Mapeamento do uso e cobertura vegetal do município de Floresta-PE

Abstract

The loss of diversity of forest communities by human actions has been reducing dramatically the natural areas of the Caatinga biome, thus increasing the concern to evaluate changes in usage patterns and land cover. This study aimed to mapper the dynamic of change in land use and forest cover in the last 21 years (1987-2008), municipality of Floresta, Pernambuco Outback aiming to subsidize forest management, beyond to monitor the transformation processes in which the town has experienced over the period. It had been used LANDSAT 5 satellite images for the periods May 1987, 1997 and 2008. In addition, SPRING application were utilized for preparation of base map data and digital processing of images. The images were segmented and classified by the algorithm Bhattacharya, in order to produce thematic maps in the following land uses: dense vegetation, semi-dense vegetation, bare soil, agriculture (crop / livestock), water bodies and riparian vegetation. Afterward were made reliability analysis of supervised classification by the Kappa index, which based on general performance, obtained a probability greater than 98% of classes had been mapped properly. Then was obtained maps of the years 1987, 1997 and 2008. For a period of 21 years the forest cover decreased from 42.21% of the total territory in 1987 to 27.38% in 2008. The exposed soil had a gain of 8.61%, the agricultural class has accumulated a gain of 5.75%, water bodies, over the time, showed an increase of 4.03% and the riparian forest had a tendency to fall over the study periods with a total loss of 3.57%. These data reflect the changes that the council has been going on forest land use. The decrease of forest cover in the town of Floresta-PE between 1987 and 2008 was the main consequence of the interaction of climate and human impact forest resources, urban expansion requirement and agricultural activities. The results demonstrate the potential of Remote Sensing techniques and GIS mapping of land use, for its convenience and speed, may be useful to support sustainable forest management aiming at definition of subsidies to public policy.