Remote sensing of water circulation dynamic to the Curuai floodplain/Amazon River / Sensoriamento remoto da dinâmica de circulação da água do sistema planície de Curai/Rio Amazonas

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

In this research it is hypothesized that once water circulation in floodplains is driven by the so called flood-pulse, optical images acquired at different water levels, disregarding hydrological year can be applied to capture time-space variability in water masses types flowing into the Amazon River floodplain. The main objective in this research was to develop and test a methodology for collecting and integrating data from several sources so as to limit and characterize the water types in the Amazon floodplain throughout the hydrological cycle. Amazon-Lago Grande de Curuai floodplain system, Pará State, was selected as test site. Water stage time series were analyzed and four floodplain critical states defined. Five campaigns were carried out at each floodplain state to acquire limnological, spectral and bathymetry data. An average of 200 samples was gathered at each state in area of around 1300 Km2 of open water. These data and a time series of optical satellite images were integrated into a geographic information system environment and submitted to several analytical methods such as kriging, spectral angle mapping, derivative analyses, segmentation among others. Variography analyses applied to limnological data allowed to characterize the water composition dynamic thought the hydrological cycle. Spectral data and limnological data analyses allowed characterizing the flood pulse effect on water type distribution at the four floodplain states. At two states (overflow and falling stage) water masses are characterized by high chlorophyll concentration whereas at the remaining states (low and rising stages) inorganic particles are dominant. Landsat/TM images allowed quantifying the area occupied by water types dominated by chlorophyll, suspended inorganic particles and dissolved organic matter. The integration of all the data allowed estimating that an average of 80 % of open water is dominated by white water through the hydrological cycle. Bathymetric data processing allowed developing models to estimate the water volume stored in the floodplain at several water stages. These data were compared to estimates derived from time series of optical images showing that they provide an alternative method for volume estimation. The information gathered during the research allowed to accept the hypothesis and to propose a conceptual model for the floodplain circulation.

ASSUNTO(S)

water circulario cupuai (pa) sensoriamento remoto curuai (pa) remote sensing variação espaço-temporal circulação da água annual variations amazon river rio amazonas

ACESSO AO ARTIGO

Documentos Relacionados