USO DE DIFERENTES MÉTODOS PARA O PREENCHIMENTO DE FALHAS EM ESTAÇÕES PLUVIOMÉTRICAS.

Studies using historical weather data, usually, is the problem of having gaps in historical series. Rainfall is a climate variable widely used in several studies related to social and environmental dynamics. Thus, the objective of this paper was to present a study of the use of different methods of gap filling in the data rainfall, for a better use of existing data in the historical series. The methods used were: regional weighting, simple and multiple linear regression and weighting based on linear regression. Therefore, was used data for four rainfall stations located in coastal plain and mountain, and four rainfall stations located in the plateau, both in the region of Joinville (SC). Was performed the gap filling of june 1987 to june 2006 and august 1976 to august 1993. The results showed that the method of multiple linear regression presented the most genuine results, demonstrating be the method more appropriate for the study region; followed by regional weighting method, regional weighting based on linear regression and linear regression. It should be noted the importance of whether to know past dates of regional rainfall distribution and always utilizing more than one method for verification to validate the results. [ABSTRACT FROM AUTHOR]