Spectral analyses applied to the identification of optical active components at Amazon aquatic systems / Análise espectral de sistemas aquáticos da amazônia para a identificação de componentes opticamente ativos

AUTOR(ES)
DATA DE PUBLICAÇÃO

2002

RESUMO

The use of remote sensing in water quality researches is important to provide data for temporal and spatial analysis in large areas. It is necessary to increase the knowledge on water spectral properties to support a more efficient application of hyperspectral sensors. The objective of this study was the analysis of spectral data derived from using field spectrometry to identify key components for the determination of water quality, as Total Suspended Solids and Chlorophyll concentration, by using statistical correlation and derivative analysis procedures, and, also the spectral continuum removal method. The study area, where samples were collected, includes some rivers (Solimões and Negro) and lakes (floodplain and "ria" lakes) from Amazonian basin. The database contained about 60 samples, identified as Black and White waters, according to visual inspection in the field. Surface water samples were collected within the Secchi Disk penetration depth, which represents the region of larger impact on the remotely sensed reflectance. The following variables were determined from water samples: Total Suspended Solids (mg/l), Chlorophyll concentration (µg/l) and Dissolved Organic Carbon (mg/l). The radiometric measurements required to derive the Bi-directional Reflectance Factor (BRF) were obtained with a 256-channel rapid-scanning spectroradiometer covering the spectral range between 400 nm and 900 nm with 2nm spectral resolution. Based on the spectral curves derivative calculation, scattering and absorption features were identified. The derivative spectra displayed two peaks related to the absorption and scattering of chlorophyll, specially observed in Black water samples. The ratio between the reflectance in the peaks of 658/690 nm, was calculated, and correlated with limnological data. This analysis allowed identifying as Clear Water samples former identified as Black Water. That showed that preliminary visual classification was not precise. The derivative analysis allowed, also, identifying a well-pronounced feature, in 720nm, related to chlorophyll concentration. Some samples presented errors in laboratory analysis concerning the determination of chlorophyll concentration. That made it difficult do carry out a more comprehensive statistical analysis. The spectral continuum removal method, although superficially assessed , allowed identifying absorption features, which indicates the possibility of its use in future water studies.

ASSUNTO(S)

água sensoriamento remoto water amazon region (south america) amazônia remote sensing reflectância reflectance

ACESSO AO ARTIGO

Documentos Relacionados