote sensing. Finally, to explore the effects of canopy structure on spectra, direct emissivity measurements were made by using a field spectrometer. The data show, for the first time, that it is possible to discriminate spectral emissivity features of plants from those of the surrounding environment. Spectral measurements made at increasing horizontal distances from several tree canopies showed progressive attenuation of the spectral emissivity features. This attenuation is ascribed to the increasing proportion of canopy voids in the instrument field of view, and to increased surface scattering effects. Errors associated with removal of atmospheric features also contributed to the loss of spectral information at greater measurement distances. Despite these problems, there is untapped potential for using thermal infrared remote sensing measurements to study plants. To be effective an airborne sensor operating in the 8-14 m atmospheric window would need high signal-to-noise and a small instantaneous field of view to enable measurements of individual leaf surfaces. Data calibration methods and spectral analysis algorithms would also require refinement to permit the extraction of subtle plant emissivity features.

ASSUNTO(S)

remote sensing cutícula thermal infrared cuticle infravermelho termal sensoriamento remoto

ACESSO AO ARTIGO

Documentos Relacionados