Avaliação de impactos ambientais da rodovia BR-242/MT (trecho Nova Ubiratã à Gaúcha do Norte) utilizando sistema neuro-fuzzy

Abstract

Responsible for cargo and individual transportation, road projects currently represent the most utilized transportation mode in Brazil. However, in view of their economic importance in the country, the construction of these enterprises also characterizes in environmental impacts source, reaffirming even more the need of environmental impacts evaluation in order to know, in a previously way, how the risks of a specific enterprise implantation can affect the physical, biotic and socioeconomic environment. However, it is up to the entrepreneur to carry out this evaluation, and, in many situations, especially during public hearings, the impacts are not presented in an explanatory manner, and, due to different situations, go unnoticed in the eyes of the population, as result of the size of the impact matrices, which constitute the most applied environmental impact assessment methodology in current developments. In light of this situation, the use of artificial intelligence techniques, especially the neuro-fuzzy technique, is able to address this public audience deficiency, as its ability to contribute both graphically and analytically with information from the interaction matrices. Wherefore, the main objective of this study was to forecast the degree of environmental impacts resulting from the implementation and operation stages of the BR-242/MT highway (Nova Ubiratã to Gaúcha do Norte stretch) through neuro-fuzzy logic, based on the qualitative evaluation of impacts identified in the interaction matrix of this project. The results of this dissertation identified 15 potentially impacting activities in the implementation stage of the evaluated BR-242/MT highway section, as well as 11 activities in the operation stage. These activities included track opening, access roads and service roads, deforestation or vegetation cover removal and land clearances, paving and earthworks, as well as pavement maintenance and conservation, among others. Regarding the possible affected environmental factors, 35 were identified in the implantation phase and 19 in the operation phase, resulting in 192 identified impacts. Concerning the estimation of the degree of impact, the surface graphics resulting from the elaborated model allowed to see the environmental impact clearly. Nevertheless, the training error found was of the order of 2.2462 x 10-7 , thus corroborating to an efficiency of the model when estimating the degree of impact elucidated by the specialists in the ad hoc consultations. Therefore, this dissertation reiterates the importance of evaluating the environmental impacts of potentially impacting ventures, as well as elucidating the importance of translating the information from these evaluations into a material accessible to the public affected by these impacts.