Análise espectral e em componentes principais: uma visão conjunta à identificação e caracterização de oscilações e ruídos atmosféricos no AZP-AL.

Abstract

Mathematical and statistical techniques like the principal components analysis PCA and the wavelets transformed WT were applied to hourly data of wind on the surface of the Zumbi of Palmares International Airport AZP-AL, between 1991-97, located in a region of coastal board in Brazil – Alagoas – Maceió. In turn, the together analysis of the PCA and the WT is motivated by the fact permit to observe the seasonal and instantaneous variability of the components of the wind on the surface and link them to the typical weather phenomena and systems. |The study is the characterization and analysis of the energy peak of the global spectrum associated with systems operating in the region, with statistical significance 10%. The temporal distribution of theses significant peaks are characterized and analyzed in scalegrams. The results of the temporal distribution of the spectral energy peaks associated with the actions of weather systems are compared to the results of the PCA in the search for seasonal patterns of occurrence. The first principal component that explains a little more than 51% (zonal component) and 43% (meridional component) of the variance evidence the activities of the east waves and breeze. The second principal component that explains a little more than 46% (zonal component) and 41% (meridional component) of the variance appears to be associated with the intertropical convergence zone ITCZ, the convergence zone of the South Atlantic CZSA, the Cyclonic vortices of high levels CVHL and the Madden-Julian Oscilation MJO. The scalegrams show that the energy spectrum associated with significant Systems Front Similars oscilations does not agree to seasonal pattern of occurrence described by the ACP. The spectral analysis found that daily cycle and Madden-Julian Oscilations are the systems of higher energy spectra show seasonally well defined implying that these are the systems that most affect the local circulation of the wind on the surface. Is observed that the results of the PCA and WT seem to be antagonistic about the daily cycle, to zonal component. So that the period of highest prevalence of occurrence, involving the rainy season, to daily cycle coincides with the lowest energy spectrum observed over the years. But in fact the results are convergent and complementary, because the lower energy are associated with the night convergence that intensify in this period. In this study are found surprising results, beyond the expected, such as: the existence of significant noise typical of the ENSO oscillation in all the years of ST, amplifying it in 1994; the fact that the energy peaks associated with MJO in some years, has greater influence on variability of the wind on the surface comparative the daily cycle.