Abstract
Rainfall is an important factor in climate definition of a locality, being the result of the whole of meteorological and geographic events. Its study is necessary due to the strong influence it exerts upon the environmental and socioeconomic conditions. Thus, is necessary to undertake studies of rainfall distribution contributing to a better socio-environmental municipal administration. Therefore, rainfall data on 42 samples of the Joinville (SC) area were used in this study, seeking to map the municipality's average yearly and monthly rainfall. At first, an exploratory analysis of the data was undertaken, followed by the homogeneity test. Afterwards, an analysis of the spatial dependence of the data through the spherical, exponential and gaussian semivariograms was conducted. Finally, the interspersed rainfall surfaces were generated by the kriging and cokriging method. To select the best surface, cross-validation and SDI analysis were used. The mapping was performed with the ArcGIS software. The results show that the spherical template, in the kriging, and the gaussian, in the cokriging, show good performance, but the kriging was the method that showed better results in validation parameters, generating an yearly average rainfall of 2,130.1 mm, and a monthly average rainfall of 183.6 mm for Joinville.
Keywords:
Semivariogram; kriging; cokriging
