Variabilidade climática do nordeste do Brasil utilizando técnicas de análise harmônica e componentes principais

Abstract

Presently, the data of the Reanalysis Project of the National Centers for Environmental Prediction (NCEPINOAA) are largely used without a critical overview, due to the scarcity of data, notably in tropical regions. In this work, a study of the climatology and the spatial and temporal variability of the NCEP Reanalysis data, was performed using Harmonic Analysis and Principal Component Analysis for Northeastern Brazil (NEB). One of the particular aspects of this study was to verify the accordance of this dataset with observed data of some climatological stations located at NEB airports, aiming at establishing a possible relationship between them as well as relating them to global atmospheric índices such as sea surface temperature (Niño 3 and Niño 3.4), Southern Oscillation and Multivariate ENSO indices among others. The monthly data used were rainfall rate, precipitable water and OLR in the period 1980-1999, for the 20°N-20°S and 20°E-80°W window. The observed rainfall data were from the Serviço de Proteção ao Vôo (SPV) network along the eastern coast of NEB. For the Harmonic Analysis, all meteorological variables above were used and the Principal Components analysis was performed using a correlation matrix esembled with standardized anomalies of the same variables, with exception of rainfall rate, but filtering out the annual cycle. The NEB region presented a strong coupling with two regions near the African West Coast. The disturbances that originate in these regions seem to influence the NEB rainfall regime, once they propagated over a “duct” of Equatorial Atlantic warm water existing between the two continents. In the NEB coastal regions, the precipitable water low interanual variability might be due to the low level convergence of the sea breeze ando f the Northern and Southern Trades. Therefore, between extreme years, such as El Niño and La Niña, this variability was relatively low, suggesting that, in case of El Niños, there is no lack of low tropospheric moisture but of a dynamical mechanism to convert moisture into convection and rain. In the NEB interior, however, the variability was high, such as over Southern Piaui and Northwestern Bahia. Through the Principal Components Analysis of SPV dataset, it was possible to verify that rainfall producing systems that acted over Eastern Coast of NEB, from Salvador to Natal, were basically disstinc from the ones that acted over Southern Bahia,evidencing the contrast between the two regions. In general, Reanalysis data underestimated themonthly rainfall totals with respect to SPV rainfall data and introduced a phase lag up to three months between the Reanalysis and the observed month of maximum rainfall in wet (La Niña) years and overestimate the reduced rainfall totals in dry and El Niño years.