Environmental fragility mapping using geospatial analysis: an application on Piracicaba and Sapucai upper river basins, Fernão Dias EPA, Mg, Brazil.

Abstract

The aim of this paper is to present a methodological approach based on geospatial analysis techniques and GIS, for environmental fragility mapping. This procedure uses the following geoenvironmental variables: NDVI vegetation index, heavy rains probability, terrain slope, structural lineaments density and road density. The weights assigned to these variables are estimated using D parameter, calculated from the Kolmogorov-Smirnov test. This test uses the geoenvironmental variables above cited, mass movements and gullies mapped on Google Earth images and checked in the field. The geoenvironmental variables maps were converted to the fuzzy values using linear membership functions. This model was tested in the Fernao Dias environmental protection area (EPA), located at southern Minas Gerais state. The results indicate that the areas located in the center-east of the EPA are the most fragile, even having significant remaining forest areas. In the west of EPA, mainly nearby Fernao Dias highway, the high fragility areas occur where urban sprawl is not planned, in croplands, degraded pastures and in the no-pavement roads high-density areas.