2019-04-172019-04-172014-01-20SILVA, Cássia Camila Silva da. Previsão de raios através de índices termodinâmicos da atmosfera, em torno de Belém do Pará. Orientador: José Ricardo Santos de Souza. 2014. 99 f. Dissertação (Mestrado em Ciências Ambientais) – Instituto de Geociências, Universidade Federal do Pará, Museu Paraense Emílio Goeldi, Empresa Brasileira de Pesquisa Agropecuária, Belém, 2014. Disponível em: http://repositorio.ufpa.br/jspui/handle/2011/10978. Acesso em:.https://repositorio.ufpa.br/handle/2011/10978The Amazon region often has most of its surface covered by cumulus type clouds which, to thrive, need a thermodynamic forcing (source of heat and humidity) and dynamic forcing (wind). The city of Belém-Pa is located in one of the rainiest regions of the Amazon, due to the strong convective activity in the region. One of the natural phenomena, besides precipitation, which causes destruction and losses to the population of the Amazon region is lightning, usually associated with the thunderstorms. The overall goal of this work is to create a statistical model based on the decision tree method, which consists of a computational intelligence technique, in order to find a relationship between thermodynamic indices of the atmosphere and the occurrence of lightning. We used data from radio soundings of the atmosphere made daily from Belém airport and published at an Internet site by the University of Wyoming. Additionally, lightning daily occurrences within a 100 km radius area around the sound launching site were drawn from the STARNET lightning monitoring network data base; which covers the entire Amazon. The results showed that the rules of the decision tree are significant because the critical values of thermodynamic indices of the model are similar to those described in the literature. The decision tree model establishes rules that can help in forecasting weather and lightning in the region ofsurrounding Belém. The study of atmospheric thermodynamic indices versus the occurrence of lightning showed a positive relationship where the evaluation of the model obtained a result of 94% correct values. This is considered good for building a statistical model representation for the occurrence and severity of electrical discharges from the atmosphere.Acesso AbertoRaiosÁrvores de decisãoBelém - PADissertaçãoCNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIASINTERAÇÃO CLIMA, SOCIEDADE E AMBIENTECLIMA E DINÂMICA SOCIOAMBIENTAL NA AMAZÔNIA