Tendências e extremos da pluviosidade na região hidrográfica do Rio São Francisco

Abstract

Studies and knowledge of the climate are fundamental for planning and managing the use of natural resources for economic activities. In the context of climate change, it is expected that the increase in the average temperature at all scales will generate impacts on the hydrological cycle and, consequently, affect the entire natural and human system. As rain represents the most important phase of the hydrological cycle and is one of the essential elements for life, it is increasingly important scientific studies that identify and analyze possible variability and/or change in the pluviometric rhythm. In this perspective, this research has as it’s general objective to study the trends and rainfall extremes in the São Francisco River Hydrographic Region, which in turn has relevance in national and regional/local contexts and because it presents adversities in terms of hydric availability and demand. It is believed that this hydrographic region is going through an increase in temperature, a decrease in rainfall and the presence of climatic extremes (mainly with more severe droughts), aggravating its situation as a whole, but mainly in the semiarid part, causing significant impacts for the populations the inhabit it. For this purpose, were used data from 124 pluviometric stations made available by Agência Nacional de Águas (ANA), whose analyzed period was from 1976 to 2015 (40 years). First, the statistical treatment of the pluviometry data was carried for knowledge of the rainfall variability. After, other appropriate mathematical and statistical tools were applied for the study of pluviometric trends, such as: Linear Regression Analysis, Mann-Kendall Test and Signal Test and, to identify and evaluate possible extreme events, the Standardized Precipitation Index – SPI (or Índice de Precipitação Padronizada – IPP) e o Rainfall Anomaly Index – RAI (or Índice de Anomalia de Chuva – IAC) were used, besides of the confidence interval analysis (maximum and minimum) calculated from the standard deviation and the average. The results confirm that there is a great spatial variation of pluviometric in the study area, that is, a decreasing standard of the rainfall from of its springs to the mouth of the São Francisco River. The average annual rainfall of the study area is around 950 mm, varying along the sub-basins from less than 550 mm to just over 1400 mm, emphasizing that the highest averages of the annual totals are located in the state of Minas Gerais and the lowest averages in sub-basins 47, 48 and 49, located in the states of Bahia, Pernambuco, Alagoas and Sergipe. It was also found that most of the sub-basins concentrate their rains between October and March, in exception of the sub-basins 48 (between December and May) and 49 (between January and August). The months with the highest monthly averages in the study area are January and December and the months with the lowest monthly averages are August and September. The episodes above of the maximum intervals of the standard deviation in the hydrographic region occur in the month of April and the month of March has the highest number of episodes below the minimum interval of the standard deviation. The SPI results indicate that 45% of the total monthly occurrences are in the normal class (N) and that the difference in percentage of the summation of the positive and negative extremes is only 5%. And in the RAI products, the very dry class (S2) is the one with the highest percentage of occurrences in the study area, besides that, the sub-basin 47, in particular, presents in this class more than 49% of the total events. About the tests of Linear Regression, Mann-Kendall and of the Sinal, the results show a reduction of the rainfall in the hydrographic region in the time interval of this study. Between the 12 months of the year, 11 showed a predominance of decreasing rainfall in Linear Regression and 10 months in each of the other two tests. Moreover, the months with significant occurrences of reduction of the pluviometry in the 3 tests are: January, February, July and September, in addition to the annual total. The occurrences that were above the maximum interval and below the minimum interval of the standard deviation per decade indicate that there is a decrease in the number of events above the maximum interval, which corresponds to a greater homogeneity of data over the decades, in addition, it was not possible to make deductions about the occurrences below the minimum intervals, due to the randomness of the results. About the SPI and RAI products per decades, in general, it was identified that the reduction of the positive extremes and the increase of the negative extremes is more easily detected in RAI than in SPI. The decrease of the SPI occurrences was also visualized in the results of the RAI and the increase of the negative extremes were verified more clearly in the RAI method than in the SPI.