Análise da continuidade espacial do comportamento da precipitação em área que engloba as mesorregiões do Triângulo Mineiro/Alto Paranaíba - Minas Gerais e Sul Goiano - Goiás

• Autor(a) principal: Silva, Lucas Rafael Tiago
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFU
• Texto Completo: https://repositorio.ufu.br/handle/123456789/24493; http://dx.doi.org/10.14393/ufu.di.2019.612
• topic: Geoestatística; Geostatistics; Mesorregião do Sul Goiano; Mesoregion of the Sul Goiano; Mesorregião do Triângulo Mineiro Alto Paranaíba; Mesoregion of the Triângulo Mineiro Alto Paranaíba; Krigagem; Kriging; Estimação de valores de precipitação; Estimation of precipitation values; Geografia; Geography; Precipitação; Precipitation; Precipitação meteorologia Triângulo Mineiro/Alto Paranaíba MG Mesorregião; Precipitation meteorology Triângulo Mineiro Alto Paranaíba MG Meso-region; Precipitação meteorologia Sul Goiano GO Mesorregião; Precipitation meteorology Sul Goiano GO Meso-region; Precipitação meteorologia estimativas; Precipitation meteorology estimates; CNPQ::CIENCIAS EXATAS E DA TERRA::GEOCIENCIAS::METEOROLOGIA

The aim of this work is to analyze the spatial continuity from the series of annual rainfall averages and the rainfall averages in the rainy season, in an area that includes the mesoregions of the Triângulo Mineiro/Alto Paranaíba - Minas Gerais and the Sul Goiano - Goiás, as well as estimating values of precipitation to the non-sampled sites in the study area. Monthly precipitation data of 52 pluviometric stations were used, available on HidroWeb from the National Water Agency (ANA). In the initial treatment, data were organized and tabulated in order to extract the pluviometric averages. The geostatistical treatment was performed in the SGEMS / AR2GeMS software, where six experimental semivariograms were elaborated for each analyzed period, which respectively contain the following intervals of lag (h) 10km, 20km, 30km, 50km and 60km. The semivariograms were manually modeled and the resulting values were validated, so that the best adjustments for the annual averages were the 20km lag (h) with 84km range and, for the rainy season averages, the 60km lag (h) with a 250km range. Ordinary kriging and indicative kriging were performed using the adjusted optimal parameters in the variographic. The data validation was done using the cross-validation method, while the estimation errors were determined by ordinary kriging. The maps and data resulting from ordinary kriging allowed us to identify in both evaluated series that the highest precipitation values, 1370mm to 1672mm, are concentrated at the area extremities, east and northwest portions, while the lowest values, 1154mm to 1369mm, are located in an intermediate range between the two portions. In the validation and errors analysis for series of annual averages, estimates errors were determined ranging from -216mm to 217.3mm, with an average of -1.99, for the series of averages in the rainy season, estimates errors were determined ranging from -225.3mm to 201.7mm, with an average of 0.14. Through the method of indicative kriging, the probability of local data deviation around the median value was set (1452mm for the series of annual averages and 1325mm for the rainy season averages), where higher deviations from 65% to 100% were set in the southeast and northwest portions and, otherwise, lower deviations, 0% and 55% are located in the intermediate range between these two regions. In both kriging methods, it was noticed that the extremes of the study area, the southeast and northwest portions, concentrate higher rainfall heights when compared to the intermediate or central portion. Also as a result of the geostatistical treatment, precipitation values were estimated for 57 municipalities that do not have stations.