Caracterização espaço-temporal de secas utilizando ondaletas e o standardized precipitation index: uma aplicação para a parecela mineira da Bacia do Rio São Francisco

Abstract

This research is focused on space-time analysis of droughts, based on rainfall data. In order to obtain time series of droughts, the Standardized Precipitation Index (SPI), a rainfall based drought index was used. The spatial analysis of the phenomenon was based on the regional analysis with L-moments. For the time series studies, the continuous wavelet transform was used. This technique shows up as an alternative to Fourier analysis, to identify periodicities in time series and to find variations of these frequencies over time. Also was sought to investigate possible influences of El Niño Southern Oscillation (ENSO) and Pacific Decadal Oscillation (PDO) on the time series of drought index obtained. It was proposed to incorporate the regional frequency analysis with L-moments for the calculation of the SPI, in order to overcome the necessity of long historical records of rainfall data in order to obtain a good estimative of the SPIs time series. Also a variety of probability distribution functions were tested, besides Gama. In addition, the regional frequency analysis came to define statistically homogeneous, which in general are spread over large areas whose sizes are inversely proportional to the mean values of rainfall. The continuous wavelet transform did not identify an increased frequency of occurrence of droughts in the study area over time. Also, through use of this technique, it was not possible to identify increased intensity of the phenomenon in the study area in recent years. Moreover, the technique did not allow making a clear connection between the droughts in the study area and macro climatic indices used (ENSO and PDO). Yet somehow, it was realized that these macro-scale phenomena may influence the intensity of drought on a local scale, as it was observed persistent cold phase ENSO and PDO in the most extreme drought events in the study area, corroborating the thesis that, when ENSO and PDO are in phase, weather extremes are more intense in South America.