Evaluation of performance of non-linear geostatistical estimation methods to estimate tonnage-grade of the Chah-‎bashe iron deposit, Yazd province‎

In this research, non-linear geostatistical methods containing block ordinary log-kriging, indicator kriging and sequential ‎gaussian simulation as well as inverse squared distance from statistical methods (for comparison) have been employed ‎to evaluate grade and ore reserve of the Chah-bashe iron deposit, Yazd provice. To achieve this goal, first statistical ‎studies of assay data of exploratory boreholes were performed through determination of descriptive statistics, drawing ‎the histograms and cumulative probability distribution diagrams of the data. For modeling the deposit, variography ‎was also carried out through drawing various directional variograms using SGEMs software, MATLAB functions and ‎short programming in this environment. Afterward, 3-D block modeling of the deposit, determination of average grade, ‎ore reserve estimation and plotting tonnage-grade diagrams were performed using Datamine software. The results of ‎the research show that sequential gaussian simulation, indicator kriging, ordinary log-kriging, and inverse squared ‎distance‏ ‏methods have the minimum estimation variance, respectively. Although all selected methods are the most ‎accurate geostatistical methods, however the results of sequential gaussian simulation method by applying various ‎filters on the data such as using residual maps, is most highly accurate and more reliable so has more validity. ‎