TÃcnicas de Sensoriamento Remoto para Monitoramento Ambiental com AplicaÃÃes em Recursos Pesqueiros no AtlÃntico Tropical

• Autor(a) principal: Antonio Geraldo Ferreira
• Data de Publicação: 2005
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFC
• Texto Completo: http://www.teses.ufc.br/tde_busca/arquivo.php?codArquivo=2867
• topic: Monitoramento ambiental; GEOCIENCIAS

In this work we studied, from data obtained from sensors on board satellites, monthly and seasonal patterns of physical and biological parameters on the Tropical Atlantic, the area between the latitudes 30 Â N - 30 Â S and longitudes 80 Â W - 5 Â E. The physical environment (temperature of the sea surface, measured by the sensor VHRR and wind, measured by sensors and SeaWinds MAI), combined with data from clorof1la it (measured by the sensor SeaWiFS), have revealed aspects of the seasonal cycle and interannual parameters on the Tropical Atlantic. It was found that areas with more TSMs is preferably heated encounter. North of Ecuador, during almost the whole year except March, April and May, where they are positioned in the southern hemisphere. For the average monthly wind fields, it was found that the Alisios the Northeast are more intense than the Southeast during the austral winter and spring, and the converse for stations in boreal winter and spring. The images of c1orofila it showed the enrichment of bodies of surface water, especially the equatorial divergence and retroflexion of the North Brazil Current. The study of three species of tuna (Yellowfin Lage, White and mandolinist), caught in the area between 15 Â N - 15 Â S and 500 W - 15 Â W through the analysis of time series, showed that the catch per unit of effort (CPUE) for Albacora Lage tend White and non-estacionaridade while mandolinist Albacora tended to estacionaridade. The differences between the monthly and annual CPUES of the three species were studied through the use of models of analysis of variance, with relations between the CPUEs and the environmental variables listed above, were correlated, allowing the fit of simple linear models and multiple, and in some cases was applied logarithmic transformation in the value of variables.