2015-08-272015-08-272009LOPES, Marcio Nirlando Gomes. Aspectos regionais da variabilidade de precipitação no estado do Pará: estudo observacional e modelagem climática em alta resolução. Orientador: Everaldo Barreiros de Souza. 2009. 117 f. Dissertação (Mestrado em Ciências Ambientais) - Instituto de Geociências, Universidade Federal do Pará, Museu Paraense Emílio Goeldi, Empresa Brasileira de Pesquisa Agropecuária, Belém, 2009. Disponível em: http://repositorio.ufpa.br:8080/jspui/handle/2011/6851. Acesso em: .https://repositorio.ufpa.br/handle/2011/6851The objective of the present work was to join different networks of surface weather stations to build a new integrated database, which it was generated a recent precipitation climatology (1978-2007) for the Pará state in high spatial resolution - 30 km, allowing better to identify the regional climatic variability that is influenced by the physiography aspects as well as large-scale climatic mechanisms of the Pacific and Atlantic Oceans. In addition, it was established an optimized configuration of the regional climatic model RegCM3 using two different cumulus parametrizations: RegCM3/Grell and RegCM3/MIT. 26 simulations (1982/83 to 2007/08) during the rainy season in Eastern Amazon (December to May) for each scheme of convective parametrization, using 30 km of space resolution, were generated. The results showed that the model is able to capture anomaly rainfall signs in the presence of extreme climatic forcing, as the El Niño-Southern Oscillation and the Atlantic dipole phases. RegCM3/MIT obtained better skill in the central region (Altamira) and reasonable performance in the Northeast (Belém), East (Marabá), Southeast (Conceição do Araguaia), and Northwest (Tiriós) sections over Pará state. RegCM3/Grell obtained reasonable skill in the Northeast, East, Southeast and Northwest areas. The North section (Macapá) was the most problematic region, with little or no sensibility presented by the model. Although RegCM3 has obtained reasonable results in most of the domain, systematic errors were detected in the regional simulations, overall those related to dry bias for RegCM3/Grell and wet bias for RegCM3/MIT in the South portion and dry bias in the North portion. These founding‟s denote the need of convection schemes adjustments to the regional conditions in eastern Amazon.porAcesso AbertoClimatologiaPrecipitação (Meteorologia)RegCM3Pará - EstadoAmazônia brasileiraDissertaçãoCNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS::METEOROLOGIA::CLIMATOLOGIA