Análise da precipitação na Bacia do Rio São Francisco

Abstract

Precipitation is the most important part of the water cycle and therefore its study is of fundamental importance, since it allows evaluating the water availability and its temporal distribution over the observed period. The study of precipitation in watersheds assists in various planning associated with the use of available water in these basins, but also promotes the discovery of possible environmental disasters associated with excessive rainfall or periods of drought. In this context, the objective of this work was to study periods of drought and rainy periods in historical series of precipitation of the São Francisco River Basin (BHSF), using several techniques such as the Standardized Precipitation Index (SPI), Markov chains, Maximum Precipitation Probability (PMP) and Sliding Window Analysis (SWA). The series correspond to daily records of rainfall of 46 years duration, which runs from 1970 to 2015. The results obtained with the application of SPI and Markov chains showed that it is possible to identify the years of droughts due to the low values of precipitation accumulated in the time scales of six and twelve months, besides calculating the time of return and the expected time of permanence for diverse climatic conditions, what made possible to know if a certain condition will occur in the near future, given the current climatic condition of the area. The adjusted probability models and the goodness of fit criteria showed that the selected distributions and the return time, in years, for each station were: GPD for the Iguatama station with 59 years of return time, BurrXII for Pirapama with 55 years, GEV to São Romão with 100 years, GGAM to Colônia do Formoso with 259 years, GPD to Pilão Arcado with 229 years, Dagum to Fazenda Tapera and Propriá, with 157 and 63 years, respectively. With the use of sliding windows, the results showed a growth trend from PMP to Pilão Arcado and Propriá and decrease to Iguatama and Fazenda Tapera. The techniques used were important for the analysis of the precipitation variable and allowed to study the behavior of each station chosen, associating with the climatic peculiarities present in each subregion.