Aplicação da geoestatística na avaliação do comportamento dos corpos auríferos na região do vale do Piancó-PB.

Abstract

The occurrence of gold in Cachoeiras de Minas located at the extreme southwest of Paraiba was discovered in 1864 followed by exploration in the auriferous regions of Pernambuco and Paraíba. Gold is concentrated mainly in quartz veins in agreement with the general foliation of the enclosing rocks. The veins are made of white quartz, often fractured and impregnated with iron oxide. THE systematic sampling performed in channel and chip form sample "covered all came from Farias and Covico, main shafts Metal bearers. In order to analyze continuity in mineralization and homogenize the sample data for the estimates, we used geostatistical techniques with their tools. The modeling of variograms were made considering the effects of levels extremes in the samples analyzed in this study. The elimination of some high grade samples resulted in more appropriate. Variographic studies revealed the presence of structures mineralization in quartz veins, with the nugget effect varying from 1.3 to 1.7, dispersion variance from 1.5 to 2.4 and the range between 11 0 and 1 85 m. In the cross-validation of data it was observed that the average estimated was close to the sample mean, however the difference between the observed and estimated variances reached a significant value. The elaborated variogram models show the influence of structures of shaping and shaping of shafts in metal distribution. Soil gold variogram modeling associated with veins showed the presence of structures with a nugget effect of 3.2 ppb, variance of 2,5 and 1,500 m range. The distribution of metal in soil and directly related to quartz veins. The contents estimated gold in the soil correspond to the sample levels but ha significant difference between estimated and observed variance. The high values ​​of estimated variance are caused by the extreme variations in sample contents. The kriging of the points and influenced by the high levels resulting in the excessive estimation of variance and estimated content. Co-kriging may decrease the influence of high contents and reduce the estimated ore variance.