O trabalho explora as relações espaço-temporal entre a série da razão isotópica do oxigênio dezoito (δ 18 O) do testemunho de gelo do Nevado Illimani (Bolívia) e a série da precipitação na região tropical da América do Sul, entre o período de 1929-1998. Com intuito, de melhorar a representatividade estatística da reconstrução da precipitação pretérita nessa região, usando o δ 18 O. Foram utilizados dados do δ 18 O dos primeiros 50 m deste testemunho e 37 séries da precipitação observadas no Brasil e na Bolívia. Adicionalmente, as médias mensais da precipitação foram calculadas e subtraídas das amostras, gerando as anomalias. A análise de dependência abrange as seguintes etapas: elaboração dos diagramas de dispersão, identificação das equações matemáticas de regressão, mensuração do erro e mensuração do índice de correlação. Ao comparar o δ 18 O com a precipitação identificou-se que os maiores índices de correlação são de aproximadamente |0,2|, o que impossibilita a reconstrução paleoclimática da precipitação pretérita a partir de uma regressão linear. Entretanto, as relações entre as variáveis evidenciam que os registros de máximo (mínimo) fracionamento isotópico estão relacionados à seca (chuvas acima da média) na Amazônia e Centro-Oeste brasileiro e que os registros do δ 18 O mínimo (máximo) ocorrem simultaneamente secas (chuvas acima da média) nordeste brasileiro, Minas Gerais, Roraima e Amazonas. Palavras-chave: glacial tropical, reconstrução paleoclimática, ciclo hidrológico Analysis of relations between δ 18 O the ice core of Nevado Illimani (Bolivia) with the precipitations in the tropical region of South America (1929-1998) A B S T R A C TThis paper examines the space-temporal relations between a time serie oxygen stable isotopes ratio (δ 18 O), obtained from an ice core in Nevado Illimani (Bolivia), and the precipitation serie in South America Tropical, the period 1929-1998. For the purpose of upgrade the statistical representativity of the reconstruction of the preterit precipitation in a tropical region, using the δ 18 O. We used δ 18 O data from the upper 50 m of this core and 37 precipitation series was collected in Brazil and Bolivia. Additionally, the time series of precipitations were calculated by subtracting of the mean monthly of original series, resulting in anomalies. The dependency analysis has four stages: elaboration of scatter diagrams, identification of mathematical regression equations, error measurement and measurement of the correlation coefficient. By comparing the δ 18 O with precipitation was identified that the correlation indices larger are about | 0.2 |, so it is impossible makes the paleoclimatic reconstruction from a linear regression used the δ 18 O. However, relations between the variables show that the δ 18 O maximum (minimum) are related to dry (rain above the mean) in Amazon and Central Brazil and rain above the mean (dry) in northeast Brazil, Minas Gerais, Amazonas and Roraima.

RESUMOComo a América do Sul se estende por diferentes latitudes e possui formas de relevo variadas, proporciona a atuação e o desenvolvimento de diferentes sistemas atmosféricos, os quais contribuem para a não homogeneidade climática da região. Portanto o objetivo deste estudo é apresentar uma revisão dos sistemas atmosféricos que atuam nos diferentes setores do continente sul-americano e que contribuem para a precipitação. Palavras-Chave: precipitação, sistemas atmosféricos, América do Sul ABSTRACT: PRECIPITATION REGIMES IN SOUTH AMERICA: A BIBLIOGRAPHY REVIEWThe distribution of South American lands on different latitudes and its diversified topography can influence the development and action of many atmospheric systems contributing to a non-homogeneous climate in this region. This work presents a review on the atmospheric systems present in the different sectors of South America contributing to the precipitation variability in these regions.

No abstract

Peres, T. C.; Maier, É. L. B. variability on South America can be associated the remote and local factors, for example, global warning and land use change (Amazonian savannization process).

Precipitation variability in Brazil was investigated using reanalysis data from the Global Historical Climatology Network-Monthly, version 3.01, within 1920-2010, provided freely by the University of Delaware. The analysis of space-temporal pattern was conducted using principal component analysis (PCA) in temporal and spatial modes. It also intends to contribute with the literature that characterizes the patterns of spatiotemporal distribution of precipitation in Brazil. Temporal grouping allowed the identification of five spatial models that represents the spatial distribution of the precipitation, explained variance of 38%. Spatial grouping mode zoned through six temporal models of the areas with partially homogeneous precipitation, explained variance of 22.69%. In equatorial latitudes, the models identified Intertropical Convergence Zone (ITCZ) latitudinal and temporal variability in North and South Hemispheres. In tropical latitudes, the models evidenced anomalous precipitation due to South Atlantic Convergence Zone (SACZ) and South America Monsoon System. In subtropical latitudes, rainfall spatial anomalous patterns were associated to the di-pole pattern between tropical and subtropical latitudes, which shows the influence of the climatic variability modes, especially El Niño-Southern Oscillation (ENSO). Therefore, it was diagnosed that the anomalies of the atmospheric mechanisms that occur seasonally, such as SACZ and ITCZ, are the ones that cause the higher frequency in the percentage of rainfall variance, which might be explained by the frequency of the occurrence. However, remote factors, such as the occurrence in different temporal scales, interannual and interdecennial, were represented in different spatial and temporal models, as the events of variability modes vary spatially, seasonally and in intensity. It is important to highlight that the modes of climatic variability may influence the atmospheric mechanisms identified in this study, nevertheless, their impact were not evidenced in the space-temporal scales analysed. The results are relevant to understand the space-temporal variability of rainfall over Brazil.