Modelagem estocástica da variação espacial sob a precipitação pluvial do estado da Paraíba

• Autor(a) principal: GOMES, Oseas Machado
• Data de Publicação: 2014
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFRPE
• Texto Completo: http://www.tede2.ufrpe.br:8080/tede2/handle/tede2/5237
• topic: Precipitação (Meteorologia); Modelagem estocástica; Variabilidade espacial; CIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA

The lack of data and poor spatial distribution of stations and meteorological stations has been a major challenge for researchers attempting to describe statistical models that can explain consistently the behavior of certain meteorological parameters such as rainfall, temperature, relative humidity among other, as well as supply this lack of data for sites not sampled. A methodology able to quantify the spatial dependence and predict certain attributes is geostatistics, which has been used in an attempt to explain the behavior of rainfall the state of Paraíba, through the technique of kriging and trend surface analysis, where it was found by variogram analysis of the existence of spatial dependence of the attribute under study. Through the kriging maps could observe the places where this dependence was stronger. The trend surface analysis was of paramount importance to explain the behavior of such phenomena, some criteria were used to select these as: The Akaike Information Criterion (AIC), Analysis of Variance (ANOVA) and the adjusted coefficient of determination (R2a). According to these criteria it was noted that the surfaces trend that best fit the rainfall data from the State of Paraíba were cubic and fourth degree surfaces. Given the methods used, it was observed, for some months, certain regions had higher rainfall, as is common during these periods the performance of some meteorological systems such as the Intertropical Convergence Zone (ITCZ) and Upper Tropospheric Cyclonic Vortex High Levels (VCAN). Estimates interpolated by the selected models were fairly representative with regard to the spatial distribution of rainfall in each location during the reporting period. The maps of the residuals show exactly the places where the smallest and largest differences occurred. Analyses were performed using R software with the aid of spatial and geoR package.