Análise geoestatística da precipitação pluvial no Estado da Paraíba.

Abstract

The objective of this work is to study the spatial variability of mean monthly rainfall of the Paraíba state using geostatiscal techniques. The data were provided by the Academic Unit of Atmospheric Sciences (UACA) of the Federal University of Campina Grande (UFCG), collected from weather stations and rain gauge located in 102 municipalities in the period 1962 to 2001. The semivariogram determined the spatial dependence of the variable under study, and then was carried out a descriptive analysis to summarize the time series and describe it in numeric terms. According to the results, all variograms showed a strong spatial dependence (IDE 75%), except for the months June to Septem- ber that showed trends in the time series and difficulty in adjusting the models of spatial behavior. According to the descriptive analysis of the data the variation coefficients showed high dispersion (CV > 20%), indicating high variability of rainfall. Maps were made by ordinary kriging for the months of January, February, October and November. The criterion was based on the values of the coefficients of determination (R2 > 93%) to obtain maps unadjusted, adjusted by the model and the differences between the unad- justed and adjusted. The Mann-Kendall test was used to analyze the presence of trends in time series and regionalize the spatial distribution of rainfall in the state with good ac- curacy. The trend surface analysis estimated the values of precipitation and the residues. The surface of the best fit to the data was the 4 th order with coefficients of determination (R2 > 80%). The analysis of variance justified the adjustment and polynomial regression surface of fourth order had been accepted with significance levels of 1% and 5%. The maps of the vectors identified the regions with the highest rainfall trends. The graphics of the residuals were important to assist in identifying sites that are not well explained by the surface model of 4th order. In summary, the geostatistical has proven an important tool for analyzing and estimating monthly rainfall data.