Proposta de método para análise da fragilidade ambiental na bacia hidrográfica do rio Jordão-MG com uso de redes neurais artificiais

Abstract

This research appears with the intention of intending a method more effective, fast and economical, for the environmental studies. Studies as the one of environmental fragility deserve cares and a wide knowledge, of several areas of the knowledge, since it should take into account a series of natural and anthropics variables. The landscape of great areas, as a hydrographic basin, consists of a complex environmental system and, therefore, worthy of methodologies that get to analyze her entirely, in order to identify your current situation and to propose solutions in critical cases, that threaten the environmental quality (human + nature). One of the justifications of that study is the lack of information about BHRJ, that possesses several and significant enterprises, as the hydroelectric plants, features scenic-landscape, diversity of use and occupation of soil and preservation areas. The present study had as objective identifies and to classify the environmental fragility in the basin of the river Jordao/MG (BHRJ) starting from the application of Artificial Neural Networks. They were made analyses of the result of RNA starting from each variable considered in the determination of the landscape sensitivity of the BHRJ, involving precipitation and superficial temperature, as climatological variables; hypsometry, slope, curvature and hillshade, as geomorphological variables; geological lithotipes and units, and distance of fractures, as geological variables; types of soils, as pedological variable; distances of urban areas, as urban variable and use and occupation of the soil as anthropic variable. To leave of those results the method it was validated proposed as viable for analysis of complex systems, since it got to generate a result here being taken in consideration all the chosen variables for use, in smaller time and with smaller cost. In the general, the result obtained with use of RNA it indicated fragile areas for BHRJ which also correspond to fragile areas in reality, considering the predominance or of an unstable geology, or of the absence of covering and protection of the soil, among other characteristics and features that can actually make the environment more or less fragile, especially in terms of sliding and/or erosion.