Dynamic of vegetal cover and land use in the city of San Francisco of Pará (Pará, Brazil) using the remote sensing technique

Abstract

This paper deals with integration of the remote sensing technique to vegetal cover and land use information about the city of San Francisco do Pará (Bragantina region, Pará, Brazil) which has a long history (more than a century) of land usage. Landsat TM5 (1995) and TM7 (1999) images were used, which were corrected and classified. The averages of gray level values were selected and used as parameters for grouping forests of three, six and ten years old (initial successional forest), 20 years old (intermediate successional forest) and 40 and 70 years old (advanced successional forest). After that, an interpretation key based on image visual characteristics (shape, texture, tonality) was created to distinguish among dense umbrophylous forest (terra firme and igapó), successional forests (initial, intermediate and advanced), pasture, crop, and exposed soil. The study revealed a decrease of 8.04% in forest areas. The land use study showed an increase of 5.80% for pasture, and a decrease of 2.81% for agricultural areas, characterizing a region that tends to show increased cattle breeding.