PREENCHIMENTO DE FALHAS EM BANCO DE DADOS PLUVIOMÉTRICOS COM BASE EM DADOS DO CPC (CLIMATE PREDICTION CENTER): ESTUDO DE CASO DO RIO SOLIMÕES-AMAZONAS

Abstract

This work has as objective filling out flaws in meteorological variabledata(rainfall) ensuring greater quality control, from 1991 to 2007, throughcorrelation with a brief data obtained from CPC (Climate Prediction Center). Theinteraction of atmospheric dynamics as a complex process requires accuracywith regard to organization and tabulation of data to infer a characteristic on theatmosphere. Therefore, the approach discussed was to analyze the data ofrainfall observed in the series (INMET), the synthetic series (CPC) for sevenmeteorological stations along the gutter of Solimões-Amazonas river. The datawere organized in spreadsheets of Excel for better tabulation. Daily files ofrainfall were collected over South America, based on data interpolated at gridpoints of 1° x 1° latitude and longitude to Brazil, using a modified Cressman(1959) scheme. A routine was created, organizing the series; the totals permonth were calculated and finally, the synthetic series built. The statisticalprocedures were based on the linear regression method. As a result, somevalues with a correlation coefficient between 0.60 and 0.90 were identified,making the filling out of flaws possible.