Abstarct: Identifying areas of probable potential is one of the goals of geochemical exploration. If such areas exist, detection of anomalous areas is important. baxsed on geochemical investigations, determining minimum anomalous or threshold values is always very important in order to identify anomalous areas. There are several statistical methods for separating and determining the anomalous areas from background. These methods are generally divided into two groups: structural and nonstructural methods. U-spatial statistics and probability diagram modeling methods were used as a structural method in this thesis. Structural methods take into account the position of sampling points and their spatial relationship without expert opinion in order to statistical analysis, separating the communities and identifying anomalous. Geochemical data expressed as a compositional data including closed number system due to presence of outliers and the skewness in the distribution of data. Compositional data consist of several variables that sum of variables amounts is usually fixed and equal unit. This means that if the value of one variable increases, the values of other variables had to be reduced. By calculating the correlation coefficient of closed number system, an increase is seen in the values of closed number system which is false and unreal. The existence of features of these data, prevent using of standard statistical techniques on them. Therefore, data must first be removed from this mode and then other statistical and geostatistical analysis be done. There are several methods for transforming data from closed to open mood such as; additive logratio transform, centered logratio transform and isometric logratio transform. Since in mining activities, the samples that are taken from the environment are very small for research, therefore, having different modes of deposit instead of an average picture can be a good help in deposits modeling. Geostatistical simulations could produce various models of the deposit with the probability percentage of each. A geostatistical simulation has various algorithms which each of them provide the best answer in certain circumstances. But the sequential Gaussian simulation method is the most practical way to interpret that has high flexibility. In this study, Lithogeochemical data were collected systematically from region of Kerman Baghqloom, and then elements of closed number system turned to open number system using conversion on analyzed elements. Using variography of open data in order to determination of anisotropy directions, simulation of certain elements was done in these directions. Anomalous areas was detected from background using U-spatial statistic method on simulated data. The results of this method are compared with probability diagram modeling outcomes in order to reaching exact results.