Banca de QUALIFICAÇÃO: ESTÉFANE DA SILVA LOPES

MULTIPLE REGRESSION MODEL AND COMPUTATIONAL MODELS FOR ACOUSTIC MAPPING

Ruído Ambiental; Regressão; Uso do Solo; Mapa de Ruído.

Environmental noise has been one of the sources of pollution that most affects the quality of life in urban areas, even having adverse effects on the health of the population. In an attempt to mitigate the effects and planned actions, one of the important tools is urban mapping. The research problem, however, is that urban mapping is a costly operation that requires economic and human efforts. Therefore, this research has the general objective of creating a noise map based on multiple linear regression as an alternative tool for evaluating urban noise. To this end, the Setor Bueno region in the city of Goiânia was chosen as the study region. The methodological procedures will characterize the region through the preparation of maps (topographic, land use, afforestation, roads and template maps). After this stage, they will be carried out on site at specific points, checking noise levels, allowing the multiple regression model to be used to be developed later. From this model, noise maps will be created and compared with the maps obtained by the Soudplan@ program, using the LAeq value. As results, maps will be obtained; noise map by multiple linear regression and noise map by acoustic computational program Soudplan@. The analyzes will be carried out based on the LAeq sound pressure levels obtained by the different noise maps. Finally, we intend to propose guidelines that help better urban acoustic assessment.