Estudo da variabilidade da chuva no Nordeste do Brasil utilizando a transtormada de ondeletas.

Abstract

The Brazilian Northeast region (NEB) has rainfall that varies according to the time systems operating in each area of ​​the Region. Using series of monthly average totals of rainfall for the period 1985 to 2018 (34 years of data) from 244 stations meteorological/pluviometric stations, cluster analysis was applied and seven homogeneous regions (RH) in relation to seasonal and interannual variability of the regime NEB rainfall. Then, the Transformed into Wavelets (TO) was applied to the totals monthly averages of precipitation, Long Wave Radiation (LOR) in each RH, and the Temperature of the Sea Surface (SST) in the areas of the Equatorial Pacific and Tropical Atlantic Oceans, for modulate the periodicity of the rains, and the phenomena that inhibit or cause them. In In the second stage, the Transformed into Cross Wavelet (TCO) was used to find the phase covariability between the variables. The techniques used made it possible to establish a relationship between precipitation, NOR and TSM in each of the regions, as well as diagnosing the predictive capacity of regional precipitation behavior. The TO and TCO spectra obtained are quite similar, since there was agreement between the performance scales, with the the most significant annual cycle in terms of energy and breadth across all HRs, in addition to scale year, interactions were observed with other time scales (intraseasonal, seasonal, biannual and interdecennial). The scale between 4 to 8 years was also highlighted in the spectra in some HR, modulating mainly the ENSO event years. The effects of ROL and TSM on rainfall is non-linear, as there were phase changes between the series over time.